Report No. 12835-IND Indonesia's Health Work Force: Issues and Options November 28, 1994 Population & Human Resource Division Country Department III East Asia and Pacific Regional Office IJ -? P _t_ __X I- - :- 4 '4 t U-t- *.---re ._' . L- j ' ,:~~~~h World Bank.. 4- : .7 14-4- CURRENCY EQUIVALENTS Annual Average 1979 US$1.00 - Rp.623 1980 US$1.00 - Rp.627 1981 US$1.00 = Rp.632 1982 US$1.00 - Rp.661 1983 US$1.00 - Rp.909 1984 US$1.00 - Rp.1,026 1985 USSl.00 - Rp.1,111 1986 US$1.00 - Rp.1,283 1987 US$1.00 - Rp.1,644 1988 US$1.00 - Rp.1,686 1989 US$1.00 a Rp.1,770 1990 US$1.00 = Rp.1,843 1991 USS1.00 = Rp.1,950 1992 US$1.00 - Rp.2,030 1993 US$1.00 = Rp.2,087 Dec. 1993 US$1.00 Rp.2,110 FISCAL YEAR Goverment - April 1 to March 31 ABBREVIATIONS, ACRONYMS AND DEFINITIONS Akper Nursing academy BAKN Public service commission BAPPENAS National Economic Planning Board Bidan Midwife Bidan di desa Village midwives Bidan pengawa Midwifery school supervisor Bintek Technical guidance CHS Consortium of Health Sciences Dl One year certificate D3 Academy-level nurse Dinas Local government office Dinkes District health staff Dokabu Chief health officer at the kabupaten level Dukun bayi Traditional birth attendant Formasi Fully funded new government health posts GOI Government of Indonesia GPs General practitioners Guru Teacher HC Health center HMO Health maintenance organization IBI Indonesian midwives' association IDHS Indonesian Demographic and Health Survey IDI Indonesian Medical Association IMR Infant mortality rate INA Indonesian Nutrition Association ISN Indicator of Staffing Needs IUD Intra-uterine device Kabupaten District Kaders Village health volunteers Kaltim East Kalimantan KOPERTIS Regional coordination bodies for private higher education Lembaga Swadana The process of converting government hospitals and similar entities into autonomous, self-managing bodies MCH Mother and child health MENPAN Ministry for the Utilization of the State Apparatus MMR Maternal mortality rate MOEC Ministry of Education and Culture MOH Ministry of Health NGO Non-government organization NHC National Health Council Norplant Contraceptive implant NTB Nusa Tenggara Barat Pedoman Manuals/handbooks Pegawai Tidak Tetap or PTT Non-permanent employees hired on contract Petunjuk Operational Guidelines PK Basic nurse PPNI National Organization of Indonesian Nurses Posyandu Monthly village preventive health sessions run by kaders and visited by puskesmas staff PTT Pegawi Tidak Tetap (non-permanent employees) Pusdiknakes Center for Education of Health Manpower Puskdiklat Center for Education and Training of Health Personnel Puskesmas Pembantu Health subcenter Puskesmas Health center Repelita Five-year Development Plan Si University degree SPK Sekolah Perawat Kesahatan, school for basic nurses Surat Wewanang Letter of competence SUSENAS National Household Expenditure Survey TBA Traditional birth attendant USAID The United States Agency for International Development WHO The World Health Organization CONTENTS Page Executive Sulmmary ............................................. Chapter 1 The Health Scene in Indonesia .1 Introduction .1 The Government System. 2 Service Delivery. 2 Funding Constraints. 3 Changing Health Challenges. 4 Underutilization of Public Services. 6 Accounting for Health Center Use. 8 Imperfect Assessment Mechanisms. 9 Private Sector Health Services. 9 Plan of the Study .12 Chapter 2 Indonesia's Doctors .13 The Policy Framework During Repelitas 11-V .13 Medical Education .13 Training Specialists .14 Demand-side Interventions .15 Private Practice ........ ............................ 17 The Contract Doctor Scheme .18 Consequences and Issues .19 Physician Distribution: Health Centers .20 Employment of Physicians in Hospitals .23 Limitations of Compulsory Service .24 Sustainability of the Contract Scheme .26 This report was written by S. Lieberman (task manager), S. Stout and A. Nyamete. Inputs to the exercise included background papers by H. Saxenian, M. Francisco and W. Stinson, and J. Soepardi, and contributions by V. Paqueo, J. Hammer, C. Bayulken, E. Iswandi, H. Djdajadi and A. Suharto. Ms. Soepardi's paper was based on a survey of private medical practice that was commissioned by the Bank and undertaken by the Indonesian Medical Association (IDI). P. Sudharto managed the survey for IDI. Missions related to the report took place in July and October 1992, and July and October 1993. J. Coon, L. Hebblethwaite and T. Do provided word processing support. The peer reviewers were M. Lewis, F. Golladay, J-L. Lamboray and W. de Geyndt. Page Demand-Side Issues .................................. 29 Making Doctor Allocations Demand-Sensitive ................. 30 Developing a Demand-Sensitive Approach to Staff Allocations ................................... 32 Supply-Side Implications and Concerns ....................... 33 Demand-side Scenarios ................................ 33 Balancing Demand and Supply ........................... 34 Medical Education-Quality Issues ......................... 35 Ownership of Medical Schools ........................... 37 Continuing Education ................................. 37 Specialist Training: Numbers, Balance and Quality .... .......... 37 Chapter 3 Nurses and Midwives ................. ................. 41 Nursing Education ................. ................... 41 Primary Health Nurses ................ ................ 41 Nursing Academies ................ .................. 44 Teaching Inputs ................. ................... 44 The Bachelor of Science in Nursing .......... .. ............ 44 Midwifery Training ................. ................. 44 Acceleration Program ................ ................. 52 Employment Patterns ................. .................. 52 Distribution by Type of Facility .......... .. .............. 52 Staffing Norms and Availability in Government Facilities .... ...... 53 Health Center Nurses and Midwives by Area and Population Density . .. 54 Employing bidan di desa ..54 Private Practice ..55 Nurses and Private Practice ..56 Bidans and Private Practice ..57 Issues and Recommendations ..57 Availability of PKs and Government Midwives in Hospitals . .58 Can PKs and Bidans be Freed up at the Puskesmas Level? . .59 Village Midwives-Economic Conditions for Success . .60 Prerequisites for Success at the Village Level ..63 Entry Requirements and Selection Procedures . .63 Pre-service Training ..65 Inservice Training and Supervision ..67 Policy as Regards Basic and Academy-Level Nurses . .68 Career Paths for Nurses and Midwives ..68 Clarifying Roles and Competencies for PKs ..69 Directions for D3-Level Training ..70 Oversight of Nursing and Midwifery Policies ..70 Policy Formulation ..70 School Accreditation and Management ..70 Licensing and Certification ..71 Page Chapter 4 Paramedical Workers .................................. 74 Training Initiatives .................................... 74 Training Facilities ................................... 74 Enrollment Levels and Distribution ........................ 76 Employment Patterns and Trends ........................... 80 Staffing Norms and Funding Mechanisms .................... 80 Employment Outcomes ................................ 80 Issues and Suggestions .................................. 82 Staffing Norms and Needs in Government Hospitals .... ......... 82 Paramedical Needs in Health Centers .85 Improving Paramedic Performance in Health Centers .... ......... 86 Policy Towards Paramedical Training ...................... 87 Chapter 5 Health Workforce Challenges and Options .................... 89 A Target-Driven Approach ............................. 89 Accounting for Low Utilization and Efficiency ................... 90 Factors Internal to Puskesmas ........................... 90 Moving to a Facility-Centered Approach ...................... 94 The Locus of Facility-Level Reforms ....................... 95 Health Workforce Implications ............................. 96 The Focus of Workforce Policy ............................ 100 Prepare Doctors for Private Sector Employment ................ 100 Identify Priority Areas for Government Health Initiatives .... ...... 101 Require and Enable Provincial and District Governments to Play a Larger and More Effective Role in Work Force Policy and Service Provision ................................ 101 Develop Alternatives to Current Delivery Arrangements .... ....... 101 Ensure Appropriate Representation of Individuals from Low Income and Outer Island Backgrounds ........................... 101 Monitor the Distribution of Staff Across Worker Categories ... ..... 101 Initiate a Phased Withdrawal from Public Ownership and Management of Medical Education and Health Worker Training Institutions ... ... 102 Strengthen Institutional Capacity for Health Workforce Policy Making . . 102 Encourage Further Development of Professional Associations ........ 103 Bibliography .............................................. 105 TABLES IN TEXT 1.1 Public Hospitals: Bed Occupancy Rates by Hospital Class and Year ..... 6 1.2 Bed Occupancy Rates, C and D Class Public Hospitals, by Region: December 1992. 6 1.3 Average Number of Daily Visits to Health Centers by Province, 1990 and 1991. 7 1.4 Proportion of Those III Visiting Different Facilities and Providers by Region, 1992 ...................................... 11 2.1 Public and Private Medical Graduates, 1950-92 .................. 14 2.2 General and Consulting Specialist Production, 1980-1992 .... ........ 15 2.3 Average Number of Days per Week and Hours Per Day in Private Practice, by Region, 1993 ..... ............................. 18 2.4 Average Fees Charged by Private Practitioners by Region, 1993 ........ 19 2.5 Average Monthly Income From Private Patients For Private Practitioners by Region, 1993 .20 2.6 Population and Doctors per Health Center by Province, 1985 and 1992 . .. 21 2.7 Clustering of Health Center (HC) Doctors by Province, 1991 .22 2.8 Doctors Employed in Publicly Operated General Hospitals, December 1993 .23 2.9 Geographic Placement of Graduating Specialists in Four Fields, 1989-1993 . ........................................ 24 2.10 Health Center Doctors: Time in Current Assignment by Province ....... 26 2.11 Distribution of Civil Service and Contract Doctors by Region ......... 28 2.12 Fifth Batch (October 1993) of Contract Doctors: Gender Specified in Provincial Requests .................. .................. 28 2.13 Utilization of Health Centers ............... ............... 30 3.1 Number of Nursing Schools and Estimated Output, by Type and Ownership, 1984-1993 ..42 3.2 Number and Output of Various Class Sizes of SPK, by Province, 1993/94 . 43 3.3 Estimated Number of Students per SPK Teacher, by Ownership and Province, 1993/94.. 45 3.4 Educational Attainment for Full Time Teachers in SPKs, 1992/93 46 3.5 Student/Teacher Ratios in D3 Academies, by Province and Ownership, 1992/93 . .......................................... 47 3.6 Number of Bidan di Desa Graduates, 1989/90-1992/93 .... ......... 48 3.7 Distribution of Nurses and Midwives (NMW) by Type, Facility, Ownership and Region, 1992 .49 3.8 Country-wide Distribution of Nursing Personnel by Type and Level of Facility, 1992 ........... ............................ 50 3.9 Norms for Number of Nurses by Type of Facility, 1992 .... ......... 51 3.10 Comparison of Average Number of Nurses per Facility to Govermnent Staff Norms, by Region and Hospital Class, 1992/93 .54 3.11 Differences Between Number of Nurses per Health Center (HC) and Current Staffing Norms, by Type and Region, 1992 .55 3.12 Distribution of Health Center (HC) Nurses and Midwives (NMW) by Area, Population Density and Number of Villages, by Region .56 3.13 Total Estimated Income from Family Planning (FP) and Delivery Services by Province, 1993 .................................... 61 3.14 Estimated Number of Births per Student at Hospital Level .66 4.1 Distribution of Paramedical Staff in Indonesia, 1992 .75 4.2 Regional Distribution of Paramedical Students and Teachers in Indonesia, 1993 .77 4.3 Percentage Change of Paramedical Students in Public and Private Schools by Category, 1989-93 .78 4.4 Percentage of Paramedical Graduates from Public and Private Schools by Category, 1989-93 .................................. 79 4.5 Average Number of Paramedics by Type of Facility ............... 82 4.6 Use of Paramedics: Characteristics of "Efficient" and "Inefficient" Hospitals ................ ... ... ... ... .. ... ... ... .. . 85 FIGURES IN TEXT 1.1 Indonesia: Trends in Real Per Capita Health Spending .... ......... 4 4.1 Total Paramedical Staff per Health Center, West Java, Indonesia, 1992 . . . 83 BOXES IN TEXT 1.1 Two Views of How Health Centers Operate ...... . . . . . . . . . . . . . . 10 5.1 Impact of Government Health Spending ....... . . . . . . . . . . . . . . . . 91 5.2 PROSALUD and the Community Doctors Scheme ................ 98 ANNEX TABLES 2A Distribution of Doctors by Kapubaten in Selected Provinces, 1992 ... ... 108 2B Doctors Employed in Hospitals, by Province, December 1992 .... ..... 110 2C Fifth Round of Contract Placements, October 1993 ................ 111 3A Trends in Nursing Work Force in Class, A, B, C and D Hospitals, 1986-1992 .112 3B Comparison of Average Number of Nurses per Facility to Government Staff Norms, by Hospital Class and Province, 1992/93 .113 3C Difference in Average Number of Nurses per Health Center (HC) and Current Staffing Norms, by Type and Province, 1992 .114 3D Distribution of Health Center (HC) Nurses and Midwives (NMWs) by Area, Population and Number of Villages, by Province .115 3E Number of Bidan di Desa (BdD) Graduates, (1989/90-1992/93) and Remaining Target for Training, 1993/94-1996/97 .116 3F Distribution of Health Center (HCs) Nurses and Midwives (MWs) in Central Java ....................................... 117 3G Distribution of Health Center (HCs) Nurses and Midwives (MWs) in Kalimantan Barat Province .118 3H Distribution of Health Center (HCs) Nurses and Midwives (MWs) in Sumatra Barat .119 31 Potential Monthly Bidan Income from Attending Births, as Percent of Expected Public Salary, 1993 ............................. 120 3J Accounting for Nurse Staffing in C, D and Private Hospitals .... ...... 121 3K Use of Nurses: Characteristics of "Efficient" and "Inefficient" Hospitals .................................. 123 4A Accounting for Paramedical Staffing in C, D and Private Hospitals ... ... 124 4B Accounting for Province-wise Variation in Paramedic Staffing .... ..... 125 ANNEX FIGURES 4A Sanitarians (D3) per Health Center, West Java, Indonesia, 1992 ... ..... 126 4B Sanitarians (DI) per Health Center, West Java, Indonesia, 1992 ... ..... 127 4C Assistant Pharmacists per Health Center, West Java, Indonesia, 1992 ... . 128 4D Administrators per Health Center, West Java, Indonesia, 1992 ... ...... 129 4E Laboratory Technicians per Health Center, West Java, Indonesia, 1992 . .. 130 4F Dental Workers per Health Center, West Java, Indonesia, 1992 ... ..... 131 4G Auxiliary Health Workers per Health Center, Indonesia, 1992 ... ...... 132 4H Distribution of Students and Paramedical Staff, Indonesia, 1993 ... ..... 133 41 Distribution of Staff and Graduates - D3 Nutritionists .............. 134 ExECuTIVE SUMMARY i. This report assesses Indonesia's referral hospitals and 16 teaching hospitals, approach to health work force development and roughly 10,000 workers in mental and and deployment, and identifies manpower- other specialized government hospitals. related challenges likely to stand out in the mid to late 1990s and beyond. The iii. The Government's program of Government of Indonesia's wide-ranging work force development and deployment health work force initiatives have been a has combined "supply" side investments in policy response to the large derived training with "demand" side interventions, demand for staff projected as a notably the hiring of health facility staff as consequence of an expanding network of regular civil servants. Supply side publicly run service points. At the heart of measures have involved spending on public this delivery system is the health center education and training facilities, support (puskesmas) which provides curative, and guidelines for private training outpatient care and undertakes preventive initiatives, tuition subsidies, curriculum health activities in service areas of roughly reform, controlled entry to some fields and 30,000 people. The puskesmas, which various licensing requirements for new directs those requiring inpatient treatment graduates. Demand side interventions have to district-level referral hospitals, is included the staffing norms for different responsible in turn for operating village- worker categories in health centers and level sub-centers, which offer limited hospitals, and various regulations and curative and preventive services. procedures governing staff hiring Puskesmas staff are also responsible for (including conscription of doctors), carrying out various community out-reach placement in facilities, supervision and activities. evaluation, compensation and benefits, promotion, reassignment and working on ii. The workforce associated with this private account. multi-tiered structure has grown rapidly. In 1974, there were fewer than 50,000 Why Evaluate Health Work Force health workers with some, often minimal Policies? training employed in government health centers and hospitals. By 1983, this iv. There are several reasons why a number had grown to roughly 84,000, review of Indonesian health work force including 44,000 health center staff and policy makes sense at the present time: 40,000 hospital workers. And by 1992, each of the more than 6200 government * Large costs. The more than health centers (puskesmas) had, on 150,000 government doctors, average, over 15 trained health workers. nurses, midwives and paramedics In all there were more than 90,000 account for a sizable proportion, puskesmas level workers. There were also nearly 60 percent in the early over 78,000 workers (including 25,000 1990s as compared to less than 40 administrative staff) employed in the 337 percent in 1986/7 and 1987/8, of the total health budget. Is this visible. This report reviews money well spent? The likelihood whether the traditional public sector of continuing constraints on focus of work force policy should goverrunent health spending makes continue, and in light of these this a pertinent question. developments, what might be useful benchmarks and instruments in a * Links to health system broader approach. perfonnance. As in other countries, health care in Indonesia Assessing Health Work Force Policy is staff-dependent and staff- intensive-the overall performance v. There is no well-established of government-run services is methodology to evaluate a program of linked inextricably to the health work force initiatives. Nevertheless, availability and effectiveness of a balanced assessment would inevitably workers operating individually and draw on two perspectives. The first, a in teams. Accordingly, concern supply-side "audit", should cover the about health system outcomes, e.g., functioning of various work force the uneven and perhaps slower pace production (education and training) and of recent infant mortality rate deployment mechanisms. Questions (IMR) decline in the country, involve whether workers of requisite skill inevitably involves questions about have been trained, placed and remain in the allocation, training and designated positions and whether various productivity of health staff. work force interventions are sustainable financially. * Policy innovations. During Repelita V, the Government of vi. The paper's findings, in respect of Indonesia (GOI) made a number of such a review of "internal" efficiency, important adjustments in health highlight some major achievements: work force policy. The most notable of these innovations were 0 National capacity to train doctors, the contract doctor scheme and nurses, midwives and paramedics creation of a new worker category, has been greatly enhanced. As a the village midwife (bidan di desa). consequence, Indonesia has Do these measures have an succeeded in expanding access to appropriate focus and design? The skilled health personnel. For potential impacts and the possible instance in 1992, there were 1.1 drawbacks of these new policies are doctors on average per facility in assessed in this report. the country's more than 6200 health centers. In 1980, half of the * A growing private sector and 4800 facilities then in operation other developments. A large and were without physicians. diverse private health sector has emerged in Indonesia during the 0 The quality of the stock of health last twenty years. In addition, workers continues to improve. For considerable institutional example, the ratio of newly development has occurred, with certified specialists to graduating health insurance firms and general practitioners (GPs) has professional and specialists' grown, while the proportion of associations becoming far more diploma holders in each emerging ii nurse cohort has risen sharply. However, this pledge may have Similar changes have occurred in made it more difficult to use this the credentials of the paramedical mechanism to guarantee a supply of workers employed in government doctors in the Eastern Islands. hospitals and health center. * The training of doctors, especially vii. Improvements in the quantity and in public universities, remains quality of health personnel have contributed highly subsidized, even though to the achievements of the Indonesian most graduates are now likely to health system, such as high levels of spend their careers in some form of contraceptive use and immunization against private practice. childhood diseases, and the decreased incidence of problems linked to Vitamin A * The quality of medical education deficiency. Nevertheless, the paper also remains uneven. Three identifies certain work force-related mechanisms-assessments of concerns: medical schools, curriculum reviews and student * Government health workers still examinations-that can be used to tend to be concentrated in urban achieve higher quality need to be and some rural areas. This has strengthened. resulted in continuing shortages of key staff in some districts, while * Entry requirements, selection other localities remain procedures and training activities overendowed with workers. for the new bidan di desa (village midwife) worker category may * The annual output of doctors, result in graduates who are not nurses and paramedics continues to sufficiently prepared sociologically, rise. However, the absorption of culturally and academically for new graduates in public sector jobs work at the village level. decreased during Repelita V and is likely to fall further during Repelita 0 As regards nurses, midwives and VI. An ingenious--but possibly paramedic workers, current temporary response--to this arrangements for school situation involves the 1992 decision accreditation, curriculum reviews, to hire newly drafted medical examinations, licensing and graduates on three year, non- scholarships remain underutilized. renewable contracts rather than as These regulatory mechanisms can civil servants. The growing size of be used to achieve qualitative graduating medical classes will improvements in worker likely make it difficult to absorb all performance. new doctors into the contract scheme. viii. Incorporating a demand-side perspective. It is not sufficient to assess * The contract scheme may prove to work force initiatives only from a supply- be problematic because of other side, internal efficiency vantage point. factors. For instance, to make This is because in principle it is possible rural assignments more attractive, for health worker training and deployment GOI has promised not to compel measures to be working well, but without recruits to serve in remote areas. achieving significant health effects or iii eliciting much client interest. Accordingly, or below. Elsewhere, especially in the health work force policies also need to be Outer Islands and among the poor and in evaluated in terms of impact criteria. An economically lagging areas of Java and examination of health system performance Sumatra, health advances have been more and impact can help determine whether the circumscribed. demand (need) for different workers has been assessed accurately and whether the xi. It is paradoxical that utilization of mix of worker qualifications and skills is public health services and staff has appropriate. remained low, well below average, in the areas of greatest need, including most ix. Of course, the performance of a Outer Island settings and the more isolated health system is influenced by many pockets of Java. The limited use of factors, e.g., local cultural practices, government hospitals is well documented. environmental variations, income levels, To illustrate, the bed occupancy ratio was and formal fees for services, which may less than 40 percent in roughly a third of overshadow the contribution of work force- the publicly run class C and D hospitals in related variables. But even if the latter January 1993. What is less well known is were not central to the performance of the how limited has been the recourse to system, it is still crucial for work force government health centers for different policy to be cognizant of health service services. For example, a 1991 survey outcomes and problems. It could be very conducted in East Kalimantan and Nusa costly to train and deploy a work force Tenggara Barat found that rural and urban which, due to culture and poverty, could residents made between one and two not be effectively utilized in many areas. outpatient visits per year to health centers On the other hand, health labor force and subcenters. By contrast, health centers policies, including the way tasks are in Malaysia are contacted, on average per defined, supervised and assessed within person, more than five times a year. Such facilities, may well have a strong influence low visit rates are indicative of low service on the system's performance-policy demand at the facility level. In 1991, the design also needs to take account of this highest health center contact rates were possibility. recorded in Java-visits were as high as 76 per day in West Java. Utilization of health x. Determining the impact of health centers was much lower outside Java, with services in Indonesia is not a average daily visits of 30 or less recorded straightforward exercise. This report uses in many provinces. several crude indicators, most notably the IMR and facility visit rate, for which large xii. To be sure, more refined indicators cross-sectional variations have been of health service impact would be helpful. observed. Considering the evidence Still, these findings raise several concerns provided in IMR trends, the pace of about the effectiveness of work force improvement appeared to have been slower policies. One issue relates to the practice after the mid 1980s than in the previous of maintaining large numbers of staff in two decades. The trend in infant mortality underutilized public hospitals and health decline is, itself, a reflection of divergent centers. This points to a flaw in the way health experiences by region and by Indonesian planners have traditionally income class. For instance, there are estimated future demand for different areas, e.g., Jakarta, Yogyakarta, Bali and worker categories. Specifically, the large parts of Central and East Java and "production" of health has been visualized Sumatra, which have achieved IMRs of 50 in almost mechanical, fixed input-output iv terms, with specified quantities (presented decisions. For instance, doctors as facility-specific staffing norms) of would need to work out whom to distinct, non-substitutable skills, e.g., see on a private basis, what fees to doctors, basic nurses, nutrition assistants charge, and how much attention to and so forth seen as determining factors in devote to public clients and to other the improvement of health status. responsibilities. Accordingly, health work force development has been conceptualized in * Workers in different staff technical rather than economic terms, with categories are all assigned large the central concern that of eliminating numbers of overlapping tasks, but anticipated staffing gaps. The agenda has are given little guidance on how to been to put into place training programs prioritize activities and use and other mechanisms that would alleviate available time efficiently. For projected personnel shortages. example, the separate curative, preventive and health education xiii. Causes of low utilization. This tasks of nurses are carefully spelled report also explores the extent to which out. Yet the amount of time to be work-force-related policies may have devoted to each task is not contributed, directly or indirectly, to low addressed in operational guidelines. levels of utilization and uneven health In practice, this has given system performance. The paper does not individual workers the opportunity offer a comprehensive analysis of the to define de facto performance factors that may influence demand for norms and to establish considerable hospital and health center services. discretionary control over their Instead, some partial hypotheses are public work scope and flow. For developed, based on micro studies of instance, nurses seem to have some facility operations. These hypotheses tie autonomy in deciding whether to low utilization to the quality of treatment emphasize curative care at the offered to clients and patients, and in turn expense of other, possibly more to the skills and motivation of health important tasks and obligations. workers, the tasks they are assigned, and the way performance is supervised and * Assessment mechanisms have been assessed. For instance, the following unable to provide the feedback policy-influenced features of puskesmas needed to improve puskesmas operations are seen as problematic: performance. Overspecification of tasks is partly at fault. This can * Short assignment periods and lead to a superficial approach to sometimes lengthy absences have supervision, one which is based on meant that doctors, in many areas, checklists rather than key have had little time to become outcomes. This is compounded by familiar with local health needs, the overuse of evaluation and to respond effectively to techniques-previously announced problems in service delivery. visits by large teams, achievement of numerous quantitative targets, * Opportunities to conduct private and self-assessment-which can be practice have meant that doctors, manipulated. Supervision basic nurses and other health mechanisms have not attempted to workers have had to contend with a harness or redirect widespread job potentially complex interplay of v interpretation and discretionary estimated 750-1500 remote and typically control over work activities. high IMR localities which need close attention. Services in these areas need to * The phenomenon of job be protected from possible budget cuts, interpretation is of special interest while innovative work force initiatives may because personal interests and be required to assure that appropriately motives, including possibilities of trained health workers are available and earning income through private effective. practice, seem especially likely to come into play when opportunities xvi. But enhanced utilization and for discretionary use of time improved puskesmas performance in high present themselves. The concern is IMR areas may be unattainable without not just that health center personnel some rethinking and testing of service reserve a large part of their day for strategies and management within health private work, but that decisions and centers. Possible directions and options behavior in their public duties may can be distilled from micro-level analyses be influenced by the aim of of the determinants of low facility enhancing their unofficial income. utilization (see Box 1.1). For example, the In the process, the performance of partial hypotheses presented in this report tasks and activities on public point to a need to institute new incentives, account may be neglected. goal-setting mechanisms and assessment procedures for health center staff, including Emerging ChaUlenges facility managers. These revised "rules of the game" would give administrators and xiv. Taken together, the supply and workers more formal authority and demand side perspectives point to some autonomy (acknowledging that they have new features of the health scene. These already claimed informally a considerable new elements will need to be reflected in zone of discretionary authority), while also health work force thinking. First, due to making use of conventional motivational budgetary constraints and rising factors, e.g., income adequacy and enrollments, the public sector will no security, peer pressure and professional longer be able to absorb all or most standards, career advancement, and holding graduating health workers. A large share workers accountable for realistic of each new cohort of trained workers is performance levels. In respect of these and now likely to work primarily in private other hypotheses, it would be desirable to practice and/or employment. This points look in depth at health centers and hospitals to a need to give greater attention to the in Indonesia which appear to have been regulatory and oversight functions and successful, judged by utilization levels, instruments of health work force policy. user surveys and other criteria. The experience of various Indonesian NGOs, xv. Public sector employment may also e.g., Yayasan Kesejahteraan, and private need to grow less rapidly once demand for ventures may yield helpful insights as well. workers is based, not on staffing norms, It is also important that policy makers look but actual or prospective utilization levels abroad for appropriate lessons (see Box within governmental hospitals and health 5.2) particularly at attempts to link staff centers. At the same time, determined income or benefits to service provision in efforts need to be made to improve public facilities, and to give users a formal utilization and health outcomes in role in facility decision making. impoverished areas. There are an vi Recommendations and request technical support. Of course, the performance of these managers needs to xvii. This paper arrives at some generic be monitored and evaluated, using district findings and recommendations which are and facility level goals and indicators as cited here. More detailed suggestions for well as measures of community support. policy adjustments are presented in the And different ways need to be explored of individual chapters on doctors, nurses and sustaining competent and credible health midwives and paramedics. center and hospital teams, headed by committed and experienced managers. xviii. Make work force policy more Alternatives in this regard would include responsive to local public health needs, creation of an autonomous national health and more selective and cognizant of corps, reliance on expanded incentive market forces and opportunties. packages for key staff, and much greater Indonesian health work force policy is use of contracting or franchising of service outgrowing an approach that was shaped by delivery to NGOs and private providers. a sense of impending personnel shortages in government facilities. Policy no longer xx. Moving to a work force approach needs to be driven by a gap-filling that is geared to local needs and demands imperative, characterized by detailed and will entail further devolution of the tasks rigid staffing norms and ambitious training assigned to provincial and district and hiring targets. Nor is it necessary to governments. For instance, local rely on extensive subsidization of medical authorities will need to assume education and staff training, graduate responsibility for the location and coverage conscription and compulsory job of facilities; allocation and deployment of assignment, and complicated and different staff categories; the availability of ineffectual service assessment practices. worker housing, means of transport and drugs; and supervision and assessment of xix. Instead, policy needs to be demand- worker performance. This reassignment of sensitive.' One way of proceeding is to responsibilities should proceed in a phased require public facility managers including manner as local governments gain new the dokabu, who is responsible for district capacities and authority. health matters, to develop realistic service delivery priorities, including addressing xxi. Finally, with work force needs in crucial public health concerns, as well as public facilities determined in this bottom- improving utilization levels and/or the up fashion, GOI can be less concerned to quality of treatment as appropriate. These sustain high levels of production of new administrators could be made responsible doctors, nurses and so forth. The large and accountable, initially in pilot districts existing pool of health workers provides a (and/or facilities), for deciding on the further reason to downplay target levels of number of staff, skill mix and task health graduates. Accordingly, subsidies to assignments in their hospitals and health public and private training institutions can centers. These decisions should be taken be cut, perhaps sharply. This would leave in light of district facility and budget the numbers of graduating doctors, nurses, constraints, as well as expected utilization midwives and paramedics to be determined levels and outcomes, and local service through individual or family calculations, priorities. Moreover, district and facility taking account of realistic education and administrators should be able to remove training costs and likely earnings in public ineffective workers and hire or obtain and private employment. Reduced additional personnel, reallocate resources subsidies will also likely induce training vii institutions, public and private, to make public of opportunities and policy options adjustments in the scope and quality of in the health field. This entity should their programs. include representatives from government agencies as well as the professional boards xxii. Include the private sector in the and health worker associations, private coverage of work force policy. Changes universities and hospitals, insurance in the health scene, including growing companies, NGOs and distinguished private private sector employment and training citizens. The mandate of this oversight activities, suggest that health labor force body should include physicians, but also all initiatives should no longer focus other health workers. exclusively on assuring service provision in public facilities. A stronger oversight and xxiv. The NHC or equivalent entity could regulatory role will be required in a sector help bring about the further maturation and that will have a growing private presence. involvement of professional health worker In this respect, the report finds that GOI's associations, guiding and inducing these regulatory machinery and capacities need to entities to play a larger role in setting be strengthened and expanded. For standards of work and care, exercising peer instance, more systematic and incisive control and providing oversight and school accreditation and evaluation performance review functions, designing procedures need to be established, and and conducting in-service and preservice comparable approaches need to be applied training schemes and handling malpractice to private hospital clinics. In addition, cases in a credible fashion. The oversight publicly owned hospitals and schools body would also need to help draw should be given greater effective autonomy boundaries between associations, define the but subjected to regulatory oversight. domain of self-regulation and provide a Finally, the use of licensing and forum for the resolution of inter-profession certification of individual staff, based on disputes. examinations and competency tests, will need to be improved and expanded. xxiii. Enhance institutional capacity for health work force policy making. Policy formulation and execution are currently diffused across and within several government entities. Some functions, especially those of developing and maintaining an overview of the manpower scene, are not getting sufficient attention. A National Health Council (NHC) or equivalent body, with a supporting secretariat, should be created to carry out strategic planning, to react to public and private sector employment conditions and consequences, to foster innovative approaches to health service delivery, to put together effective accreditation and licensing procedures and to work out other policies regarding health education opportunities and quality, and to inform the viii Endnotes As an example, the lack of close attention to actual or prospective service demand may interfere with ongoing efforts to train, deploy and establish large numbers of village midwives. The intention is to hire these field workers on three-year contracts, allowing for a single three-year extension, in the expectation that private practice will eventually provide their principal livelihood. However, based on reasonable assumptions about likely earnings from midwifery and family planning services, this report concludes that many bidan di desa will be unable to become economically self-supporting. ix CHAPTER ONE THE HEALTH SCENE IN INDONESIA 1.3 Analogous measures have been applied to Introduction nurses and various paramedical positions, e.g., laboratory technicians, nutritionists and so forth. 1.1 During the last twenty years, the For instance, the Ministry of Health (MOH) Government of Indonesia (GOI) substantially conducts pre-service and in-service training in its expanded access to publicly provided health own nurse and paramedic schools, provides services. This heightened service availability subsidies and sets standards for private training was achieved primarily by enlarging the network programs, specifies staffing norms for different of health centers (puskesmas) and subcenters facilities and defines procedures and regulations (puskesmas pembantu) tasked with delivering governing staff hiring and placement in curative (outpatient) and preventive care in rural facilities, supervision and evaluation, transfer and urban localities. Between 1978 and 1993, possibilities, compensation and benefits, and the number of health centers increased from promotion. 4353 to 6588, while the number of subcenters grew from 6636 to 18,816. Services made 1.4 This report assesses Indonesia's approach available in publicly run hospitals have been to health work force development and enhanced as well. For instance, the number of deployment, and identifies manpower-related first line referral (class C and D) hospitals challenges likely to stand out in the mid to late increased from 260 in 1978 to 300 in 1993, with 1990s and beyond. The current chapter total bed capacity rising from 23,962 to 29,449 highlights salient features of the health in this period. landscape. Attention is directed first at the governrment-run health system-the multi-tier 1.2 This expansion of services has been delivery structure is described and funding supported by policies and interventions which trends are reviewed. Discussion then turn to together comprise an ambitious program of health indicators, facility utilization patterns, and health work force development and deployment. characteristics of private health care providers. For doctors, this effort encompassed "supply" side measures such as investment in public 1.5 Subsequent chapters address policy medical schools, encouragement of private instruments and issues in respect of different medical education, subsidies in government staff categories, beginning with doctors schools, curriculum reform, controlled entry to (Chapter Two), and continuing on to nurses specialist training and various licensing and midwives (Chapter Three) and various requirements for new graduates. Related paramedical workers (Chapter Four). Chapter "demand" side interventions have included Five pulls together key findings and staffing norms established for government recommendations, and takes up generic topics hospitals and health centers, conscription and and concerns in regard to current health work lifetime employment of medical graduates and force policy. These crosscutting issues special benefits and allowances for those serving include low average facility utilization and in remote areas. worker productivity, ways of reconstituting and focusing delivery systems in high IMR percent) between 1978 and 1993. There are and low IMR localities, and the suggested more than 90,000 trained health workers ends and means of work force "planning" in a employed at the puskesmas level. As with the hybrid, demand-driven delivery system. different hospital classes, puskesmas staff allocations are based on specific norms for 'The Government System different worker categories. In Java, a health center team usually includes at least one 1.6 Service Delivery. The locus of doctor (and sometimes two or three) and in government-provided health services in some cases a dentist, six nurses and midwives, Indonesia has changed during the last twenty nine auxiliary health workers, and a driver, years. Through the early 1970s, the publicly- handyman and several administrative workers. run system was largely hospital-based, with Outside Java, the puskesmas contingent is non-hospital staff working in a relatively small typically somewhat smaller but still sizable. number of peripheral facilities, and on a Subcenters are typically staffed by a nurse- limited scale, in vertically managed public midwife and an auxiliary worker. health programs in rural areas. The hospital sector has grown, and remains an important 1.9 The (senior) doctor in the puskesmas component of the health system. The plays a key role. He is the primary medical government's 300 class C and D referral authority and guarantor of quality. In this hospitals employ almost 27,000 health workers capacity, the doctor typically sees at most 25 and nearly 11,000 administrative staff. percent of the clients who need curative care, Another 26,600 health and 14,000 on a referral basis. The doctor monitors the adminstrative workers are employed in the 37 clinical work of nurses and auxiliary class A and B referral and teaching hospitals, employees who see the bulk of the visitors to while an additional 5000 employees are on the the facility. The doctor also serves as the staff of various public mental and other facility manager and executive officer. In this specialized hospitals. capacity, he is the overall coordinator and supervisor of all staff activities, including 1.7 However, the centerpiece of the those of subcenter and outreach workers, and government system is now the puskesmas the principal planner, priority setter and which provides curative, outpatient care and spokesman vis a vis district officials and the undertakes preventive health activities in rural local community. and urban service areas of roughly 30,000 people. The puskesmas, which directs those 1.10 The work of the puskesmas physician- requiring inpatient treatment to district level in-charge is aided by the clear and detailed job (class C and D) referral hospitals, is descriptions and assignments for different responsible in turn for operating village-level workers, and by recommended work standards subcenters, each covering about 7000 people, and reporting relationships which are set out which offer limited curative and preventive in various manuals (Pedoman) and operational services. Puskesmas staff are also responsible guidelines (Petunjuk). Health center (and for carrying out various community outreach physician) performance are assessed, activities, including supporting posyandus, the meanwhile, through the quantitative targets -nonthly village sessions devoted to preventive specified for all major activities, and through care for children and mothers.' other evaluative instruments. 1.8 As mentioned, the number of health 1.11 One such mechanism is the annual centers grew by more than 2000 (over 50 classification, known as "stratification", of 2 health centers into groups with good, includes all of the district office's section satisfactory or less than satisfactory ("fair") heads. During the latter exercise, which in performance. The exercise, which is intended some districts occurs monthly, elsewhere to give feedback on the overall operations of quarterly or twice yearly, the entire spectrum the puskesmas, is based on achievements, of program activities and achievements is relative to targets, for the 12 patient services reviewed. offered by facility staff, and quantitative indicators for various supporting activities. 1.14 Funding Constraints. Expansion of Results are often publicized, and followed up the publicly-run health delivery system in with remedial steps, e.g., filling position Indonesia has proceeded despite the limited vacancies or sending staff for training. The availability of government funding. Chronic best health centers receive certificates, and the resource shortfalls were not anticipated when best puskesmas doctor in each province is the rapid build-up of health facilities was invited to meet the President. In one province conceived in the mid- 1970s. Indeed, central (East Java), the worst health center is given a government spending (excluding foreign loans black ribbon, and the doctor-in-charge is and grants for salaries) on health rose by 172 required to provide a public explanation at the percent in real terms during the first four district health office. years of Repelita III (1979/80-1983/84). However, this increase proved to be short- 1.12 A parallel evaluation mechanism lived, as public spending decreased sharply operates through the accrual, by individual during the final year of Repelita III, and doctors (and other staff), of credit points based continued to fall during much of Repelita IV on performance as well as professional (1984/85-1988/89). It was only during development, e.g., attending a conference, Repelita V (1989/90-1993/1994), that real per having a paper published and so forth. capita government health spending exceeded Promotion occurs once a required number of levels attained during Repelita III. In fact, points are earned. Some provinces have real per capita health spending was only six converted this assessment exercise into a way percent higher in 1991/92 than in 1982/83 of honoring outstanding performance. For (Figure 1.1). Thanks to chronically low instance, NTB awards a "best doctor" spending levels, Indonesia continues to lag certificate based on demonstrated professional behind countries of similar per capita income skills, job performance, leadership qualities as regards government health outlays. For and service in a rural area. example, per capita government health expenditure in China and India exceeded that 1.13 Finally, formal and informal feedback in Indonesia, while outlays in Thailand and is conveyed to health center doctors through Malaysia are large multiples of the Indonesian numerous technical support and supervisory figure. visits by district health staff (Dinkes). For instance, technical guidance (Bintek) from 1.15 In this context, the health work force section heads in the district office is provided has absorbed a rising share of the health through monthly coordination meetings and budget--salaries accounted for nearly 60 semiannual visits. Sometimes a Bintek visit percent of health spending in the early 1990s coincides with the periodic (thrice yearly) as compared to less than 40 percent in 1986/7 "team supervision" in which a district section and 1987/8. At the same time, resource head and staff provide overall review and constraints have affected health work force program guidance. Each health center is also development in several ways. One immediate visited by the "integrated team" which consequence has been to induce MOH to 3 Indonesia: Trends in Real Per Capita Health Spending 1' is. J8 - 4' 20 a- I3 I i IS 1 SI I I SI it 13 44 83 Is 87 Is as so 91 Year Source: Minisuy of Health budget documems moderate its plans to expand the health work scheme, incoming general practitioners are no force. For instance, the number of longer hired as civil servants using formasi government health sector workers, including allocated by BAKN, the country's public non-medical staff, was expected initially to service commission. Instead, doctors are grow from 169,000 in 1983/84 to 284,00 in offered fixed term (three-year) contracts, with 1988/89, nearly 500,000 in 1993/94, and over the possibility that some contract scheme 700,000 in 1998/99. This projected staff "graduates" could eventually be employed on build-up of 23,000 net new recruits per year a permanent basis. Proposals under Repelita during Repelita IV and even greater annual VI to avoid any net increase in the overall size increments in subsequent years, was derived of the civil service are likely to spur even from a forecasting exercise which predicted greater recourse to contractual arrangements. extreme manpower shortages throughout the MOH is currently considering extending 1990s (Gani, et.al., 1993). However, because contractual hiring to dentists, midwives and of funding limitations, new hiring has several other staff categories. proceeded at a pace which, while still very rapid, is below that envisaged. Thus, the Changing Health Challenges number of new "formasi," i.e., fully funded new government health posts, grew by roughly 1.17 Despite relatively low levels of 11,000 annually during Repelita IV and by government spending, the health status of 14,000 a year during Repelita V. Indonesians has continued to improve-though progress has been uneven. This is seen in the 1.16 A n e v e n more s t r i k i n g erratic course of the infant mortality rate acknowledgement of resource constraints is (IMR), which is a useful, broad-gauge subsumed in the contract doctor scheme that indicator of health performance. IMR fell was initiated in February 1992. Under this from an estimated 112 deaths per thousand 4 live births in the early 1970s (reference period vehicle and employment-related accidents, and 1970-74) to roughly 78 in the early 1980s chronic illnesses such as cancer and (1981-86). But, during the last decade, a cardiovascular ailments which are attributable strong downward trend has not been to lifestyle adjustments linked to rising discernible, and the risk of infant death may incomes. At the same time, communicable have stabilized at around 70 deaths per diseases have not disappeared and some, like thousand live births. tuberculosis, various sexually transmitted diseases and now AIDS, pose urgent 1.18 The uncertain pace of mortality decline challenges to the health system. is a reflection of divergent health experiences and accomplishments by region and by income 1.20 Finally, an unappreciated dimension of class. For instance, there are a number of the health scene emerging in Indonesia's areas, e.g., Jakarta, Yogyakarta, Bali and dynamic "center" is the likely increasing trend large parts of Central and East Java and in morbidity rates, defined as the population Sumatra, which have achieved IMRs of 50 or share who report that they are sick at any below. In these rapidly urbanizing and given time. Rising sickness rates are increasingly prosperous settings the risks of important because they may foreshadow, dying, especially at younger ages, from better than mortality indicators, the trend in infectious diseases and birth-related causes demand for health services. The inverse have fallen due to the impressive coverage of association that is implied between mortality immunization, family planning, and other rates and prevalence of illness has been services and thanks to the large share (e.g., noticed in several countries, e.g. Japan, over fifty percent in Bali and North and West Hungary, Britain, and India (Riley, 1990; Sumatra) of deliveries attended by trained Gopalikrishna Kumar, 1993). The personnel. IMR also varies across income explanations typically offered are consistent groups. As an example, in 1991 infant with Indonesian conditions, especially in areas mortality for mothers who had completed where IMR is 50 or below. The incidence of some secondary school, a reasonable proxy for extended duration, degenerative or long- household income, was one half that for those convalescence illnesses has risen. There is who had not completed primary school (GOI, earlier detection of health problems. Rising 1992). incomes have brought greater sensitivity to and expectations about health status, and an 1. 19 In Indonesia's rapidly growing "inner" expansion of the number of conditions defined islands, death is now more likely to take place as sickness. Lastly, there are various at older ages. For example, the 1992 national economic incentives which persuade people to household expenditure survey (SUSENAS) present themselves as unwell. found that adults over age 60 accounted for 46 percent of all deaths in 1991 in Java and Bali, 1.21 Elsewhere, especially in the Outer as compared to 11 percent in Maluku, Irian Islands and among the poor and in Jaya and East Timor. In areas of IMR of 50 economically lagging areas of Java and or less, deaths in the 60 and older age group Sumatra, health advances have been more may already account for over 75 percent of all circumscribed. One telling indicator is that mortality. This altered age pattern of death still occurs predominantly in the younger mortality is tied to changes in the age cohorts. For instance, children age four epidemiological profile. Death in and under accounted for 55 percent of all economically favored areas of Java, Bali and deaths in 1991 in Maluku, Irian Jaya and East Sumatra is increasingly linked to motor Timor, 44 percent in East and West Nusa 5 Tenggara, and 38 percent in Kalimantan, but of public hospitals fell between 1984 and only 22 percent in Java and Bali (GOI, 1992). 1993, and are generally quite low (Table 1.1). In these settings, the burden of traditional Class C and D hospitals reported the lowest diseases, e.g., malaria, diarrhea, respiratory rates. Occupancy rates fell after 1989 in infection and nutrition-linked disorders nearly 40 percent of the class D facilities. remains substantial, while birth-related They were particularly low in Sumatra, problems continue to cause significant loss of Kalimantan and Sulawesi (Table 1.2). There life. were 47 (27 percent) such facilities with utilization rates of 30 percent or less. Underutilization of Public Services 1.23 Recourse to government health centers 1.22 The finding that health gains have been for curative and preventive care also occurs on limited in some settings is paradoxical in that a limited basis, not only in the Outer Islands many services and facilities in the affected but in the country overall. For example, only areas and elsewhere remain underutilized. For a quarter of all rural Indonesians, 24 percent example, bed occupancy rates for all classes in Java and Bali and 27 percent elsewhere, Table 1.1: PUBLIC HOSPITALS: BED OCCUPANCY RATES BY HOSPITAL CLASS AND YEAR 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 - 1- --- - - - Class A 75.2 75.5 77.0 71.4 75.0 72.5 70.1 64.6 60.1 60.4 Class B 78.5 68.7 70.2 66.4 61.4 60.6 64.2 62.1 60.4 60.1 Class C 59.4 59.6 59.4 58.4 58.0 57.3 58.4 58.4 58.1 55.4 Class D 48.0 49.5 48.7 48.7 45.3 44.8 45.9 47.1 45.1 42.9 Source: Ministry of Health. Table 1.2: BED OCCUPANCY RATES, C AND D CLASS PUBLIC HOSPITALS, BY REGION: DECEMBER 1993 Hospital Class Region C D Java (including Jakarta) and Bali 59.6 51.3 Sumatra 44.5 36.8 Kalimantan 50.9 36.0 Sulawesi 57.2 37.6 Eastern Islands 51.1 46.0 Total 55.0 40.7 Source: Ministry of Health. 6 visited the puskesmas when ill in 1992. These Table 1.3: AVERAGE NUMBER OF DAILY figures were somewhat lower than those VISITS TO HEALTH CENTERS BY PROVINCE, reported for 1987, e.g., 31 percent in Java 1990 AND 1991 and Bali and 26 percent for the Outer Islands 1990 9 (van de Walle, 1992). A 1991 survey conducted in East Kalimantan (Kaltim) and Jaat 37d 44.6 Nusa Tenggara Barat (NTB) found that rural and urban residents made between one and WestJava A 76.3 two outpatient visits per year to health centers Yozvakarta 65.8 43 2 and subcenters. By contrast, health centers in East Java 78.4 75.3 Malaysia are contacted, on average per Bali NA 33A person, more than five times a year. W-iatd 6, 2 68 7 1.24 Information on how frequently health centers and subcenters were visited is available in the 1990 national household expenditure WestSumatra 54.3 53.0 survey (SUSENAS), in MOH service Riau 43.2 42.9 statistics, and in the special 1991 survey in Jarn2i 19.9 23.4 Kaltim and NTB. The 1990 SUSENAS round South Sumatra NA 22.2 indicated that on average each rural health Bengk.lu 17.6 18.7 center received only 23 patient/client visits a Laungu 51.8 31.5 day and each subcenter only six contacts a WPiuhfod 4n 0 day. Contacts included patients seeking ____hm_nt__ _ _ _ __________ curative care, as well as visits for immunization, family planning and other West A23. 23.9 mother and child health services.2 In 1990 South and 1991, health centers themselves recorded East 19.9 31.3 44 and 50 visits per working day on average Central 22.4 NA respectively, with a reported range of four to i .722 6 . 257 78 patient contacts a day (Table 1.3). In _ ___ ___ 1991, the highest health center contact rates NorthSulawesi NA 54.1 were recorded in Java-visits varied from 45 Central 21.6 174 per day in Jakarta to 76 per day in West Java, where some districts reported over 100 South Sulawesi NA 5.2 contacts a day. By way of comparison, South East 17.9 16.4 Thailand's 536 community hospitals, which Weiahtp. . n207 17. 4 provide a similar mix of services, reported 72 visits per day in 1987. Visit rates were also EastNusa 3.6 39.0 high on average in East Java, though several Maluku NA 15.5 districts reported daily contact rates of less [in -fnva NA 4 6A than 50. On the other hand, utilization of Weighted 3.6 22.4 health centers is much lower outside Java withl 43 i -- average daily visits of 30 or less recorded in a number of provinces. For instance, there Source. Health Profile 1990 and 1991. were only 16 visits a day on average in 1991 to the 79 health centers in Southeast Sulawesi. Utilization rates also varied widely within 7 province. For instance, the 26 health centers 1.27 Another potentially important factor is in Kotabaru Kabupaten in South Kalimantan the affordability of health services. Here, the recorded only 11 visits a day in 1991, while fees collected for health services may be an the 23 centers in Banjar Kabupaten reported important deterrent to facility use. The poor, 45 contacts a day. Utilization of subcenters is especially, might have trouble finding the cash also typically quite low. A 1993 study in to cover the standard visit charge (Rp. 300 in NTB found that nearly half of all subcenters the early 1990s) as well as the additional fees recorded fewer than eight patient visits per that are sometimes collected for specific day. services. Clearly, detailed work is needed on the extent to which payments for services have 1.25 The 1991 Kaltim-NTB survey found been a discentive to facility utilization.4 It is higher, but still not impressive levels of daily worth noting, though, that fees, often set at use. Median visits per health center and modest levels, are collected routinely by subcenter in Kaltim were 56 and 9 private health providers, many of whom are respectively, while median daily figures for visited by low income patients. centers and subcenters in NTB were 61 and 12 respectively. Even more revealing though 1.28 In fact, fees may matter less than local was the variability in average visit levels. A concerns about the availability of needed handful of health centers had 150 more patient equipment, medicines and health workers. contacts a day, but a significant share, roughly These physical inputs contribute to the quality a third in each province, had fewer than 32 of health services provided to patients. No daily visits. Similarly, although a few doubt, many facilities need improving in these subcenters received 40 or more contacts a day, respects. For instance, the Kaltim-NTB a large proportion, 44 percent in NTB and 28 survey found many health centers to be percent in Kaltim, handled seven or fewer deficient in cleanliness, ventilation and visits a day on average. sanitation, while some lacked electricity as well as certain types of equipment, 1.26 Accounting for Health Center Use. instruments, supplies and drugs. Still, the It is not a straightforward task to determine same survey found that despite occasional which factors and mechanisms are responsible lapses, most facilities had sufficient drugs, for puskesmas utilization patterns in Indonesia. instruments and equipment to provide adequate Take access, for example. Proximity has an levels of service, while most health centers obvious bearing on facility contact rates, had at least one doctor, ten nurses and especially for poor clients and if the paramedical workers. Similarly, most (89 puskesmas was newly established. And percent) facilities in other provinces had at although many new health centers and least one doctor (Tables 2.6 and 2.7), and subcenters went into operation in the 1980s most health centers had twelve or more and early 1990s, wide variation remains nurses, midwives and other trained health staff between and within provinces in the average as well as four or more administrative/all- number of people nominally served by each purpose workers. facility. Nevertheless, visit rates are often low in places which have public facilities nearby. 1.29 A related set of factors, also subsumed The Kaltim-NTB survey found that 83-95 in the concept of health care quality, pertains percent of potential clients lived within "easy to the way available materials and staff inputs access" of a puskesmas.3 Yet, annual use are combined and used to produce specific rates and facility caseloads remain low in these services. It is, of course, difficult to obtain provinces. systematic, survey-based findings on such 8 facets of service delivery as client-provider of numerous quantitative targets, and self- interactions, the effectiveness of patient assessment-which can be manipulated and examinations, diagnoses and prescriptions, and abused. Moreover, established guidelines, job how efficiently staff use their time. However, descriptions and procedures, including in-depth studies of the functioning of assessment mechanisms, take little or no individual health centers offer an alternative account of the range of expectations, motives way of visualizing how ineffective or and incentives which affect the behavior of inappropriate procedures and management puskesmas employees and actual or potential techniques can result in low facility utilization. health center "customers". For instance, At least two such analyses are available for supervision mechanisms have not attempted to Indonesia (Box 1.1). These studies suggest harness or redirect job interpretation and that puskesmas activities have been disrupted, discretionary control over work activities by and local perceptions of service quality have health staff. And only limited attention is been weakened by some persistent operational given by supervision and evaluation phenomena. These unwanted and unintended instruments to the quality of service provision, outcomes seem to derive from continuing including the context in which care is challenges in respect of how available staff are dispensed-there are no regular mechanisms to deployed and managed. In particular, it follow up individual cases, to take stock of would appear that problems have been client's reactions and views and to consult encountered in arriving at an effective mix of with local community figures. Finally, staff roles and skills, motivation and measures of activity-specific and facility-level incentives. resource use and efficiency indices are generally not calculated. This interferes with 1.30 Imperfect Assessment Mechanisms. identification of activities and resource which The suggested linkage, derived from micro- are underutilized, and makes it difficult to studies, between low client utilization and work out which combinations of staff and factors internal to the health center is other inputs are cost-effective. somewhat surprising. This is because various feedback devices and supervision and Private Sector Health Services assessment procedures have been established (see paras. 1.10-1.13). Yet these instruments, 1.31 Another important feature of the health which are backed by detailed and specific scene in Indonesia is the increasingly pivotal operational guidelines and data on program role played by the private sector in delivering activities, have not brought about sharp health services. Included here are all services improvements in facility utilization. This may provided on private account, not only by be due to deficiencies in existing evaluation doctors but by the large number of nurses and arrangements. For instance, the detailed task midwives who see patients privately. Some of descriptions provided in the pedoman and the dimensions of the private health sector petunjuk may be partly at fault. This over- emerge in survey data. For instance, specificity can lead to a cursory and household expenditure surveys have shown superficial approach to supervision which, that the share of the sick treated by private guided by checklists, focuses on steps rather providers rose sharply between 1978 and 1987 than outcomes, and which discourages (van de Walle, 1992). This phenomenon was initiative taking and problem solving by staff. observed in most income groups and regions- This problem is compounded by the overuse in urban Java, the share of the poorest 40 of certain evaluation techniques-previously percent who consulted private health service announced visits by large teams, achievement suppliers when ill rose from 13 percent in 9 Box 1.1: Two VIEWS OF How HEALTH CENTERS OPERATE Two micro-level studies of puskesmas operations are worth noting. The first, by Katherine Neilan, was designed to complement the Kaltim - NTB survey, conducted by RAND/Lembaga Demografi, which measured quantitative inputs, e.g., equipment, drugs, numbers of staff, and so forth, available at each facility. Neilan looked at the process of service provision in ten health centers in Lombok, NTB, in August and September 1993, relying on interviews with staff and patients and direct observation. One finding was that doctors were not satisfied with puskesmas conditions, and complained about inadequate water supply, equipment and furniture. Yet what seemed most salient to Neilan were various deficiencies in the process of providing care to health center visitors. For instance, she observed that even when water was readily available, it was seldom used by staff, e.g., to wash hands before and after a delivery. Other process-related deficiencies cited in the study included frequent reuse of non-sterile needles; perfunctory physical examination and inconsistent diagnosis and prescription, especially by nurses and paramedics; overuse of drugs and injections; and limited attention by staff to instructing patients about appropriate behavior, possible side-effects and so forth. A second study, based on participant observation and interviews conducted by R. Sciortino in several facilities in Central Java in 1989-1990, reaches strong and very critical conclusions about the functioning of health centers. Sciortino observed that the effectiveness of puskesmas staff was disrupted and even undermined by some persistent operational phenomena: actual working hours were constrained by a "flexible" interpretation of official opening and closing times; personnel were not deployed according to competency or stated roles and assignments; activities were carried out haphazardly according to the availability and even the inclinations of staff; there were overt and implicit conflicts between workers; the biomedical knowledge of nurses was inadequate, while their treatment of patients over-emphasized injections and put little weight on examination and communications; puskesmas personnel favored medical-technical activities, in part because these lent themselves more readily to target fulfillment than participatory and educational tasks; staff were generally unwilling to become involved in various community-based health delivery activities and did not view patients and local villagers as partners in a health development process; fees were increased arbitrarily over official levels; there were huge reporting and recording requirements; the need to reach quantitative targets often led to rushed service provision, and at times to fictitious representation; and planned supervision visits by district officials were usually greeted with carefully orchestrated performances of supposedly routine activities, e.g., baby weighing at the posyandu. Sciortino found that local residents did not place a high value on services offered by the puskesmas. Wealthier households did not appear interested in community health initiatives, while poorer families lacked the time to come to posyandu sessions, felt they received nothing concrete from various outreach events and instead were "overdosed" with unchanging health education messages. Patients seemed inured to inferior, often rude treatment at the puskesmas -- "the popular assumption is that the standards in the health centers are lower; the instruments, contraceptive means and drugs are considered of lesser quality" (page 189). Respondents were confident, though, that they could get considerate and unrushed care in a pleasant and informal setting in the private practice of doctors, bidans and nurses. Sources: K.H. Neilan, "Case Study of Ten Puskesmas in Lombok, NTB, Indonesia," Project Overview and Preliminary Report, Stanford Medical School, September 11, 1993; and R. Sciortino, Caretakers of Cure, Amsterdam, Jolly Publishers, 1992. 1978 to 29 percent in 1987. Use of private important in rural areas, especially in the providers continues to be more common in Eastern Islands (Table 1.4). urban areas, especially in Java and Sumatra. On the other hand, government facilities, 1.32 Of course, private practitioners have primarily health centers, are relatively more always been present in Indonesia to some 10 Table 1.4: PROPORTION OF THOSE ILL VISITING DEFERENT FACILITIES AND PROVIDERS BY REGION, 1992 Sumatra Java and Bali Kalimantan Sulawesl Eastern Total Islands Urban Rural Urban Rural Urban Rural Urban Rural Urban Rural Government l Hospitals 6.7 2.9 5.3 1.4 6.7 1.7 8.4 3.2 12.8 2.5 4.0 Health Centers 19.3 23.0 18.7 24.3 21.8 22.0 16.9 24.0 25.3 35.1 22.0 Total _ _26.0 25.9 24.0 25.7 28.5 23.7 25.3 27.2 _ 38.1 37.6 26.0 Private Hospitals 3.7 0.8 1.8 0.5 2.9 0.3 2.9 1.3 2.2 1.6 1.4 Doctors 17.8 5.6 19.7 9.4 17.3 3.4 18.8 6.7 16.0 3.7 12.0 Clinics 3.4 3.6 2.5 2.5 1.3 2.1 0.8 3.7 2.2 6.2 3.0 Nurses/ Midwives 9.0 15.4 7.8 15.9 5.7 14.7 8.2 13.7 7.8 10.1 11.8 Traditional Birth 4.1 7.9 1.5 3.3 1.8 4.1 5.3 6.1 3.1 8.2 4.0 Attendant Other 1.2 2.9 1.1 1.3 1.8 3.9 4.5 5.2 2.2 3.6 2.1 Total 39.2 36.2 | 34.4 32.9 J 30.8 29.5 [40.5 36.7 | 33.5 33.4 34.3 Self Treatment - - _ l Self/no I l Treatment | 34.9 37.8 41.5 41.4 140.8 47.8 34.2 35.9 28.4 28.9 38.7 TOTAL 100.1 99.9 99.9 100.0 100.1 100.0 00.0 99.8 100.0 99.9 99.0 Source: 1992 Household Health Survey Module. extent. The scope and importance of private some important changes are taking place. The health supply activities may have been number of part and full time private work underestimated in the past, possibly because of opportunities for doctors and other workers the part-time and seemingly marginal nature of has increased, especially in urban centers but the commitment and the resultant income. also in more prosperous rural settings in Java, The part-time pattern is still the rule for the Bali and Sumatra. This is apparent in the large proportion of health staff still employed large number of private hospitals and clinics by the government, including the sizable that have opened in Jakarta and other contingent of contract doctors. However, metropolitan centers--there are over 22,000 11 full-time health workers employed in private epidemiological profiles within the population, hospitals. Extensive recourse to private and the mixed public-private delivery system practitioners is also documented in household in which many government facilities and surveys (Table 1.4). Secondly, private personnel are underutilized. The rest of this earnings have become the principal source of report assesses the government's health work income for many health workers. In fact, force initiatives in light of these "stylized private income earning opportunities very facts". Chapters Two, Three and Four focus likely play a large, even determining role in on the sustainability and effectiveness of work the labor supply and other decisions of doctors force training and deployment policies in and other health workers. respect of doctors, nurses, and midwives, and paramedical workers. In Chapter Five, Plan of the Study attention turns to possible ways of assuring that health work force policy is responsive to 1.33 This chapter has discussed some generic concerns, such as the uneven impact enduring as well as several altered dimensions of the public delivery system, the need for of the health scene in Indonesia. These demand-revealing mechanisms, including features include the tight constraints placed on much greater facility-level accountability, and government health expenditures, divergent the regulatory requirements of the expanding private health sector. Endnotes 1. The puskesmas is supposed to provide or carry out 18 services: mother and child health, family planning, community health nursing, occupational health, nutrition, environmental health treatment of infectious disease, health education, school health, sports medicine, immunization, dental services, ophthalmic services, basic laboratory work, care of the elderly, traditional medicine, emergency care, and recording and reporting. Not all of the 18 services are provided in each health center. 2. Puskesmas staff do more than provide curative and other services at health centers and subcenters. They also undertake posyandu, school and household visits and carry out various outreach activities. Accordingly, differences in health center visit rates may derive from differences in the performance of outreach tasks, i.e., staff in health centers with low visit rates may be busy with immunization and other posyandu-based activities. It seems more likely though that such intensive outreach work would result in more frequent referrals and other visits to health centers. Therefore, we would expect puskesmas visits to be a good indicator of how effectively staff were performing in their tasks. 3. A user exit survey found that only two percent of patients lived more than five kilometers from the nearest health center, and 80 percent lived within a kilometer of the facility, while 80 percent needed 20 minutes or less to reach the puskesmas. 4. The recently completed second round of the Kaltim-NTB survey should allow researchers to gauge the impact on service use of health center fee increases during 1992 and 1993. 12 CHAPTER TWO INDONESIA'S DOCTORS risen and now accounts for nearly 30 percent of 2.1 There were roughly 4000 doctors in all new doctors produced in Indonesia (Table Indonesia in 1970. This figure represented a 2.1). An increasing proportion of students, significant increase over the limited number- especially in private universities, are women. well under one thousand-of physicians in the The share of female graduates rose from 40 country when Independence was declared in percent in 1986 to 49 percent in 1992. Some 57 1945. Still the supply of doctors in the early percent of private medical school graduates in 1970s was much less than what was needed to 1992 were women, as compared to 44 percent staff a health system that was programmed for for public schools. rapid expansion. Accordingly, GOI turned to a mix of policies to increase the number of 2.3 Related policy measures include tuition qualified general and specialist physicians and to subsidies and other forms of support. Medical induce doctors to work in public hospitals and education for students in public universities is health centers. Various aspects of this policy heavily subsidized. Few students receive package have been refined and adjusted during scholarships, but public university fees are very the last two decades. Yet the broad outlines of low relative to the full costs, to the government, GOI's physician strategy remain unaltered. This of providing training. Public students pay strategy is assessed below. between US$ 1500 and US$ 2000 to attend seven years, on average, of medical school, The Policy Framework during Repelitas II-V excluding books and living expenses. Private students pay an estimated US$ 15,000 in fees to 2.2 Medical Education. One important set complete their training, which usually takes nine of interventions has aimed at reaching and or ten years. Public universities must have sustaining large annual flows of medical considerably higher training costs per pupil than graduates. To this end, nine new public medical those in private schools, because of their better schools were established in the 1950s and 1960s, facilities and higher teacher per student ratios. and three more were opened in the late 1970s Thus, the difference between these two cost and 1980s, bringing the total number to 15. figures, about US$ 13,000 per student, is a Students enter after high school and graduate as conservative estimate of the public subsidy to general practitioners (GPs). The annual output medical students. GOI has also provided some from public schools has continued to rise, subsidies to private universities, mainly through averaging 460 per year in the 1960s, 842 in the support for teachers' salaries. In 1992, 1970s, 1160 in the 1980s, and over 1350 during approximately a third of the full time instructors the early 1990s (Table 2.1). GOI also agreed to in private schools were financed by government the establishment of private medical schools. grants. Local governments occasionally fund The first such school produced its initial improvements in physical facilities for medical graduate in 1971. There are now 14 private schools. schools, located for the most part in Java. The combined annual output of these schools has 13 Table 2.1: PUBLIC AND PRIVATE MEDICAL GRADUATES, 1950-92. Public Private Total 1950-59 688 0 69 (annual average) 1960-69 4603 0 460 (annual average) 1970-79 8418 363 878 (annual average) 1980-83 3892 1039 1232 (annual average) 1984 1136 284 1420 1985 1066 194 1260 1986 1135 221 1356 1987 1538 256 1794 1988 1465 181 1646 1989 1412 286 1698 1990 1301 298 1600 1991 1389 335 1724 1992 1368 1 530 1898 Source: Consortium for Health Sciences. 2.4 Cumulative subsidies for the nearly 1900 examination for the six private universities in the doctors who graduated in 1992 were estimated at Jakarta area. over Rp. 38 billion (US$ 18 million).' This was 60 percent of the amount, Rp. 63 billion 2.6 Several other institutions play a role in (US$ 29.9 million), spent through the determining medical education policy. MOH is development and routine budgets on training for responsible for all teaching hospitals and remains other health staff in 1992/93. the major employer of physicians. The KOPERTIS are entities set up by the Directorate 2.5 The principal regulatory body for General of Higher Education to support private medical education is the Consortium of Health universities in each region. The Indonesian Services (CHS) which operates administratively Medical Association (IDI) and various specialist under the Ministry of Education, but includes associations have become influential voices in representatives from MOH. CHS evaluates the medical education arena. proposals for new medical schools, and undertakes periodic reviews of the medical 2.7 Training of Specialists. Steps have also curriculum and teaching practices. CHS does been taken to increase the outflow of medical not, however, play a role in testing students. specialists. Annual production of specialists has Each state university sets examinations for its nearly doubled since 1979 (Table 2.2) with the own students, while private medical students increase divided almost equally between the four must take a test developed at the nearest public general specialties-internal medicine, pediatrics, medical school. For example, the University of surgery and obstetrics and gynecology-and all Indonesia develops and organizes the other fields, often referred to as the consulting 14 Table 2.2: GENERAL AND CONSULTING SPECIALIST PRODUCTION, 1980-1992 1980 1984 1988 1992 General Specialists /a 119 152 135 215 Consulting Specialists /b 118 149 142 215 Total 237 301 277 430 General Specialists as 50% 50% 49% 50% share of total /a Internists, pediatricians, surgeons, and obstetricians/gynecologists. /b All other specialists. Source: Consortium for Health Sciences. specialties. The ratio of newly certified centers. Meanwhile, several associations- specialists to graduating GPs grew from 19 radiology, surgery and orthopedics-have percent in 1980 to 23 percent in 1992. Twenty- established national examination boards. The four specialties are now recognized by the Indonesian Surgeon's Association set up such an government, while each year since the late 1970s examination board in 1992, and has already has seen the graduation of a specialist in a new administered the written and oral test several field. The associations that have emerged in times. Some medical schools reported that the different fields recognize additional national board has helped them to improve their subspecialties. For instance, the Indonesian surgery programs. Other associations, e.g., Pediatrics Association has defined fourteen pediatrics, obstetrics/gynecology and pediatric subspecialties. anesthesiology are working to establish national examination boards. 2.8 Unlike undergraduate medical education, all specialist training occurs in public 2.10 Demand-side Interventions. GOI has universities. Applications are channeled through used various policy tools to make the medical the CHS to schools indicated by the prospective profession attractive to prospective graduates. student. Medical schools make selection Here, the ambitious coverage targets adopted for decisions based on the application and an health service delivery have played a role by interview. In addition, compulsory service in engendering a strong derived demand for general remote areas is supposed to improve admission and specialist physicians. The planned facility chances. In practice, it is unclear what weight per population coverage norms and the staffing universities give to remote area service relative standards for each type of facility are crucial to academic criteria. Fees, which are normally parameters in this regard. paid by the sponsoring agency, e.g., MOH or the military, are heavily subsidized. Most 2.11 The largest program-driven demand for students finance their personal expenses from doctors occurs at the puskesmas level. With their salaries which they continue to receive over 6600 health centers in operation (up from during the four to five year training program. 4353 in 1978), the targeted coverage ratio of one puskesmas per 30,000 people has already been 2.9 Specialist associations have begun to play reached, at least in aggregate terms. Each a role in assuring the quality of training health center is expected to employ 1-3 doctors programs. CHS routinely calls on the relevant depending on the range of services available and association to review proposals for new training the number of people served. Those units 15 covering more than 30,000 people may be training. In the past, all graduating GPs were allocated more than one doctor. All this required to enter government service, as civil suggests an institutionally mandated demand for servants, normally as puskesmas doctors. health center physicians of well above 6600. Fulfillment of this obligation was a condition for obtaining a license to practice; licenses were 2.12 Many doctors are needed as well in issued by provincial health authorities once the government-run hospitals. MOH has specified newly appointed GP had assumed his position in that each of the country's 132 class D public the puskesmas. GOI has also used other hospitals (averaging 52 beds) and 168 class C measures to make government service, public hospitals (averaging 134 beds), should specifically deployment to health centers, more have three and eleven GPs respectively. In attractive. For instance, the length of mandatory addition, each class C facility is supposed to service, normally five years, in the GP's first have seven specialists, one for each of a core posting was reduced to three years in most group of specialties. Minimal staffing standards areas outside of Java and Bali, and two years have also been defined for the class A and B in remote areas such as Irian Jaya. hospitals which have teaching, research and other specialized functions. The 33 class B 2.15 Adjustments were made as well in the public hospitals (averaging 457 beds) are each compensation received by health center expected to have eleven GPs and 37 specialists. physicians. In addition to the usual civil The four class A public hospitals (863 beds service benefits, i.e., salary and allowances, each) are to employ specialists only-each health coverage and pension rights, puskesmas facility is to have over 100 specialists on hand. GPs typically received free housing and a special incentive for rural service, with 2.13 Finally, a number of other government payments graded according to the remoteness jobs have been designated for medical graduates. oth henter For exmple,eM s For instance, doctors are expected to head the health office in each of the country's 350 revised its salary scale for doctors early in districts. Supporting administrative position, i.e. 1992. The pay of doctors working in cities as section chiefs with differing responsibilities, was set at Rp 296,000 (US$ 140) a month, within the district office, are also designated for while those working in rural health centers medical graduates. Similarly, the kanwil were to earn Rp 496,000 (US$ 235) and those (MOH) and dinas (local government) health serving in very remote areas got an additional offices at the provincial level include Rp 200,000 (US$ 95) a month. In addition, administrative and technical jobs which are doctors as well as other puskesmas staff in reserved for and typically filled by doctors. A some if not all provinces are allowed to retain significant number of physician jobs exist as a share of the health center's earnings from well within the public medical schools and in the fees. In Kaltim in 1992/93, such "incentive" headquarters offices of MOH itself. Most payments amounted to an estimated Rp. medical school positions are held by specialists, 75,000 (US$ 34.50) per doctor per month-- while the MOH slots are often filled by general doctors were allocated 28 percent of retained physicians with the M.P.H. degrees. income. Lastly, doctors must complete their initial health center assignment before they can 2.14 GOI has used its powers of conscription enter a specialist training program. As noted, to secure a steady flow of medical graduates an additional reward promised to puskesmas from public and private universities, and doctors, especially those in more remote specialists to jobs in public facilities. All areas, was more favorable treatment of their specialists must complete three years of applications for positions as specialist government service, normally in a class C residents hospital, once they have finished their advanced 16 2.16 Private Practice. As a further benefit, 2.18 Doctors differed, of course, in the MOH has allowed government doctors to amount of time they put into private work, conduct private practice after the close of the and in other aspects, including the fees they official work day, i.e. after two p.m. The charged and the income earned in private intention was to enable physicians to practice. Physicians tended to work somewhat supplement their civil service earnings, fewer days and shorter hours in urban areas, thereby making it easier to attract people to with specialists, especially those in Jakarta, rural areas and adding to the stability and spending less time in their private capacity sustainability of the health system. Different than GPs (Table 2.3). The time invested in SUSENAS rounds and other surveys have private practice was especially large for those documented how widespread is the use of working in health centers, with doctors in private medical practitioners by Indonesians some regions, e.g., Bali, Sumatra and from each income class and region. But little Kalimantan, spending over six days a week for is known about who is involved in private up to four hours a day on private work. One practice, average patient loads and fees and anthropological study suggests that doctors other relevant details. Accordingly, this study weigh various factors in deciding where to sponsored a survey to determine key features situate their private practice, often employ of private medical work. The survey was other health center staff in a supporting role, conducted by IDI using a questionnaire that and take great pains to create an informal and was mailed to a sample of 4000 members in pleasant context for patients (Sciortino, 1992). February 1993.2 2.19 City-based doctors including specialists 2.17 Survey results indicated that private seem to compensate for shorter time practice was pervasive and represented an commitments with higher average fees. Not important source of income and a significant surprisingly, it was Java (including Jakarta) influence on the expectations and behavior of and Bali-based specialists who charged the doctors. Most doctors, e.g., 80 percent of highest fees, while the lowest fees were GPs and 90 percent of specialists, 84 percent reported by puskesmas doctors working in of those with health center jobs, 80 percent of Sumatra, Java and Bali (Table 2.4). The latter those working in hospitals, 90 percent of group received a greater share of their fees in university-based physicians, and 93 percent of kind and treated a larger proportion of patients those with administrative assignments, for free. Most doctors indicated that fees conducted private practice. Most (53 percent) were set with a view to patients' ability to of those employed as doctors in private pay. On average, doctors saw over 230 companies also had a private practice. patients a month, with GPs and specialists in Involvement with private practice increased Bali reporting nearly 300 patient visits per gradually with age, i.e., from 75 percent for month, while physicians in the Outer Islands those under 30 to 77 percent for those in the saw fewer than 200 per month. 30-39 interval and over 86 percent for those in the 40-60 age range. A large share (85 2.20 Average monthly earnings from private percent) of those who had retired from the practice varied by region, between GPs and civil service continued to see private patients. specialists and amongst different specialist Finally, private practice was more extensive in categories (Table 2.5). Java, including Java, Bali and Sumatra (85 percent) than in Jakarta, and Bali emerged as the most Jakarta and Kalimantan (78 percent), Sulawesi remunerative locations for private practice. In (62 percent) and Irian Jaya (53 percent). fact, rural health center GPs in Java and Bali earned more on average in their private work 17 Table 2.3: AVERAGE NUMBER OF DAYS PER WEEK AND HOURS PER DAY IN PRIVATE PRACTICE, BY REGION, 1993 I Jakarta Java BalI Sumatra Kalimantan IOther Total Days Hours Days I Hours |Days Hours |Days Hours Days Hours Days I Hours I pays|Hours All Doctors __ Rural - - 15.7 13.7 15.7 12.8 16.1 13.6 1 6.1 13.7 15.8 13.8 15.8 1 3.6 Urban 5.0 2.8 5.4 13.3 5.0 2.7 j5.7 13.4 1 6.1 2.9 15.8 13.4 5.5 3.1 GPS Rural - - 15.8 13.8 15.8 2.8 6.1 3.6 6.1 3.7 15.8 3.9 5.8 3.6 Urban 15.2 2.9 15.5 13.4 15.7 2.6 5.8 13.5 1 6.2 3.1 15.9 13.3 5.6 1 3.2 GPs (Health Center) Rural - 5.8 3.8 5.9 2.5 6.0 37 6.1 3.7 5.8 4.0 5.8 3.7 Urban 5.4 2.6 5.8 3.5 6.0 2.2 6. 1 4.1 7.0 6.3 3.9 5.9 3.5 S ialists 4.5 12.8 15.2 13.0 15.2 13.0 15.2 12.8 15.5 12.0 15.3 13.5 15.0 13.0 Source: IDI Survey of Private Practitioners, 1993. than urban physicians including specialists, in (pegawai tidak tetap or PTT). This policy different Outer Islands. Monthly income for change was required, in part, because of the specialists in Jakarta, Java, and Bali was fiscal imperative of restricting overall civil substantially, sometimes three times higher service growth. MOH now has the flexibility than that for GPs. Internists and to hire PTT doctors without waiting for the obstetricians/gynecologists earned the highest authorization, i.e., the formasi, for new private practice incomes. positions. Moreover, it can now induct smaller batches of GPs several times a year- 2.21 Finally, private practice accounted for problems had arisen in absorbing large a large proportion of total income for incoming doctor cohorts on a once-a-year government doctors in all locations. The basis. The first group of 924 contract doctors share of private earnings within total income were hired in February 1992 and will varied from an estimated 70 percent or higher "graduate" in February 1995. There were two for specialists in Jakarta, Java, and Bali to 25- additional batches in 1992, 865 and 597 70 percent for rural health center physicians in doctors, and two more in 1993, comprising the Outer Islands. 805 and 955 physicians. 2.22 The Contract Doctor Scheme. Since 2.23 While in service, contract doctors have 1992, MOH has relied on quasi-contractual the same rights and obligations as regular arrangements to mobilize physicians for government employees. However, the salary service in public facilities. Newly graduated and special allowances of PTT physicians are GPs are no longer hired as civil servants. far more attractive than the salary and benefits Instead, their compulsory assignment is received by civil service doctors. The initial handled through a non-renewable three-year batch of contract doctors were offered a base appointment as "non-permanent employees" salary of Rp 300,000 (US $142) a month as 18 Table 2.4: AVERAGE FEES CHARGED BY PRIVATE PRACTITIONERS BY REGION, 1993 (in Rupiah) Jakarta Java Bali Sumatra Kalimantan Other Total All Doctors 11 Rural - 4546 3611 3197 4494 | 5227 5578 Urban 7956 T 5658 5031 4521 4945 [ 4541 4473 Rural - J 3923 3581 3147 4494 5146 4186 Urban 5587 4622 3579 4325 4267 | 3752 4522 GPs (Health Center) | Rural - J 4029 3507 3036 4494 5186 4268 Urban 4000 3589 1 3250 [ 4750 - T 3724 1 3704 Specialists I All 12329 8809 8589 5500 8000 7947 9457 Pediatricians _ 7854 Surgeons 11045 Internists 11500 Obstetricians/ Gynecologists ___ 10969 Others I _ I _ 9058 Source: IDI Survey of Private Practitioners, 1993. well as special allowances of Rp 200,000 (US 2.24 Another distinguishing feature of the $95) for service in remote areas and Rp new system is the stated intention of placing 400,000 (US $190) for assignments in very doctors in one of the locations they themselves remote health centers. PTT doctors also have chosen. Previously, the MOH Personnel receive housing, income tax and travel Bureau had greater flexibility and leeway in allowances and other benefits, the right to specifying assignments. Incoming PTT conduct private practice after working hours, doctors will now be asked to make three and "priority" consideration as candidates for selections from a list that typically includes specialist training. The salary scale has been their province of study, areas in adjacent revised upwards on two occasions since provinces, and more remote locations in February 1992. The latest batch of PTT specified Outer Island provinces. Government doctors (October 1993) was offered a base officials have stated that no one will be forced salary of Rp 500,000 (US $237) per month as to serve in a remote area. well as bonuses of Rp 350,000 (US $166) and RP 550,000 (US $261) for service in remote Consequences and Issues and very remote locations. By contrast, a civil service doctor working in a puskesmas 2.25 What consequences have followed from earns about Rp 400,000 (US $190). GOI's medical education and physician 19 Table 2.5: AvERAGE MONTHLY INCOME FROM PRIVATE PATIENTS FOR PRIVATE PRACTITIONERS BY REGION, 1993 (in thousands of Rupiah) Jakarta Java Bali Sumatra Kalimantan Other Total All Doctors Rural | 1 1030 1211 694 783 823 | 953 Urban 1713 1923 1283 743 942 884 1602 GPs Rural - 938 1130 694 J 783 771 884 Urban 1246 | 1824 840 718 973 793 | 1384 GPs (Assignment to Health Centers) Rural - 939 1236 578 783 374 | 754 Urban 513 | 965 826 1122 - 663 | 873 Specialists All 2601 2160 2508 867 800 1483 2203 Pediatricians l l | | 2389 Surgeons l l | 1282 Internists l l | | 4088 Obstetricians/ Gynecologists _ l _ _ 3837 Others l l _ _ 1542 Source: IDI Survey of Private Practitioners, 1993. employment initiatives and what policy options virtually unchanged, increasing from 31,138 in are worth considering? The following 1980 to 32,267 in 1985 and then falling to paragraphs consider the impact of these 30,608 in 1992. Moreover, this ratio rose in measures from several perspectives including 15 of the country's 27 provinces between 1985 observed changes in the distribution of and 1992 (Table 2.6). doctors, the sustainability of physician policies, and the effectiveness of service delivery in government facilities. Some policy 2.27 On the other hand, advances have alternatives are also outlined. occurred in the number and distribution of puskesmas doctors. In 1980, half of the 2.26 Physician Distribution: Health country's 4800 health centers lacked Centers. One significant policy achievement physicians. By 1985, 83 percent of health has been the continuing improvement in access centers had at least one doctor; on average, to doctors based in health centers. These each puskesmas had .9 doctors. Further gains gains are not due to notable increases in the had been registered by 1993. There were availability of health centers. Overall, the 6848 GPs, spread over 6242 health centers population: health center ratio has remained (1.1 doctors per puskesmas). Only nine 20 Table 2.6: POPULATION AND DOCTORS PER HEALTH CENTER BY PROVINCE, 1985 AND 1992 1985 1992 Population Doctors per Population Doctors per per Health Health per Health Health Center Center Center Center D.I. Aceh 20,418 0.66 22,701 0.91 Sumatra Utara 33,137 0.97 34,708 1.15 Sumatra Barat 26,000 0.93 24,672 1.03 Riau 27,326 1.13 28,933 1.40 Jambi 22,737 0.91 26,454 1.01 Sumatra Selatan 32,401 0.99 29,954 1.09 Lampung 47,515 1.02 46,860 0.95 Bengkulu 11,848 0.85 11,180 1.08 DKI Jakarta 27,961 1.35 35,926 1.38 Jawa Barat 49,972 0.98 32,109 1.20 Jawa Tengah 44,446 0.99 41,222 1.19 D.I. Jogyakarta 29,376 1.09 33,871 1.08 Jawa Timur 37,991 0.88 39,702 1.08 Kalimantan Barat 20,547 0.74 25,455 1.12 Kalimantan Tengah 12,391 0.61 25,770 0.84 Kalimantan Selatan 18,460 0.68 18,744 0.79 Kalimantan Timur 14,110 0.64 13,679 1.29 Sulawesi Utara 20,833 0.97 19,734 1.24 Sulawesi Tengah 21,375 0.82 23,024 1.14 Sulawesi Selatan 29,730 0.80 25,994 1.07 Sulawesi Tenggara 15,696 0.59 5,973 0.84 Bali 31,783 1.01 25,251 1.74 Nusa Tenggara Barat 34,625 0.92 27,024 1.08 Nusa Tenggara Timur 22,774 0.53 26,100 0.69 Maluku 16,330 0.60 15,978 0.79 Irian Jaya 10,856 0.32 19,498 0.65 Timor Timur 10,230 0.61 15,110 0.88 Indonesia 32,267 0.90 30,608 1.10 Source: Ministry of Health. 21 provinces, compared to 22 in 1986, had not of health center physicians in urban areas is achieved the staffing norm of one doctor per seen in many provinces, e.g., Sumaa Barar, health center (Table 2.6). Bengkulu and Kalimnntan Tengah (Table 2.7). A related phenomenon, the clusterng of 2.28 Of course, doctors are not spread doctors in a fifth of the disticts, is seen in uniformly widtin provinces. A concentration many provinces as well. Ihis grouping of Table 2.7: CLUsTENG OF HEALTH CENMR (HC) DCrORS BY PfoVINCE, 1991 % of af HCs % of rural HC % HC doctors % HC doctors % nu HC Provinc wlo doctors w/o doctors in urban ares in top 20% La with two or doctos D.L Aceh 4.5 4.3 6.0 38.0 20.2 Sumat Utal 15.7 16.2 4.0 42.4 10.5 Sumatr Bart 6.7 7.0 27.4 36.6 1.4 Rim 4.2 3.7 16.2 33.8 20.6 Iambi 6.7 5.4 16.5 22.4 1.3 Suiafm Sek= 11.7 11.5 18.3 28.7 13.6 Beagkuln 2.4 1.1 41.4 26.1 7.7 LAMPtmg 5.0 5.6 14.5 31.9 7.8 DKI Jdakc 10.5 N/A 100.0 24.4 N/A Jnm Bust 4.4 4.0 12.0 39.2 7.0 Jla Tesk 4.4 4.6 11.4 36.1 7.0 DI Yogkrta 7.4 8.0 17.3 24.4 8.6 mn Tulnr 4.4 4.8 13.9 35.9 13.2 Bali 5.1 5.2 7.9 31.6 76.0 Nass Tauu 7.6 7.4 0.0 28.7 2.1 Nun TemS 46.0 48.2 0.0 30.4 0.6 TImor T=mu 27.5 19.8 0.0 32.6 5.2 Kaim;ta But 4.8 4.6 14.8 27.5 7.2 Krli;MatM 22.3 26.8 44.8 36.2 2.8 rulrn; Saba= 42.6 20.5 21.8 38.2 0.8 Kali;manta TM 8.0 9.4 15.9 38.8 28.0 Sulawea Utar 4.1 1.6 15.4 40.0 4.8 SulWwe Tagah 1.2 1.2 0.0 28.6. 17.7 Sulaemd Selatn 2Z24 47.4 18.3 44.2 0.7 Sulswesi Tugzra 22.7 21.3 0.0 34.1 2.1 Maluku 15.4 14.0 18.2 37.7 17.8 Ii Jays 59.6 52.8 0.0 41.3 0.6 Indonesia 11.2 11.2 18.0 33.7 8.6 Source: Ministry of Health, 1991 Health Can Survey. L& The proporton of all doctors working in the top 20% of districts, ranked according to the total number of govemment physicians. 22 physicians in particular districts, is present in special hospitals, e.g., those for mental the provinces which still lag in terms of the health, tuberculosis, leprosy and cancer overall doctor per health center index. As a patients. The remaining doctors (an result, the shortage of doctors that estimated 14,033 in 1993) are characterized much of the country in the 1970s employed in general hospitals of continues in a relatively small number of varying size and functional orientation districts. The problem is particularly acute in which are run by MOH, provincial and Irian Jaya, Central and South Kalimantan, district governments, the military, and East Timor, Central Sulawesi and Maluku private foundations and businesses. (Annex Table 2A). 0 There are 11067 doctors practicing in 2.29 Employment of Physicians in 337 government-run general hospitals. Hospitals. Government policies have also This number includes 2522 GPs, two- brought an increase in the number of doctors, thirds of whom work in class C and D GPs as well as specialists, working in hospitals (Table 2.8). Nearly 70 hospitals. There were an estimated 10,500 percent of the specialist physicians doctors in Indonesia in 1985 based in 683 work in class A and B hospitals. hospitals. By the end of 1993, the number of Indeed, the four class A hospitals hospitals had grown to 950 and the number of employ 1237 (23 percent) of the doctors had increased to 13,425, including specialist doctors. Another 2280 (43 4672 GPs and 3088 residents in specialist percent) work in the country's 33 class training programs. The remaining 5665 B hospitals. doctors were specialists, half of whom were internists, pediatricians, surgeons, or * Specialists tend to work in Jakarta. obstetricians. For instance, nearly a third of all cardiologists, over a third of the 2.30 The distribution of hospital-based country's anesthesiologists and roughly doctors, particularly specialists, has several a quarter of the pediatricians, surgeons noteworthy factors: and internists are based in Jakarta. Nearly 30 percent of Indonesia's * Roughly six percent of hospital resident interns are being trained in physicians work in an assortment of Jakarta. Table 2.8: DOCToRS EMPLOYED IN PUBLICLY OPERATED GENERAL HOSPITALS, DECEMBER 1993 Hospital Class Number of Total Number of GPs Residents Specialists Hospitals Doctors A 4 3029 59 1733 1237 B 33 4568 820 1468 2280 C 168 2823 1247 1576 D 132 647 396 251 Total 337 11067 2522 3201 5344 Source: MOH tabulations. 23 Table 2.9: GEOGRAPHIC PLACEMENT OF GRADUATING SPECIALISTS IN FOUR FIELDS, 1989-1993 Jakarta Java and Bali Eastern Sumatra and Total Provinces /a Kalimantan Pediatricians 43 111 46 90 290 Surgeons 10 78 32 72 192 Ob/gyns 22 96 37 81 236 Internists 23 114 25 61 223 Total 98 399 140 304 941 Percent 10% 42% 15% 32% 100% Compared to: 1990 Population 5% 57% 13% 25% 100% Distribution /a Through September 1993. Source: Ministry of Health. - After completing their studies, all time basis in private medical specialists must work in government facilities.3 class A, B, C or D hospitals for three years unless they return to military * Finally, one important goal for service. Assignments are not graded Repelita VI is to raise some or all of according to remoteness. Cumulative the country's 132 class D (no placements of specialists in pediatrics, specialists) hospitals to class C (seven surgery, internal medicine and specialists) standards. This upgrading obstetrics/gynecology during the 1989- has already begun in selected hospitals. 1993 interval appear to have favored For instance, there are already 82 the Outer Islands (Table 2.9). pediatricians and 33 surgeons assigned Nevertheless, it is striking that ten to class D facilities. Still, attainment percent of graduating specialists in of this Repelita VI goal would require these fields have been placed in large increases, e.g., up to 139 Jakarta. The current stock of internists, 122 obstetricians/ specialists is dominated by Java, gynecologists and 174 radiologists, in including Jakarta, and Bali, which staffing over a five-year period. account for nearly three quarters of specialist doctors working in hospitals 2.31 Limitations of Compulsory Service. (Annex Table 2B). Administrative allocation of these large numbers of GPs and specialists to government * There are 2473 doctors, including 946 facilities has been costly and difficult to specialists, listed as full-time sustain. Costs have included the salary and employees of private hospitals. Of allowances paid to an expanded number of course, many public sector doctors, physician civil servants working in health GPs and specialists, work on a part- centers, hospitals and provincial and district 24 administrative positions. An additional unemployed doctors have established their element, discussed earlier, has been the large own private practices, in rented or borrowed subsidy to medical education in public and premises in kainpongs or low income private universities. Specialist physician neighborhoods. Many of these untrained training is subsidized heavily as well. doctors appear to have done well in terms of net earnings. 2.32 A justification for the subsidy to undergraduate and postgraduate medical study 2.34 Because of increasingly lengthy waiting has been the requirement of compulsory periods and the availability of alternative government service, at times in remote health work, obligatory service was resented and centers, for GPs or provincial hospitals in the resisted by a growing number of doctors in the case of specialists. Whatever their initial late 1980s and early 1990s. Indeed, it was not rationale, MOH's deployment arrangements unheard of for "idle" doctors to use various became increasingly cumbersome during the tactics to postpone their puskesmas late 1980s when difficulties were encountered assignments or to attempt to negotiate in absorbing the annual outflow of medical placements in attractive locations. graduates. During the 1985-1987 interval, the Compulsory service was perhaps especially average medical school graduating class, unwelcome to private medical school graduates 1,345 students, was less than the 1,361 who had to finance much higher tuition and formasi authorized each year. But this related expenses than public students over a situation had changed abruptly by 1990-1992, longer period of study and examinations. when the average graduating cohort reached 1,674 while the annual number of funded new 2.35 It was this diminished enthusiasm that positions fell to 1,047. very likely led to MOH's efforts to make compulsory service more appealing. The 2.33 A small number of the roughly 627 duration of assignments in difficult areas was doctors who were not immediately "drafted" reduced and provisions for leave and in- by MOH were taken by other government service training were liberalized. And departments or the military. But most were allowances for postings in remote areas were required to wait, sometimes for two years or added to the benefit package. By introducing longer, until directed to begin their rewards for service in difficult areas, MOH compulsory service. The jobless, still-to-be recognized, at least implicitly, how important dispatched doctors were not authorized to private income earning opportunities had practice medicine since they had not become for doctors, even for those just completed their puskesmas assignments. emerging from medical school. Additional Nevertheless, quite a few of these young allowances were not offered to GPs and physicians, especially the 400 or so graduates specialists working in facilities in urban and from medical schools in Jakarta, found periurban areas and in relatively prosperous temporary work in private clinics and rural settings, e.g., Java and Bali. This was hospitals. One not uncommon practice was understandable since, as discussed above, such for licensed doctors to open clinics which doctors could count on a regular flow of would then be staffed by unemployed medical private patients at remunerative fees. This graduates. In fact, many of the 24-hour probably accounts for the limited turnover of clinics and private hospitals which have doctors working in health centers in urban sprung up in Jakarta have recruited newly areas and in Java overall. For instance, credentialed doctors who were waiting to be almost two-thirds of the puskesmas doctors in assigned. It is also thought that some Semarang and Boyolali districts of Central 25 Table 2.10: HEALTH CENTER DOCTORS: TIME IN CURRENT ASSIGNMENT BY PROVINCE (In percentages) Less than 6 months 7-12 months 13-24 months More than two l ____________________ _____________________ y ears East Kalimsintan 32 26 23 19 I (1992) Nusa Tenpaara 25 39 13 23 Barat (1991) West Java (1993) Ciajur 15 6 37 43 Sumedang 13 13 28 46 East Java (1993) Semarang 26 0 9 64 Boyolahi 4 24 8 64 Irian Java (1993) L Jayapura 12 28 32 28 Source: Kaltim/NTB survey and MOH records. Java and over 40 percent of those in Cianjur under the present system, the contract doctor and Sumedang districts of West Java had been scheme. Indeed, current arrangements rely in their jobs for more than two years in 1993 more explicitly on individual self interest as an (Table 2.10). allocational device. Thus, allowances for PTT doctors have been enhanced several times, and 2.36 On the other hand, allowances were compensating job security and pension benefits seen as necessary to draw doctors to less enjoyed by civil servants and associate career developed areas, especially to health centers in considerations have been withdrawn. which opportunities for private practice were constrained because of low average income, 2.38 Sustainability of the Contract and possibly low population densities and Scheme. Due to these and other features, the competition from traditional healers. sustainability and effectiveness of the contract scheme, as currently construed, are in doubt. 2.37 In short, by providing special One concern relates to GOI's ability to bear allowances, MOH acknowledged limitations to the costs of hiring, through increasingly its ability to use compulsion or administrative attractive contracts, virtually all Indonesian means to deploy physicians to improve access medical graduates. The most recent class to health services. The government, in effect, numbered 1898, and future cohorts may be entered into continuing negotiations with even larger as students in several recently doctors, with the latter responding to income established universities begin to graduate. earning possibilities and other factors, e.g., Thus, each incoming PTT class is likely to be educational opportunities for dependents, that at least 80 percent larger than the average figure in individual location and employment number taken in via formasi in the 1989-91 decisions. What is essentially a bargaining interval. The implied increase in process has carried over and even intensified compensation costs may be even higher 26 because of the larger salaries paid to contract 2.40 Adding to these concerns is the doctors. However, the lack of pension growing reliance on contract physicians in obligations to PTT doctors may yield some areas outside Java and Bali. Since the scheme offsetting savings. was initiated, almost half of the 4145 PTT doctors have been sent, in five batches, to 2.39 Absorbing large and still growing remote and very remote areas, mostly in the graduating cohorts may also lead to Outer Islands. Through October 1993, administrative strains and personnel problems. contract doctors accounted for a third of health For instance, the special incentives paid to center physicians in the Eastern Islands as PTT doctors have not been well received by compared to 16 percent in Java and Bali civil service physicians who still make up a (Table 2.11). However, the recent large share of the health work force. (September 1993) government pledge not to Moreover, allowances have not quieted compel graduates to serve in remote areas may concern among contract doctors, some of have already created problems in inducing whom are still disposed to explore means of doctors to serve in the Eastern Islands and avoiding unwanted assignments. Here, there elsewhere. This may be seen in the sharp are reports of graduates refusing offered drop in the proportion of recruits who accept postings, preferring to wait for more attractive remote area assignments and in the difficulties opportunities to open up. Some physicians are some provinces, e.g., Irian Jaya, East Timor, fearful that provincial governments, which Aceh, Central Kalimantan and Southeast have been authorized to handle local Sulawesi, have experienced in obtaining deployment of PTT doctors, will be tempted doctors for available jobs (Annex Table 2C). to economize by including difficult and The problem may have been compounded by unattractive locations in the set of "normal" stated provincial preferences for male doctors assignments which do not offer special to fill positions in very remote areas (Table allowances. (Some provinces are making 2.12). Since nearly half of all new graduates efforts to win over medical graduates, by, for are women, this limits greatly the pool of example, sending officials to brief students available recruits. Indeed, gender as well as and faculty members on health center other factors may account for the placement of conditions and opportunities.) Doctors are greater number of PTT doctors in Jakarta and also perplexed and insecure about what awaits West Java than had been requested (Annex them when they finish their compulsory Table 2C). service. They have been advised that only a small share of each PTT batch, initially a 2.41 Similar issues are likely to arise when quarter, and then progressively less with each MOH begins to assign specialists to class D graduating cohort, will be absorbed into the hospitals slated for upgrading to class C civil service, while the rest will need to find status. While specialists remain civil servants, work in the private sector. However, contract their often lucrative private income earning physicians are not being offered practical prospects make them prime candidates for information on income earning prospects in relocation to major urban centers and different regions or jobs, e.g., private provincial capitals once their compulsory, insurance companies, or assistance in three year hospital assignments come to an acquiring skills, such as elementary end. The same considerations may make them accounting, that would be useful in developing less than willing to accept assignments in a private practice. hospitals located in remote or impoverished areas. MOH may need, therefore, to develop special salary and benefit packages to attract 27 Table 2.11: DISTnuBUrION OF CIVIL SERVICE AND CONTRACT DOCTORS BY REGION General Contract Total DEPKES Regional share Contract physicians, civil physicia physiians of total physicians as service, 199 /a through 10/93 physicians /b share of total Lc Jakarta 2,567 116 2,683 14% 4% Elsewhere in Java and Bali 6,348 1,211 7,559 40% 16% Sumatra and Kalimantan 3,816 1,793 5,609 29% 32% Eastern Ilands 2,129 1,025 3,154 17% 32% Indonesia 14,860 4,145 19,005 100% 22% /a No new general physicians have been hired into the civil service since 1991. /b For comparison, population was distributed across Indonesia as follows: Jakarta, 5%; elsewhere in Java and Bali, 57%; Sumatra and Kalimantan, 25%; and Eastern Islands, 13% as of 1990. Lc This measure is a rough approximation. It assumes no civil service physicians have left service or been reassigned across regions since 1991, and that no contract physicians have been reassigned across regions or left before their service was up. Source: Ministry of Health. Table 2.12: FlH BATCH (OCTOBER 1993) OF CONTRACT DOCTORS: GENDER SPECIFIED IN PROVINCIAL REQUESTS MALE FEMALE GENDER NOT SPECIFIED TOTAL NORMAL 240 147 199 586 REMOTE 148 91 98 337 VERY REMOTE 85 0 22 107 TOTAL 473 238 319 1030 Source: Ministry of Health. specialists to former class D hospitals. itself has already been revised several times. As discussed, the latest refinement, which 2.42 In summary, MOH has continued to enhances allowances while giving conscripted make adjustments in the mix of subsidies, doctors a greater say over where they will compulsion (conscription and administrative serve, seems to entail costly and unsustainable allocation) and financial and non-financial financial outlays. The current version may incentives it has used to widen access to also run the risk of leaving some remote and doctors in Indonesia. The contract scheme 28 impoverished areas without adequate health allocation and deployment mechanisms, have services. a bearing on demand. 2.43 In principle, the government could 2.45 In this regard, low average levels of respond at the present time by reinstating facility use indicate that current deployment some of the mandatory features of the contract policies, including staffing norms, scheme, and narrowing or elirninating the conscription, mandatory postings, short- discretionary power it has conceded to new duration assignments, and the right to practice graduates to select their assignment areas. privately, may not have a sufficiently high This, however, would very likely engender payoff in health terms. One question is different forms of evasion and non- whether current or prospective utilization compliance. A major administrative effort levels justify placement of seven core would then be needed to implement and specialists in different class C hospital. Exact enforce provisions of the scheme. information is not available on the number of Alternatively, MOH could consider further cases which different class C hospital weakening or even doing away with specialists typically treat. However, compulsory medical service, which has created persistently low bed-occupancy rates (Table resentment and led to avoidance and 1. 1) suggest that specialists in many class C resistance.4 Instead of the "stick," health facilities are not very busy.5 For example, policy makers could rely exclusively on there were 76 seemingly underemployed benefits or expected earnings packages to doctors (including 46 specialists) working in attract physicians to different postings, the five class C hospitals with occupancy rates including those in remote areas. For example, of less than 30 percent. In such settings, it salaries (or grants) to directly recruited would perhaps be more cost-effective to rely doctors in some settings may need to be set at on GPs who have acquired some specialist levels comparable to the earnings of Java- skills through pre-service or in-service based doctors with public and private practice training. (This would probably entail some income; as additional inducements, paid leave curriculum changes and adjustments in in- to a designated province, and subsidized service training options--see discussion children's education in boarding schools could below.) The advantage of this approach is be provided as well. that it would reduce demand for specialists from within the government system and make 2.44 Demand-side Issues. In deciding the it possible to cut the subsidy to specialist future of the PTT scheme, it is not only the training. Staff utilization is an issue in class costs of maintaining over 6000 doctors in D hospitals as well. For instance, there were government facilities which should receive 219 doctors (including 37 specialists) working attention but also the productivity, in the class D hospitals which had utilization performance and health impacts of these rates of 30 percent or less. trained workers. In other words, supply-side policies relating to training and deploying 2.46 Similar questions arise in respect of the doctors should not be reviewed without taking 4200 doctors assigned to health centers. As account of demand-side considerations. discussed in Chapter One (see Table 1.3), Levels and likely trends in the demand (need) there is considerable variation in the average for doctors in hospitals and health centers number of daily puskesmas visits, both should be identified. It is also important to between and within provinces. Contact rates determine which factors, including various tend to be much lower outside Java and Bali, though there were districts in these regions 29 which reported low average puskesmas 2.48 Making Doctor Allocations Demand- utilization. These low visit rates suggest that Sensitive. The upshot of this discussion is doctors in many health centers may be that staffing targets for doctors in many underemployed. The same finding emerged government facilities seem to have been set from estimates of the extent to which too high relative to actual demand. That is, physician staffing patterns contribute to current norms, e.g., one to three doctors per observed differences in utilization rates. In puskesmas, three GPs for each class D this exercise, 1991 kabupaten-level data were hospital and 11 GPs and seven specialists for used for Bali, East and West Java, Jambi, every class C hospital, appear excessive in Kalimantan Selatan and Timur, Sulawesi view of low average levels of facility use. Tenggara, and Irian Jaya. A regression was Existing staffing guidelines, therefore, may specified linking the mean number of daily encapsulate exaggerated and possibly visits per facility in each kabupaten (V) to the unsustainable levels of demand for costly average number of doctors per puskesmas (D), skilled manpower. the average number of people in the coverage area (P), and a variable (M, the proportion of 2.49 This finding that demand has been centers equipped with maternal and child overestimated is, however, of only limited health kits) representing the availability of policy relevance in and of itself. Health medical instruments, drugs and other authorities also need to know to what extent materials. Also included in the specified staffing standards exceed actual requirements, equation was a variable designating whether and how to recalibrate norms or find facilities were located in provinces outside alternative ways of assuring that demand-side Java and Bali (0). considerations are reflected in staffing patterns. 2.47 Regression results (Table 2.13) indicate that utilization levels varied directly but only 2.50 In fact, there are no straightforward, weakly with the number of doctors on hand. experience-tested ways of deviating from the Findings suggest a low degree of norms-driven approach that has underpinned responsiveness of utilization to placement of work force policy to date. One option, which doctors. For example, a ten percent increase is already in use to some extent, is to allow in the availability of physicians would increase staffing standards to vary geographically or average daily visits by only 3.2 percent, while according to other criteria. This practice, placing a second doctor in all facilities would which is exemplified in MOH's use of raise utilization by only 18 client visits per somewhat smaller health center teams in the day, i.e., 30 percent.6 Outer Islands, still suffers from the weakness that such norms may be inconsistent with actual or potential staffing needs in many facilities. Table 2.13: UTILIZATION OF HEALTH CENTERS [ V = 1.90 + 17.98 D + 0.71 P + 42.77 M - 14.87 1 (0.12) (2.02) (3.05) (5.72) (-1.79) [t - statistics in parentheses; R-squared = 0.54; number of observations = 94] Source: Ministry of Health service statistics and the 1992 Health Center Survey. 30 2.51 A second alternative, illustrated in the setting exercise which reflects known budget Indicator of Staffing Needs (ISN) methodology constraints. Several preconditions need to be which has been discussed in Indonesia, gets met for this approach to work effectively. For closer to reality by determining personnel instance, managers need to be able to adjust requirements on the basis of the actual volume staff assignments, and transfer or remove of outpatient and inpatient visits to facilities. ineffective workers. Administrators also need One problem with this approach relates to its to know they will be held accountable for complexity--a huge computational effort is facility performance and for the effectiveness imposed on central or provincial planning of their staffing decisions.7 units which had relied previously on unchanging, uniform norms to guide staff 2.54 A second advantage and opportunity allocations. A second concern relates to how inherent in a facility-grounded approach is the client demand is estimated. Here, there are attention that would necessarily be given to several risks. The ISN approach may create operational procedures and phenomena, and facility-level pressures to overstate patient care management challenges within hospitals and loads in order to maximize staff numbers. On health center. As discussed in Chapter One the other hand, this methodology may in (paras. 1.28-1.29), these process-related effect, legitimize existing, unsatisfactory factors may well have a large bearing on service delivery patterns, and may give utilization through their impact on the range excessive weight to current demand which and perceived quality of services offered to may be low due to unfavorable (but clients. Of course, limited access to facilities, remediable) local perceptions of service difficulties faced by the poor in paying service quality. Differences in infrastructure and fees and the non-availability of drugs, geography may also be overlooked in this equipment and staff all probably contribute to approach. low average service utilization. Yet none of these factors seems to account for the low 2.52 Because of these problems, it seems utilization rates in the many facilities which worthwhile to explore alternative ways of seem conveniently located and adequately obtaining demand-sensitive estimates of supplied with inputs.8 requirements for doctors (and other staff). The suggested route is to design mechanisms 2.55 A facility-based demand estimation which allow facility-specific requests to process will afford opportunities to identify emerge. These are different ways, none well and address management and supervision established in developing countries, of practices and staff behavior which may be pursuing such an approach. But before diminishing service effectiveness and creating discussing possible options, some advantages unfavorable local perceptions of service and opportunities inherent in a facility-based quality. In this regard, existing staffing and approach should be stated. deployment arrangements may contribute to persistent operational difficulties. Here, a 2.53 One benefit is that facility-generated source of problems is the short duration requirements would be more realistic since service pattern that was established in the they would derive from the detailed plans and 1980s. Unlike Java and Bali where five-year projections of hospital and health center assignments were the norm, replacement of managers. Obviously, such estimates should puskesmas doctors every 2-3 years or sooner not be a "wish list" of desired staff, but would became a fact of life in the Outer Islands. need to emerge from a detailed analysis of This is illustrated in the large proportion local health problems linked to a priority (roughly 60 percent) of doctors in Kaltim and 31 NTB who had been in their jobs for a year or satisfactory care during working hours and less; by contrast, 40-64 percent of the Java- would then be induced to attend their private based doctors had been in position for over practice. two years (Table 2.10). Short assignment periods and sometimes lengthy absences for 2.57 Developing a Demand-Sensitive leave and training have meant that doctors Approach to Staff Allocations. A demand- have had less time and possibly little responsive approach to determining staffing inclination to establish linkages with local requirements for doctors and other workers communities, and to become familiar with needs to be instituted in Indonesia. This will local health patterns and needs. Puskesmas require more than the introduction of doctors in such settings also have had fewer differentiated staffing standards or planning opportunities to size up and improve the tools like ISN. Instead, what are entailed are performance of available staff. some perhaps substantial adjustments in the way district health programs and facility-based 2.56 Another cluster of problems derives activities are planned and managed. Likely from the doctor's right to conduct private changes include greater decision making and practice. This contract-protected option can allocative authority, and heightened afford significant money making opportunities accountability for district, hospital and for doctors because of the relatively short puskesmas administrators, systematic priority official working day and due to the availability setting (especially in health centers) and in most settings of patients willing to pay for stronger incentives and more demanding private care. Another incentive to private supervision for staff. These and other steps practice is the fact that government salaries can be expected to affect the range and quality and allowances are largely seniority-linked of available service demand, and in this rather than effort or performance-based. fashion, to result in reality-tempered, facility However, this opportunity for doctors brings specific staffing requirements. possible risks for puskesmas operations. This is because private practice requires puskesmas 2.58 The aim of these adjustments is fairly doctors to face a potentially complex interplay specific-a system of hospitals and health of decisions. For example, health center centers in which administrators, including physicians need to decide which patients or district level officials, have the incentives and medical conditions should be treated on a instruments needed to arrive at realistic and public or private basis, and what fees should sustainable staffing requirements. But what be charged for different private clients or routes and processes can be employed to services. And doctors need to work out how arrive at this end point? Here, there are much energy and intensity to apply to different several approaches which should be pursued public responsibilities, e.g., curative care, simultaneously as part of an eclectic, supervision and outreach, and administrative opportunistic process of policy development. tasks, as compared to private practice, and what their priorities and obligations are to 2.59 First, in-depth review is needed of why different categories of patients and to the local the government-run health system has still not community. One troubling hypothesis, due to generated much public interest or utilization in Sciortino (see Box 1. 1), is that doctors tend to many localities, and what are the perceptions curtail their curative work in the puskesmas in of patients, non-users, and staff in those areas part because of the demands and rewards of as regards the range and quality of services. administrative duties, but also in the At the same time, the experience of those expectation that many of the ill would not get health centers and the specific services that 32 have been embraced by communities and expanded job descriptions which allow scope brought significant health impacts needs for initiative taking and discretionary use of careful analysis. For example, what was authority; and new incentive arrangements noteworthy about the doctors-in-charge of aimed at attracting doctors to difficult areas successful health centers? How have they for lengthy periods. motivated and energized their staff and achieved credibility in the local community? 2.62 Finally, pilot efforts should be used to What role was played by the dokabu in such test delivery systems which have shown useful areas? results in other countries. This could include efforts to strengthen service delivery by 2.60 In addition, the assessment of facility linking staff income to performance, and by and staff performance needs to be giving local communities advisory and review strengthened. For instance, performance powers. One potentially promising reviews could be based on activity and mechanism would be to contract out health facility-level indicators of service costs and center activities, ranging from specific tasks to impacts. Such exercises could begin with overall operations, to qualified NGOs or simple measures of staff time, drug and private individuals and firms-this approach is resource use, combined with information on being tested currently in Mexico and several the number and range of client contacts and other countries. NGOs could be required to outcomes, and indications of which workers compete for service opportunities, submitting are performing well and why. Indicators proposals that would be judged as regards the could then be refined further, e.g., through scope, focus and quality of services they follow-up studies of patients and longitudinal intended to make available, expected staffing surveys of health conditions in the area. and resource needs, fees and unit costs and the Information on the service delivery process process and impact objectives that had been needs to feature in such efforts, including established. The providers who were selected monitoring of diagnosis and treatment, referral could then be monitored and evaluated patterns, attention to health education, and according to agreed milestones and indicators. indicators of patient satisfaction. Staff beliefs and goals in respect of treatment processes Supply-Side Implications and Concerns need to be evaluated as well. The assessment process should also take note of community 2.63 Demand-side Scenarios. To sum up, expectations and concerns, using consultations the costs and problems associated with with local residents, especially the poor.9 supplying doctors through the PTT mechanisms and low average utilization of 2.61 Revised assessment techniques and facilities and staff together point to the need performance indicators, together with findings for policy changes, including adjustments in from best and worst case analyses, can be the way health care is managed and services used to identify problems and areas of are delivered. Health centers, for instance, potential improvement. These should be should focus on building competent teams that pursued through a series of closely monitored provide wanted, high quality services. These experimental efforts. For instance, trials facilities need to be headed by committed and could test the effectiveness of different ways experienced managers. These would normally of motivating and guiding facility managers; be individuals age 30 or older, who are not changed supervision and evaluation trying to enrich themselves but who should not procedures; scaled-down service packages; have to pay a financial penalty because of strengthened priority setting mechanisms; their willingness to work for lengthy periods 33 in possibly difficult areas. This argues against centers, especially if some facilities were using recently graduated TT doctors, on one- turned over to NGOs, private practitioners or time (compulsory) three year contracts, in even non-physicians with public health skills. such assignments. Instead, health authorities Moreover, the government may decide to could hire doctors with appropriate focus on health center and hospitals in poorer characteristics using attractive incentive settings, relying on a smaller number of packages. Other countries have used this experienced, professional health managers, approach successfully. For instance, Canadian hired on contract or through a national health provinces have used different pay scales, corps. But whatever staffing requirements settlement allowances, continuing education emerge in this fashion, it is unlikely that there expenses and scholarships for later study to will be sufficient demand to ensure effective interest doctors in remote area assigmnents. use at the puskesmas level of most graduating The Chilean government has used similar medical students. Therefore, there is little benefits to attract physicians to rural areas. need to retain the compulsory service obligation in its present form or scale. 2.64 Incentive packages need to be designed to appeal to the large pool of 2.66 Balancing Demand and Supply. potential applicants. This would include not Demand-side changes may need to be only the current PTT doctors whose contracts accompanied by changes in the number of will be expiring shortly, but also existing civil doctors and specialists produced in the service physicians, private practitioners, and country. By some criteria, Indonesia still in exceptional cases, newly graduating medical faces a shortage of physicians. For instance, students. As mentioned, the benefit package the population per doctor ratio of roughly would need to compare favorably with real net 6200, although much below the 1970 figure earnings available to established doctors in (27,000), is well above that in many urban and periurban areas in Java. With such developing countries, e.g., Egypt (1320) and incentives, authorities should be able to India (2460). However, doctors may not be establish rigorous entry qualifications, allocated efficiently in countries with lower selecting candidates with a flair for the job, ratios. For instance, population ratios empathy for local community members, good typically include all qualified staff regardless management and public health skills, and a of what jobs they perform, their workloads willingness to see the job through. The appeal and tasks and how services are organized of this work may be enhanced by establishing (Shipp, 1993). Accordingly, such ratios do a special, autonomous national health corps of not represent a valid or useful standard for professional health center managers who work planning and policy development purposes. in difficult locations. Indonesia itself provides a good case in point. Although the population/doctor ratio has fallen 2.65 Once new contracting mechanisms and sharply, there have been absorption problems incentive packages were in place, demand for and concerns about how effectively physicians doctors would be responsive to results, are used in many assignments. specifically via delivery arrangements which generated interest and led to enhanced 2.67 Instead of striving to reach a certain utilization and health impacts. However, the ratio, it would be better to let the numbers of actual demand for doctors that would emerge medical graduates and specialists be governed following such changes is difficult to predict. largely by income earning opportunities for For instance, in one scenario fewer doctors doctors in public and private jobs. This can would be needed in government health be accomplished, in part, by reducing sharply 34 subsidies to public and private medical 2.70 Such phenomena have been noticed in education. Such a policy change, which a number of countries. For instance, the should be pursued as part of a wider reform of number of physicians rose from 20,000 to higher education subsidies, would require 166,000 in Mexico between 1960 and 1990. students to finance a much larger part of their A 1986 survey found that a quarter of all training costs. This could lead to a decrease doctors were jobless, underemployed or in applications to undergraduate medical employed in non-medical activities. Mexico school and specialist training programs, if responded by creating a commission with private rates of return fell significantly as a representatives from the ministries of result of rising tuition fees. An offsetting education and health, health care institutions factor would be the opportunity, if compulsory and universities. This group agreed to reduce service is ended, of entering the labor market enrollment and contain the number of medical immediately after graduation. There would be schools. Elsewhere, Egypt reacted to a reduced flow of graduates if student attrition physician overproduction by cutting medical rates increased or if medical schools retained enrollment in half in 1982. Other countries current admission standards or imposed higher have responded to actual or threatened doctor criteria for program completion. oversupply by promoting outmigration, restricting immigration and reducing 2.68 Large increases in tuition would need physicians' working hours (see World Bank, to accompanied by efforts to assure that poor 1993). students have the opportunity to attend medical schools. This can be achieved by offering 2.71 Medical Education: Quality Issues. many more scholarships to low income While fewer graduates may be absorbed, those individuals, expanding student loan schemes, who do work in health centers will need to and reserving some places for qualified acquire management, public health and applicants from particular provinces. Loans community development skills and also have should be made available to students attending reached a satisfactory level of clinical public as well as private universities. Loans competence, including an adequate command and scholarships could be linked to career of surgical, obstetric and other basic specialty decisions. For instance, beneficiaries could be skills. Of course, those destined to work in asked to make a longer term commitment of at private practice or hospitals will need solid least seven years of service in health centers in clinical training as well as some grounding in remote areas. non-clinical matters, including accounting and other skills useful in operating a private 2.69 Medical schools, especially those that practice. are publicly run, may also need to make anticipatory adjustments, including cutting the 2.72 It is unclear what proportion of medical size of entering classes. Such steps may be students leave university with the required required because demand for medical range of skills. What is known, however, is education may not decline sufficiently since that the quality of medical education remains the status and prestige associated with uneven. In general, public medical programs becoming a physician may overwhelm market have sustained a higher standard of education signals such as tuition increases or falling than those in private schools, although there relative earnings. The result could be are some low quality public institutions and unemployment and other symptoms of some higher quality private programs. Private physician "oversupply". schools typically have special disadvantages. They rely on part-time teachers and have 35 weaker linkages to teaching hospitals and less accreditation status if objective standards are satisfactory physical facilities. Although pass not met. rates have improved in several private programs, continuing poor examination 2.74 CHS also has the responsibility of performance is one of the factors accounting periodically reviewing the national medical for the longer time private students take to education curriculum. With WHO support complete their studies. over the past twenty years, CHS has striven to revise medical curriculum and teaching 2.73 There are three mechanisms in place practices to better prepare physicians to work for assuring medical education quality: in community outreach and ambulatory care in assessments of medical schools, curriculum rural areas and to give less emphasis to reviews and student examinations. All of hospital-based, curative care. Even with the these mechanisms need to be strengthened. reorientation toward community medicine, For instance, medical schools are evaluated training for health center work is lacking in currently at the initial stage when the program several areas: field experience at the is being established. CHS assesses the puskesmas-setting is restricted to just a few proposal and conveys its evaluation to the weeks in medical school; community-outreach Directorate General of Higher Education in activities are particularly limited and graduates the Department of Education. Sometimes this from some universities have seen very few review process is bypassed. Only in one cases of endemic diseases that may be extreme situation has the Consortium problems in the puskesmas area. CHS has the recommended that a medical school be closed. responsibility for reviewing and periodically CHS's initial assessment is important, but updating the curriculum, but it has no quality could be better assured over time by authority to enforce the changes in curriculum. introducing a regular system of accreditation That is the responsibility of the individual of both medical schools and teaching hospitals. university. Again, ongoing accreditation Ongoing accreditation systems can give processes would provide the opportunity to institutions valuable feedback about how to evaluate how well universities are improve their functions can help to raise implementing the curriculum, and whether any quality to a certain minimum standard. rethinking is required of the content and pace Accreditation can be the responsibility of of medical training. For example, the amount private, voluntary agencies, as in the United of time devoted to basic sciences and to States. Or it can be under the Department of clinical medicine needs to be reviewed Education, as is done in many countries in periodically. And thought needs to be given Europe. The Philippines has used to how to equip GPs with rudimentary accreditation systems to improve the quality of "specialist" skills, so that they can cope private higher education (over 85 percent of effectively in class D hospitals.'" all higher education in the Philippines is private). In order to be accredited, the 2.75 Examinations are a potentially institution must have educationally appropriate important mechanism for maintaining quality. goals and the resources to achieve those At present, state universities have the objectives, and must have demonstrated that it autonomy to test their medical students, while is achieving those objectives currently, and private university students must take provided evidence that it will continue to examinations developed by the nearest state achieve those objectives in the future. Review university. Under this system, students in bodies need to have the authority to revoke public universities do not take a standardized test, and that for private schools is 36 standardized only for the private students in drug-resistant microbes, suddenly emerge. that region. Pass rate information is not Some practices taught in medical school in the available to the public. CHS does not even 1960s and 1970s are now viewed as compile this information. A uniform national ineffective. In this context, low-cost examination for all medical students, public mechanisms to provide physicians, both public and private, after each main phase of their and private, with continuing education medical training would help assure uniform opportunities are potentially important. These standards. Collecting and publishing pass mechanisms could consist of drug formularies rates and the length of time taken to complete (guidelines for drug prescribing) tailored to studies would help potential applicants decide Indonesia, practice guidelines, newsletters, between schools and would put pressure on journals, and short seminars and courses, new or weak schools to improve. either handled as distance education or in traditional settings. There are no systematic 2.76 Ownership of Medical Schools. Some programs in Indonesia for continuing of the adjustments recommended above may education for physicians, either on a voluntary eventually require a reexamination of the or mandatory basis. Physicians in the public number and role of different medical schools, sector do receive some in-service training. particularly those that are publicly controlled. IDI has sponsored some training and could It is not obvious that Indonesia needs 29 develop practice guidelines over the medium medical schools, some with small class sizes, term to assist public and private physicians. including 15 publicly owned institutions. GOI could consider providing more training Some restructuring (including, for example, opportunities as well as written materials, merging schools or rationalizing programs) of perhaps by contracting a professional this picture may be warranted in three to five organization or medical school to carry out years. In the mean time, it is important that this work. public schools, like privately owned institutions, be accorded extensive autonomy 2.78 Specialist Training: Numbers, in financial, administrative and educational Balance and Quality. Annual production of spheres, and be held accountable through specialists is still relatively low in absolute budgetary and accreditation processes and terms. Still, concerns need to be raised about market tests for results. The appropriate role the rapid rate of increase in the pool of of the government in medical education could specialists and about an emerging imbalance be taken up within a larger review of the between the four main general specialties, i.e., comparative advantage of publicly owned pediatrics, surgery, internal medicine and institutions in education training. In this obstetrics/gynecology and the consulting respect, one could envisage an outcome in specialties which include all other fields (para. which some public schools are privatized, 2.6 and Table 2.2)." Several issues in while the remaining ones focus on graduate respect of training and certifying specialists training and research in medical fields of also need to be noted. strategic national interest. 2.79 The rapid increase in the output of 2.77 Continuing Education. Ways of specialists is linked to MOH's policy of strengthening physicians' skills once they are placing a surgeon, pediatrician and so forth in in the labor force need to be explored. The each first-line public hospital. As discussed, medical field is always changing. New there remain large numbers of specialist prevention, diagnostic, and treatment therapies vacancies relative to staffing norms for class C are developed, while new problems, e.g., hospitals. For instance, it would probably 37 take more than five years at current graduation have a more costly practice style. And as and attrition rates to fill the 140 or more more physicians continue on to specialist currently available surgeon's positions in class training, fewer general physicians remain. It D hospitals. However, low bed occupancy needs to be kept in view that the bulk of the rates in C and D hospitals suggest that the health care needs of the population can be case load may not be sufficient to justify a full handled by well trained primary care nurses, complement of specialists. A reexamination general physicians, and surgeons, of the number of specialists needed coupled pediatricians, internists, and obstetrician/ with more stringent controls over total output gynecologists in appropriate ratios to each seem to be in order. Here, a key step would other. If a specialist needs to be consulted, be to allow hospitals to decide, relative to the vast majority of problems can be handled their budgets and expected revenue flow, what by a general specialist. For cost effective number and mix of specialists they need and care, general physicians should outnumber what they can afford. Government hospitals general specialists, and general specialists could also begin to hire and fire specialists should well outnumber consulting specialists. directly, as in the private sector, and to offer private specialists fee-sharing opportunities on 2.81 Recognizing this, the Canadian a part-time basis. This would be consistent province of Ontario has established guidelines with the decision, under the Lembaga for the distribution of physicians: 40 percent Swadana program, to convert public hospitals primary care physicians, 40 percent general into autonomous entities. The Center and specialists, and 20 percent consulting provincial governments could help by making specialists. And many Canadian analysts funds available for salary supplements to believe that the proportion of consulting attract specialists to unpopular locations. specialists should be reduced even further. Efforts to slow production of new specialists Of course, there are important differences would be strengthened by moves to increase between Canada and Indonesia. Canada has a tuition payments for those being trained in much older population with a larger share of specialty fields. Scholarships can be linked to costly chronic and degenerative diseases. It a readiness to serve, for lengthy periods, in also spends much more on health care than provincial hospitals. Indonesia. These are all factors that would result in a higher demand for specialist 2.80 This leaves the issue of balance physicians and consulting specialists than in between GPs and specialists, and among the Indonesia. Nonetheless, there is a lesson from different specialties. Specialists, especially how much attention Canadian policymakers consulting specialists, typically provide more give to the question of manpower balance, and costly care than primary care nurses and GPs. the fact that consulting specialists are limited A high density of specialists has been shown to half the number of general specialists as a in the United States and Europe to increase the measure of containing costs and ensuring good frequency of unnecessary and often risky quality care. The United Kingdom approaches procedures, resulting in higher health care this question through a central government costs and reduced quality of care (World body. In the Netherlands, payers limit the Bank, 1993). This phenomenon of "supplier- number of specialists eligible for induced demand" is usually abetted by reimbursement (Schroeder, 1992). widespread insurance coverage and by the willingness of patients to allow doctors to act, 2.82 Finally, specialist associations play a in effect, as their agents. Moreover, when strong role in Indonesia in assuring the quality specialists operate as general physicians, they of specialist training programs, increasingly by 38 organizing national examination boards. CHS countries. Standards could be set for hospitals also relies heavily on the relevant specialist to be accredited as specialist teaching associations to review proposals for new hospitals, which could then offer places to training centers. The associations have been GPs. Program graduates would have to take conservative as regards opening new training the same national board examination as centers because of the standards they want to university trained individuals. maintain in terms of staff per student ratios, and the case-mix and the numbers of patients seen by specialists-in-training (only the largest hospitals provide a case-mix sufficient to train specialists). IDI has made the useful suggestion of organizing specialist training in hospitals, a model that is used in many Endnotes 1. The 1,368 public students were each assumed to have cost the government the Rp. equivalent of US$13,000; the 530 private students were each assumed to have cost the Rp. equivalent of US$2000. 2. Multi-stage random sampling was used to select an initial group of 4000 from IDI's roster of 26,447 members. One thousand questionnaires were sent to doctors in Jakarta and Surabaya, who account for a quarter of IDI's members, using an updated address list. The remaining 3000 were chosen from 64 IDI branches which had first been selected using a random sampling technique. Twelve percent (483) of the questionnaires were returned unopened, because of erroneous addresses. These questionnaires were replaced using the adjacent name on IDI's master list. In all, 955 questionnaires were returned and processed. 3. MOH has also occasionally assigned public sector doctors to work in private hospitals while continuing to receive civil service salaries. This practice is being curtailed. In 1994, MOH decided that by age 50 such doctors must either retire from the civil service or return to work in public facilities. 4. Doctors seem to prefer to be based in areas with better income earning prospects and educational facilities, and in health centers which are well supplied with staff and equipment. To illustrate, a regression was specified using kabupaten data for eight provinces (see para. 2.50). The average number of doctors per puskesmas (D) was related to the number of people in the coverage area (P), the average number of non-physician staff working in each facility (TS), an indicator (M) of the availability of equipment and drugs and various locational characteristics, i.e., the proportion of health centers in remote (R) or transnigration (T) areas: D = 0.485 + .003 P + 0.48 TS + .129 M - .004 R - .01 T (3.68) (1.42) (4.40) (1.55) (-2.22) (-2.28) [t - statistics in parentheses; R-squared = 0.41; number of observations = 94] Results, which may be distorted by area heterogeneity bias and by a confounding of supply-side and demand side effects, suggest that doctors are "drawn" to facilities which are more attractive in terms of work environment and opportunities for private practice. 39 5. An earlier analysis argued that bed occupancy ratios were low in class D hospitals because of a lack of specialists (World Bank, 1989). Low ratios in class C hospitals, nearly all well endowed with specialists, suggest that the causes of low facility utilization lie with management problems and perceptions about service quality rather than in the availability of personnel. 6. The estimates may be biased since they were derived, using ordinary least squares, from a single cross-section. One possible distortion, called area heterogeneity bias, can arise with such a data set because of the non-randomness of program design and placement. Programs typically respond to locational characteristics, but these attributes usually influence program outcomes as well. This bias can be corrected by taking account of the endogenous placement of programs. But the required techniques depend on large data sets covering several years. For more details, see Pitt, Rosenzweig and Gibbons, 1993. 7. In 1986, MOH introduced a manual and revised some procedures relating to "microplanning" of family health services in the health center. Research is needed on the impacts on utilization and health outcomes of this methodology in localities where it was applied faithfully. The approach does not appear to meet the preconditions specified in para. 2.53. 8. Few clues are revealed in the regression results reported in Table 2.13. Take for instance, the availability of medical instruments-a ten percent increase in access to such inputs results in only a 3.8 percent gain in visits per day. Facility-level differences in the endowment of staff as well as instruments and other resources do not seem to be the key consideration. 9. There is now considerable experience in different countries with methods of characterizing and evaluating program-community and program-client interactions. Relevant techniques include patient reports, exit interviews, focus group discussions, role playing and participant observation. For further details, use R. Simmons and C. Elias, "The Study of Client-Provider Interactions: A Review of Methodological Issues," the Population Council Programs Division, Working Paper No. 7, 1993, and M. Wensing, R. Grol and A. Smits, "Quality Judgements by 40 CHAPTER THREE NURSES AND MIDWIVES Indonesia. A second has been to build an 3.1 Nurses and midwives comprise the effective and accessible community and family largest health worker category in Indonesia, health program that operates through health accounting for roughly 44 percent of health staff centers and subcenters and outreach efforts. A in government hospitals, 35 percent in health third goal has been to achieve a sharp drop in centers and subcenters and 39 percent in private reproductive risks. hospitals. There were an estimated 113,000 nurses and midwives working in hospitals and 3.4 This chapter surveys training and government health centers in December 1992. employment intiatives and outcomes in respect Like doctors, many nurses and midwives also of nurses and midwives in Indonesia. The work privately on a part-time or full-time basis. chapter concludes with a discussion of policy issues and options relating to this important 3.2 As with doctors, GOI has used various segment of the health work force. training and employment-related initiatives to broaden access to nursing and midwifery Nursing Education services. But despite their large numbers, MOH has not pursued a single, consistent approach to 3.5 During the last 15 years, government training and deploying nurses and midwives. policy has brought about a significant expansion Unlike the policy towards doctors, government in nurse training capacity. The current annual thinking on nurses and midwives has gone outflow of new nurses is nearly 10,000, up from through several phases and changes in direction. an estimated 6771 in 1984 and 8436 in 1989 This may be due to the wide range of tasks and (Table 3.1). services which are performed explicitly or implicitly by nurses in Indonesia. These 3.6 PrimaryHealthNurses. Approximately functions include: direct patient care or service 75 percent of Indonesia's nurses and midwives delivery under the supervision or instruction of graduated from a Sekolah Perawat Kesahatan a doctor or other workers; direct care or service (SPK) which they entered following completion delivery involving independent judgement and of junior secondary school. This three year action with little or no supervision; teaching in basic nurse (PK) program was initiated in 1975 pre-service or in-service training programs; as the successor to diverse training activities that various secretarial, administrative and then produced over 20 different types of nurses. supervisory tasks, and some non-nursing work In the early 1980s, GOI enunciated a policy of including laboratory and pharmacy-related gradually phasing out SPKs and requiring a activities. secondary school for entry into a nursing program. This policy has not been pursued. 3.3 Policy shifts have also reflected the Instead, new public SPKs were opened. In pursuit of several distinctive and not always 1993, there were 118 publicly run SPKs (not compatible goals. One objective has been to including military schools) producing an establish fully and to elevate the status and estimated 4,366 primary health nurses a year, up qualifications of the nursing profession in from the 3219 PKs emerging from 87 41 Table 3.1: NUMBER OF NURSING SCHOOLS AND ESTIMATED OUTPUT, BY TYPE AND OwNERSHP, 1984-1993 Estimated Annual Output* Type 1984 1989 1993 1985 1989 1993 SPK Government 107 117 129 3,959 4329 4773 Private 59 80 54 2,183 2,960 1,998 Subtotal 166 197 183 6,142 7,289 6,771 Akademi Government 8 18 30 296 666 1,110 Private 9 13 48 333 481 1,776 Subtotal 17 31 76 629 1,147 2,886 Si Govemnment 0 0 1 0 0 37 Total 183 228 260 f 6,771 8,436 9,694 Notes: * Based on average class size of 40 students; attrition rate of 7.5 percent. a. Second Health (Manpower Development) Project., Staff Appraisal Report, (Report No. 5442-IND. Annex 1, Table 3). b. Roemer, 1990, Table 1, p. 6. c. Pusdiknakes, Rekapitulasi Peserta Didik Program Pendidikan, TA 1993/1994. September 1993. government schools in 1984 (Table 3.1). This made to equalize access to nursing education increase in PKs graduating from public schools between regions (Table 3.2). Training has been offset to some extent by a decline in guidelines for SPKs specify an average class size privately enrolled basic nurses. There were 54 of 40 students, which means that the system is private SPKs in 1993, producing an estimated currently capable of producing on the order of 1,998 primary health nurses, while in 1984, 7,320 PKs per year. However, as noted below, there were 59 private schools with an estimated recent changes in nursing and midwifery policy annual output of 2,183 PKs. Overall production call for stretching this capacity in most settings of PKs (6771) in 1993 was ten percent larger as a part of the effort to accelerate the than the 1984 total. production of village midwives (bidan di desa). 3.7 SPKs have been established in most 3.8 Entry examinations for public and private provinces, with considerable effort having been SPKs are given separately and organized on a 42 Table 3.2: Number and Output of Various Class Sizes of SPK, by Province, 1993/94 Public SPK Private SPK All SPK Enrollment Enrollment Total Enrollment Capacity planned for Capacty planned for Capacity planned for Total Province No. at 40/class BDD Program No. at 40/class BDD Program No. at 40/class BDD Program /a Planned DKTJakarta 5 200 240 4 160 160 10 400 880 1,320 Jawa Barat 14 560 1,000 6 240 260 20 800 1,260 1,260 Jawa Tengah 11 440 720 9 360 360 20 800 1,080 1,080 Yogyakarta 1 40 80 3 120 120 4 160 400 600 JawaTimur 19 760 880 5 200 240 25 1.000 1,160 1,160 Bali 2 80 120 2 80 160 200 Subtotal 52 2,080 3,040 27 1,080 1,140 81 3,240 4,940 4,940 DTAceh 6 240 480 2 80 160 9 360 720 720 Sumatera Utara 9 360 720 8 320 520 20 800 1,400 1,400 Sumatera Barat 3 120 320 3 120 120 7 280 480 480 Riau 2 80 120 0 2 80 120 120 Jambi 3 120 120 1 40 40 4 160 160 160 SumateraSebatan 5 200 280 3 120 160 9 360 520 668 Bengkulu 4 160 160 4 160 160 160 Lampung 4 160 280 1 40 40 5 200 320 320 Subtotal 38 1,440 2,480 18 720 1,040 60 2.400 3.880 3,880 Kalimantan Barat 2 80 0 2 80 0 0 Kalimantan Tengah 1 40 45 1 40 40 2 80 85 85 Kalimantan Selatan 3 120 80 1 40 40 5 200 160 160 Kalimantan Timur 2 80 120 0 2 80 120 120 Subtotal 8 320 245 2 80 80 11 440 365 365 0 Sulawesi Utara 2 80 120 2 80 120 4 160 240 240 Sulawesi Tengah 2 80 2 80 0 Sulawesi Selatan 4 160 200 4 160 200 9 360 480 626 Sulawesi Tenggara 1 40 80 1 40 40 2 80 120 120 Subtotal 9 360 400 7 280 360 17 680 840 840 0 NTB 2 80 160 0 0 2 80 160 160 NTT 3 120 120 0 0 3 120 120 120 Maluku 2 80 160 0 0 3 120 240 240 Irian Jaya 5 200 240 0 0 5 200 240 240 Timor Timur 1 40 40 0 0 1 40 40 40 Subtotal 13 520 720 0 0 14 560 800 800 Total 118 4,720 6,885 54 2,160 2,620 183 7,320 10,825 10,825 Source: Rekapitulasi Peserta Didik PPB Tahun 1993/1994 Notes: a/ Includes estimated additions from provinces regional basis. Privately run schools charge per instructor, assuming three classes of 40 PKs higher tuition than public schools, but some and one class of 40 bidans per school; the ratio government SPKs levy high non-tuition fees, can be as low as five students per teacher when e.g., for uniforms, orientation, and equipment. part time staff are included (Table 3.3). These The SPK curriculum covers basic nursing and ratios are lower in Java and Bali than in the limited maternal and child health skills, and Outer Islands, where student/teacher ratios are offers some exposure to the ecological and social usually more than 15. This indicator appears to sciences. The most recent revision of the be somewhat less favorable among private SPKs. curriculum was in 1985. 3.12 About 50 percent of the faculty in SPKs 3.9 Nursing Academies. A second large are nurses with D3 degrees (Table 3.4). Forty group of nurses, a fifth of the national total, has percent have the equivalent of a DI (certificate) earned a D3 degree after graduating from a three degree. Most SPKs rely on part time faculty, year post-secondary school program which they typically clinical staff of the hospital with which attended following the completion of senior the school is associated, for the provision of secondary school. The academy-based program clinical and higher level trainers. About 55 was initiated in 1965. Most academies were percent of all part time trainers have attained at established in the late 1980s when 16 least the SI degree. It is likely that the bulk of government and 39 private schools were opened. these are medical staff from participating In the 1984-1993 interval, the outflow of D3 hospitals. Not surprisingly, this has meant that nurses from government schools more than teachers in schools in the Outer Islands have less tripled and that from private schools grew more formal training than those working in Java and than five fold. The current annual flow of Bali. academy graduates (from publicly run schools) is over 1100, which is nearly twice the figure 3.13 Student per teacher ratios in the Nursing for 1989 (Table 3.1). About ten percent of these Academies are typically above the desired range graduates have been prepared to become of five to ten pupils per instructor. The ratios midwife trainers. D3s now account for 30 are lower in publicly owned schools in Java, percent of all nursing graduates. Bali and several other provinces (Table 3.5). These ratios are considerably less favorable in 3.10 Most nursing academies are in large the private academies in many provinces. cities and regional capitals. Apart from high school graduates, nursing academies also accept 3.14 The Bachelor of Science in Nursing. some SPK graduates who have work experience. Some 200 nurses have graduated from the There are separate Indonesia-wide examinations University of Indonesia's four year Bachelor of for public and private schools which are given Nursing (S1) program. Admission to the on different days. Students may take both university program is based on a national examinations and apply to government and examination. Apart from high school graduates, private schools. Only 25 to 35 percent of this program, which admits about 40 students a applicants are accepted-competition is generally year, also takes some D3 graduates who have less rigorous at the private school level although nursing experience. In addition to the more prestigious non-government schools are University of Indonesia, four other public highly selective. universities are in the process of establishing S1 programs. 3.11 Teaching Inputs. International experience suggests that student-teacher ratios 3.15 Midwifery Training. Midwives are no should not exceed ten, but ideally should be no longer trained separately in Indonesia as was the more than five or six. In Indonesia, these ratios case prior to the decision to consolidate the in the SPKs average about 16 full time students many different training activities into the PK program. Basic midwifery is included in the 44 Table 3.3: Estimated Number of Students per SPK Teacher, by Ownership and Province, 1993/1994 Province Public SPKs Private SPKs All SPKs Student Student per Student per Student Student per Student per Student Student per Student per per FTT EFT EFT per FTT EFT EFT per FTT EFT EFT PT=25% PT=50% PT=25% PT=50% PT=25% PT=50% DKT Jakarta 9 7 5 16 8 6 14 9 7 Jawa Barat 10 7 6 15 8 6 12 9 7 Jawa Tengah 11 8 6 9 13 10 10 11 8 Yogyakarta 15 8 6 14 8 6 14 8 6 Jawa Timur 14 10 7 15 9 7 13 9 7 Bali 11 9 7 12 9 7 11 11 9 Average 12 8 6 14 9 7 12 9 7 DTAceh 15 10 8 14 8 7 15 8 6 Sumatera Utara 14 11 9 9 13 10 11 11 9 Sumatera Barat 13 12 10 25 6 5 18 8 7 Riau 14 11 9 14 9 8 Jambi 8 7 7 8 18 16 SumateraSelatbn 16 9 6 7 15 11 10 10 8 Bengkulu 20 15 12 20 7 6 Lampung 11 9 7 27 5 5 14 9 8 Average 14 10 9 10 6 5 14 10 8 Kalimantan Barat 40 16 10 28 5 5 32 4 4 .6h. Kalimantan Tengah tJ> Kalimantan Selatan 15 11 9 46 3 3 20 7 6 KalimantanTimur 15 10 8 23 6 6 18 7 6 Average 10 7 6 23 3 3 12 5 4 Sulawesi Utara 34 28 24 11 11 8 19 8 7 Sulawesi Tengah 17 10 7 20 7 6 18 7 5 Sulawesi Seatan 15 10 8 28 5 5 20 6 5 Sulawesi Tenggara Average 22 16 13 20 7 6 19 7 6 NTr 15 10 7 24 6 5 18 7 6 Maluku 30 21 17 27 6 5 Irian Jaya 26 18 14 26 6 5 TimorTimur 21 19 18 21 7 7 Average 23 17 14 23 6 6 Overall Average 16 12 9 17 6 5 16 8 6 a. Sample includes miwtary schools, for which data is not shown Notes: FTT = Full rfim Trairer PTT = Part Time Trainer, EFT = Equivaent Full Time Source: Minitry of Heatth Table 3.4: Educational Attainment for Full Time Teachers in SPKs, 1992/93 All SPKs Ave. No. FTT Ave. FTT wl Ave. FTT wl Total FTT Province w/ D3 Percent DI* Percent Si, S2 or S3 Percent DTAceh 4.60 47% 4.20 43% 0.90 9% 9.70 Sumatera Barat 3.63 45% 4.25 52% 0.25 3% 8.13 Sumatera Utara 5.42 45% 5.58 47% 0.96 8% 11.96 Riau 6.00 60% 4.00 40% 0.00 0% 10.00 Jambi 11.50 61% 1.50 8% 6.00 32% 19.00 Sumatera Selatan 6.50 50% 4.13 31% 2.50 19% 13.13 Bengkulu 6.29 62% 2.71 27% 1.14 11% 10.14 Lampung 6.80 63% 3.40 31% 0.60 6% 10.80 DKT Jakarta 5.61 59% 2.91 30% 1.04 11% 9.57 Jawa Barat 7.25 62% 3.61 31% 0.89 8% 11.75 Jawa Tengah 6.38 40% 7.00 44% 2.52 16% 15.90 Yogyakarta 6.80 65% 3.40 33% 0.20 2% 10.40 Jawa Timur 4.70 43% 5.03 46% 1.09 10% 10.82 Bali 5.40 44% 6.40 52% 0.60 5% 12.40 Kalimantan Barat 4.00 46% 4.43 51% 0.29 3% 8.71 Kalimantan Tengah Kalimantan Selatan 3.17 42% 3.67 49% 0.67 9% 7.50 Kalimantan Timur 5.25 60% 3.50 40% 0.00 0% 8.75 Sulawesi Utara 4.60 34% 9.00 66% 0.00 0% 13.60 Sulawesi Tengah 4.43 53% 4.00 47% 0.00 0% 8.43 Sulawesi Selatan 4.73 68% 2.09 30% 0.18 3% 7.00 NTT 1.50 19% 6.38 80% 0.13 2% 8.00 Maluku 3.50 58% 2.25 38% 0.25 4% 6.00 Irian Jaya 3.50 58% 2.50 42% 0.00 0% 6.00 Timor Timur 5.00 67% 2.00 27% 0.50 7% 7.50 Overall Average 5.06 50% 3.92 39% 0.83 7% 9.81 Source: Ministry of Health. Notes: FTT = Full Time Trainer; DI*= SPK degree + one additional year -- the equivalent of a Dl degree; Si, S2 and S3 are undergraduate, masters and doctoral degrees. 46 Table 3.5: Student/Teacher Ratios in D3 Academies, by Province and Ownership, 1992/93 Public Private All Academies Studentsl EFT EFT. Stud6nts/ EFT EFT Students/ EFT EFT Province FT PT=25% PT=50% FT PT=25% PT=50% FT PT=25% PT=50% DTAceh 48 46 44 96 91 87 Sumatera Barat 9 6 4 40 14 8 15 8 6 Riau Jambi 8 5 4 8 5 4 Sumatera Selatan 6 4 3 20 10 7 9 6 4 Lampung 6 5 4 60 30 11 9 7 Average 15 13 12 30 28 15 28 24 22 DKJakarta 6 4 3 10 7 6 9 7 5 Jawa Barat 5 4 3 19 14 11 10 8 6 Jawa Tengah 7 5 4 8 6 5 8 6 5 Yogyakarta 8 3 2 16 10 7 12 6 4 Jawa Tlnur 8 5 4 8 5 4 8 5 4 Bali 9 5 3 _ 9 5 3 Average 7 4 3 12 8 6 9 6 4 Kalimantan Barat 60 21 13 60 21 13 Kalimantan Selatan 19 10 19 10 Kailmantan Timur Average Sulawesi Utara 9 5 4 9 5 4 SumateraUtara 7 5 4 13 11 9 11 8 7 Sulawesi Selatan 8 6 4 8 6 4 Average 8 5 4 9 6 5 NTB NTT Maluku 5 3 2 5 3 2 Irian Jaya 20 17 14 20 17 14 Average 13 10 8 13 10 8 OverallAverage 11 8 7 21 18 11 15 12 10 Source: Ministry of Health Notes: FT= Full time trainer, PT= Part time trainer, EFT= Equivalent full time Table 3.6: Number of Bidan di Desa Graduates, 1989/90-1992/1993 No. villages Bidan di No. BDD Total Planned Enrollment Net +1- Net /- Net N - without a Desa needed to No. Total from from from Health Center trained achieve SPK Capacity Capacity Capacity Capacity Province d9/90-92/93 1BDDhvilLage at 40/dass 1993/1994 1994/1995 1995/1996 1993/1994 1994/1995 1995/1996 DKTJakarta 0 21 -21 10 400 880 (1119) (1052) 480 Jawa Barat 5,469 2,649 2,820 20 800 1,260 1,271 1,145 460 471 345 Jawa Tengah 6,736 2,707 4,029 20 800 1,080 989 1.052 280 189 252 Yogyakarta 161 177 -16 4 160 400 210 293 24G 50 133 Jawa Timur 6,314 2,560 3,754 25 1,000 1,160 1,228 1,076 160 228 76 Bali 413 342 71 2 80 160 194 163 80 114 83 Subtotal 19,093 8,456 10,637 81 3,240 4,260 3,892 3,729 1,020 652 489 DTAceh 5,220 1,065 4,155 9 360 720 386 526 360 26 166 Sumatera Utara 4,836 2,365 2,471 20 800 1,400 1,312 1,486 600 512 686 Sumatera Barat 2,882 799 2,083 7 280 480 537 525 203 257 245 Riau 870 290 580 2 80 120 125 122 40 45 42 Jambi 1,182 430 752 4 160 160 160 180 0 0 20 Sumatera Selatan 2,086 924 1,162 9 360 666 488 512 306 128 152 Bengkulu 804 305 499 4 160 160 138 233 0 -22 73 Lampung 1,491 471 1,020 5 200 320 205 179 120 5 -21 Subtotal 19,371 6,649 12,722 60 2,400 3,880 3,351 3,763 1,480 951 1,363 OD Kalimantan Barat 4,364 345 4,019 2 80 0 109 100 -80 29 20 Kalimantan Tengah 906 215 691 2 80 85 130 123 5 50 48 Kalimantan Selatan 1,994 549 1,445 5 200 160 202 212 -40 2 12 Kalimantan Timur 742 260 482 2 80 120 103 112 40 23 32 Subtotal 8,006 1,369 6,637 11 440 365 544 552 -75 104 112 Sutawesi Utara 936 406 530 4 160 240 240 235 80 80 75 Sulawesi Tengah 1,153 312 841 2 80 0 0 173 -80 -80 93 Sulawesi Selatan 705 784 -79 9 360 626 451 360 266 91 0 Sulawesi Tenggara 597 178 419 2 80 120 93 95 40 13 15 Subtotal 3,391 1,680 1,711 17 680 840 784 863 160 104 183 NTB 378 314 64 2 80 160 197 131 80 117 51 NTT 1,431 389 1,042 3 120 120 251 257 0 131 137 Maluku 1,495 414 1,081 3 120 240 218 261 120 98 141 Irian Jaya 652 271 381 5 200 240 326 291 40 126 91 Timor Timur 303 160 143 1 40 40 147 80 0 107 40 Subtotal 4,259 1,548 2,711 14 560 800 1,139 1,020 240 579 460 Total 54,120 19,702 34,418 183 7,320 10,145 9,710 9,927 2,825 2,390 2,607 Source: Kegiatan Pendidikan Bidan Tahun 1993/1994, Dilingkungan Pusat Pendidikan Tenaga Kesehatan, DepKes, October, 1993 Table 3.7: Distribution of Nurses and Midwives (NMW) by Type, Facility, Ownership and Region, 1992 All Categories Public Hospits Private Hospitls Healfth Centers AUl FacilWes Pop per Region No. % all NMW No. % all NMW No. % all NMW No. % all NMW All NMWH 1 24,197 21% 17,054 15% 18,334 16% 59,585 53% 1,870 0 8.734 8% 5,065 4% 12,453 11% 26,252 23% 1,570 m 2,598 2% 881 1% 4.078 4% 7,557 7% 1,598 IV 4,434 4% 1,372 1% 5,688 5% 11,494 10% 1,091 V 3,173 3% 425 0% 4,637 4% 8,235 7% 1,572 Total NMWs 43,138 38% 24,797 22% 45,190 40% 113,123 100% 1,682 Basic Nurse (PK) Public Hospitals Prate Hospis Heafth Centei AlI Facilities Pop per Region No. % afl PK No. % all PK No. % all PK No. % all PK PK 1 19,495 23% 13,363 18% 11,296 13% 44,154 52% 2,524 0 6,602 8% 3,954 5% 9,069 11% 19,625 23% 2,100 m 2,038 2% 74e 1% 2,888 3% 5,672 7% Z129 IV 3,478 4% 1,009 1% 4,110 5% 8,597 10% 1,459 V 2,598 3% 361 0% 3,809 4% 6,768 8% 1,912 Total PK 34,211 40% 19,433 23% 31,172 37% 84,816 100% 2,243 lidwives (Bidan) Public Hospitals Prvate Hospitals Health Cenbn All Fadlities Pop per Region No. %ai SkBdn No. %al Bkdan No. %al lBSin No. %all Bldan BSdan I 2,882 13% 1,774 8% 7,011 31% 11,687 52% 9,552 n 1,523 7% 595 3% 3,347 15% 5,485 24% 7,543 in 412 2% 84 0% 1,145 5% 1,641 7% 7,358 IV 805 3% 264 1% 1,572 7% 2,441 11% 5,137 V 419 2% 40 0% 756 3% 1,215 5% 10,652 Total MW 5,841 26% 2,757 12% 13,831 82% 22,429 100% 8,481 D3 Nurses (Akpor) Public Hospials Prvaeb Hospitals Health Centers All FacilIte Pop per Region No. %aol 3 No. %all D3 No. %allD3 No. %all D3 D3 1 1,820 31% 1,917 33% 27 0% 3,764 64% 29,608 0 609 10% 515 9% 37 1% 1,162 20% 35,473 m 148 3% 51 1% 45 1% 244 4% 49,486 IV 351 6% 99 2% 6 0% 458 8% 27,501 V 156 3% 24 0% 72 1% 252 4% 51,359 Total D3 3,084 52% 2,807 44% 187 3% 5,878 100% 32,361 SourCe: Ministry of Health. 49 Table 3.8: Country-wide Distribution of Nursing Personnel by Type and Level of Facility, 1992 Nunce (PK) Percent MUidife (BSan) Percent D3 Percent All NHW Percent Ave AR NMW per Type of Faciliy No. of al PK No. all Bidan No. all D3 No. all NMW No. Beds BOR Occupied Bed A 2,813 289 1.16 252 4.57 3,147 2.82 3,450 B 8,942 10.77 1,598 6.90 983 17.82 11,523 10.32 14,825 64.24 C 9,433 11.36 1,784 7.70 741 13.43 11.958 10.71 20,153 58.00 D 5,129 6.18 1,048 4.53 406 7.36 6,583 5.89 10,501 43.90 Toba 28.117 31.46 4,899 20.29 2,382 43.18 33,198 29.72 48.729 41.54 1.64 Othe Public 8.287 7.57 1,135 4.90 341 6.18 7,783 6.95 11,175 38.79 1.79 ANl Pulic Hoeplils 32,404 39.04 5,834 25.19 2,723 49.36 40,961 38.88 59,904 40.16 1.70 Pridvat Hoepasb 19,433 23.41 2.757 11.90 2,807 47.25 24,797 22.20 46,223 47.25 1.14 H"eth Cenbts 31,172 37.55 14,569 82.91 187 3.39 45,928 41.12 OveralTota 83,009 100.00 23,160 100.00 5,517 100.00 111,686 100.00 106,127 43.70 1.42 Source: Miniby of Health. Table 3.9 Norms for Number of Nures by Type of Facility, 1992 . Hospitals Category of Nurse Type A Hospital Tyve B Hospital Type C Hospital Type D Hospital D3 Nurse (Akper) 135 50 9 3 Basic Nurse (SPK) 1420 468 80 20 Bidan 43 28 5 5 Implied Nurse/Doctor Rati 20tol 13tol 5tol 7tol Health Centers Level I (Urban) Level I (Rural) Level II (Urban) Level II (Rural) evel IlIl (Urban) Level IlIl (Rural) 30,000 Doctor 3 2 2 1 1 1 +1 Basic Nurse 6 8 6 7 6 6 +1 Bidan 3 5 2 5 2 5 +1 Total Nurse and Bidan 9 13 8 12 8 11 +1 Source: Ministry of Heafhi. SPK curriculum. For example, SPK students through programs to be implemented at the SPK are expected to assist in five normal deliveries level. Overall, the acceleration program calls during their three year course. In addition, for a 25 percent increase in total production some SPKs initially offered an additional one capacity at the SPK level. Plans include year course in midwifery for nurses with at least provision for additional equipment and teaching two years of experience. This post-SPK materials at the school level, as well as the program was terminated in 1980, but revived in allocation of one minibus per school to make it 1989 with the development of the bidan di desa easier to acquire field experience. (village midwife) training program. This Student/faculty ratios at participating schools program aimed at creating 18,900 bidan di desa will be doubled for at least 18 months while new by the end of Repelita V. These workers would D3 level faculty are trained. be placed at the village level for the provision of maternal and child health and family planning Employment Patterns services. To accomplish this goal, the requirement for at least two years of experience 3.17 With more than 113,000 nurses and as a PK was dropped, and approximately 90 midwives working in hospitals and health centers percent of all PK graduates were admitted to the in Indonesia, the population per nurse/midwife additional year of midwifery training, directly ratio is about 1,682. This figure is about the from their PK level programs (see Table 3.2). same as the average ratio in all middle income In addition, the Center for Education and countries (1,640), but is more than twice as high Training of Health Personnel (Puskdiklat) as that found in Thailand (550), and Malaysia provided in-service training for 180 bidan di (380), and about one half the regional average in desa in three centers during 1991/92. The the East Asia and Pacific region (2720) [World program expanded to cover a total of 10 training Bank, 1993]. What is striking, though, is how centers and 300 midwives in 1992/93 and 26 quickly the supply of nurses and midwives has training centers and about 780 midwives in grown and continues to grow. There were 1993/94. 45,000 nursing staff working in the government system in 1983/84. By 1993, the nurse work 3.16 Acceleration Program. A January 1993 force had increased by 250 percent, a ten Ministerial Decree (Ministry of Health Decree percent annual rate. Growth continues at No. 69/MENKES/SK/I/1993) mandated a roughly this pace. Where have these health stepped up training scheme for village midwives. workers found jobs? The initiative aims at placing one bidan di desa in each of the approximately 67,000 villages in 3.18 Distribution by Type of Facility. Indonesia. These midwives are to be the Information is available on the three categories primary sources of both basic health care and of nurses (basic nurses or PKs, and academy maternal and child health at the village setting. nurses or D3s) and midwives working in public MOH has calculated the total number of bidan di and private hospitals and in government health desa that need to be trained to achieve the centers (Tables 3.7), and on recent trends in Acceleration Program's goals (Table 3.6 and nurse staffing in class A, B, C and D hospitals Annex Table 3E). These estimates, which target (Table 3.8). Unfortunately, there are no robust villages which do not have health centers, call estimates of the total number, or distribution of for the training of nearly 10,000 new bidan di basic, D3 or midwives employed in the private desa each year to the end of 1996. To achieve sector in non-hospital settings, such as this goal, class sizes will be doubled in the polyclinics, private medical practices or in majority of schools, and students from selected employer managed facilities: provinces will be assigned to training programs in provinces with relatively greater school * Basic nurses (PKs) constitute about 75 capacity. In addition, plans call for the training percent of the total number of nurses and of 5,846 D3 graduates as bidan teachers, midwives, midwives make up about 20 percent, 52 while the academy-level nurses (D3s) make up percent are in type C and D hospitals, with only just 5 percent of the nursing work force in 17 percent working at the type D level (Table hospitals and health centers. D3s now comprise 3.8). more than seven percent of the nurse work force in A, B, C and D hospitals, up from three * The nursing staff in A, B, C and D percent in 1986. hospitals increased by 25 percent (6511) between 1986 and 1992 (Annex Table 3A). Growth was * About 53 percent of this work force are especially rapid in class A (42 percent) and B based in Java and Bali (less than the population (50 percent) hospitals. share of this region), about 23 percent are in Sumatra, and the remaining 25 percent are 3.19 Staffing Norms and Availability in distributed among the more remote and less Government Facilities. In hospitals, the PKs densely populated Outer Island provinces. Of undertake simple clinical nursing and patient the total, about 38 percent work in public care, bidans assist in deliveries and provide hospitals, about 40 percent in health centers, and various family planning services and Akpers are the rest in private hospital settings. The bulk expected to take on supervisory and management (68 percent) of those employed in the private tasks and to direct care of difficult cases. A hospital sector are found in Java and Bali; only class D hospital with 60 beds and limited or no about 10 percent of private hospital based specialist services is expected to have 20 PKs, employment occurs in Kalimantan, Sulawesi or three D3s and five midwives, for a total of 28 the Eastern Islands. nursing staff (Table 3.9). Staffing standards are more generous at the class C (94 nurses), B * For nurses and midwives employed in (546) and A (1598) levels, reflecting the public facilities, there is a fairly even differing functions assigned to these hospitals. distribution between hospital and health center Based on these norms, class A and B facilities level employment. Thirteen percent are found are supposed to employ 55 percent of the nurses in Type A and B hospitals, 17 percent are in working in publicly run general hospitals, with Type C and D hospitals and 41 percent work in the four class A units accounting for 17 percent health centers (Table 3.8). of all nursing staff. Class D hospitals are allocated, according to staffing norms, less than * About 77 percent of all PKs work in 13 percent of nursing personnel. public hospitals (40 percent) or health centers (37 percent) and 23 percent are employed in 3.20 Using such norms to evaluate the private hospitals. Only 12 percent of bidans availability of nurses is problematic. In work in private hospitals with the rest employed particular, standards need to be calibrated in public hospitals (26 percent) and mainly against reasonable estimates of service delivery health centers (62 percent). Both basic nurses efficiency and quality (see discussion below). and midwives are distributed in rough proportion Nevertheless, norms do provide a simple to population across the five major regions of yardstick to assess the availability of staff in the country. government facilities. The degree to which these standards were attained in government * D3 nurses are heavily concentrated in hospitals in different regions can be summarized hospital (including private hospital) employment, by juxtaposing actual employment and standard with fewer than three percent working at the patterns (Table 3.10 and Annex Table 3B). This health center level; about 65 percent of all D3 comparison shows that current nurse and nurses are found in Java and Bali and Sumatra. midwifery staffing exceeds that suggested by the Some 44 percent of all D3s are employed in norms in most provinces for class C and D private hospitals and 52 percent are employed in hospitals. Sumatra is the only region with a public hospitals. Of the latter, 51 percent are significant nursing "shortage" in class C employed in type A and B hospitals, and 49 hospitals. No region is experiencing such a 53 Table 3.10: COMPARISON OF AVERAGE NUMBER OF NURSES PER FACILITY TO GOVERNMENT STAFF NORMS, BY REGION AND HOSPITAL CLASS, 1992/1993 Region /a Class A Difference Class B Difference Class C Difference Class D Difference NMW/ from norm NMW/ from norm NMW/ from norm NMW/ from norm Hospital (1598) Hospital (546) Hospital (94) Hospital (28) Java/Bali 1,306 -292 337 -209 90 -4 45 17 Sumatra 126 -1,472 245 -201 76 -18 32 4 Kalimantan 116 -430 98 4 30 17 Sulawesi 397 -1,201 203 -343 109 15 45 93 Eastem slands /b 174 -372 114 10 38 45 TotllWeightod Av. 784 -814 268 -278 91 -3 38 10 Source: Derived from Ministry of Health data. deficit at the class D level. At the class A level, Acceleration Program presumes that health the prestigious Dr. Cipto Mangunkusomo center bidans would be the primary trainers for (Jakarta) and Dr. Soetomo (Surabaya) hospitals the Program C type of bidan di desa, the locally have close to the number of nurses specified, but recruited primary school graduates who would the two recently reclassified A level facilities in "apprentice" with health center staff. Medan and Ujung Pandang have far less than the number of PKs and so forth called for in staffing 3.22 Health Center Nurses and Midwives by standards. Finally, nursing strength in class B Area and Population Density. The availability hospitals in every province is below that of nursing and midwifery staff relative to specified in MOH's staffing norms. population density, area and, as a crude indicator of potential workload, the average 3.21 Staffing standards are also available for number of villages per available staff also needs health centers-these are differentiated by rural to be explored (Table 3.12 and Annex Table and urban areas and by level (Table 3.9). For 3D). Not surprisingly, the size of the catchment regional comparison purposes, average values of area served by a particular health center varies these norms across health centers are used, significantly across the country, with coverage regardless of type or level. These call for about areas tending to be significantly larger in the 10 nurses (either D3 or SPK, but mainly PK) Eastern Islands, and in Kalimantan. However, and three midwives for each puskesmas. This average nursing staff availability per health level of staffing intensity is rarely achieved center does not vary significantly in the same (Table 3.11 and Annex Tables 3C, 3F, 3G, direction. Thus, the potential total workload, 3H).I Actual staff placements at the health and the time and distance involved in performing center are lower than targeted levels in all nursing and midwifery functions are much regions; "shortages" are relatively higher in Java higher in less populated areas. For example, and Bali, Sumatra and Kalimantan than in the health center nurses and midwives in Kalimantan Eastern Islands. Shortfalls from targeted would be responsible for coverage areas more staffing ratios for bidans are smaller than for than 10 times larger, on average than those SPKs, amounting to fewer than one midwife per implicit in the government norms for Java and health center in most of the country. Shortages Bali. of bidan staff at the puskesmas level are particularly pronounced, however, in the Eastern 3.23 Employing bidan di desa. As Islands. Most health centers in Irian Jaya lack mentioned, MOH is using the Acceleration midwives-this is noteworthy since the proposed 54 Table 3.11: DIFERENCES BETWEEN NUMBER OF NURSES PER HEALTH CENTER (HC) AND CURRENT STAFFING NORMS, BY TYPE AND REGION, 1992 Region la Total # Total PK+D3 Difference from Total Bidan Difference from Total NMW of HCs per HC Norn (10:1) per HC Norm (3:1) per HC Java/Bali 2,902 3.9 -6.1 2.4 -0.6 6.3 Sumatra 1,311 7.0 -3.0 2.6 -.04 9.5 Kalimantan 569 5.2 -4.8 2.0 -1.0 7.2 Sulawesi 539 7.6 -2.4 2.9 -0.1 10.6 Eastern Islands /b 560 6.9 -3.1 1.6 -1.4 8.3 Total 5,881 5.3 -4.8 2.6 -0.4 7.7 Source: Derived from Ministry of Health data. /a Data presented are regional weighted averages. Province specific data are in Annex Table 3D. /b Includes NTB, NTT, Maluku, Irian Jaya and Timor Timur. Program to increase the supply of midwifery Tidak Tetap (PTT) scheme. As of November, services at the village level. The aim is to place 1993, the terms and conditions for the one trained bidan in each village in Indonesia. contracting procedures which would govern the At the end of 1993, 19,702 village midwives had scheme for bidan di desa had not been fully been trained and assigned to villages (Table developed. Discussions indicated that each 3.6). Of this number, 80 percent (15,900) bidan di desa would receive roughly Rp 250,000 actually took up positions. This crash program (US$ 118) per month as basic salary, with was designed according to the following increments of Rp 600,000 (US$ 284) per month principles: allocation policies would assign each for those in very remote areas, Rp 400,000 - bidan di desa to her own community or at least 500,000 (US$ 190-237) for those in remote to the same kecamatan or kabupaten, to assure areas and Rp 300,000 (US$ 142) for those in compatibility between the midwife and the less difficult areas. Contracts would be prepared recipient village/community; recruitment of for a three year period, with extensions for an candidates for the training programs, and additional three years likely. Specification of responsibility for the assignment of newly areas by degrees of remoteness and difficulty, trained bidans would be a provincial and the terms of reference for the contract were responsibility, collaboratively managed by the not available. Current salary scales for newly provincial health office, the provincial BKKBN graduating SPKs or bidan di desa are for rank office, the Bappeda and the concerned school; II/a, or about Rp 110,000 (US$ 52) per month. those hired would become full time civil The proposed contract would not provide for servants; newly appointed midwives would be long term civil service benefits, such as a provided with a basic bidan kit prior to their pension plan. deployment; and bidan di desa would be expected to remain at the village level for an 3.25 Private Practice. All categories of agreed (not specified) length of time, and would nurses and midwives, including traditional birth be permitted to conduct private practice. attendants (dukun bayi) conduct private practice, although, with the exception of private 3.24 These principles remain in force, with midwifery, these activities are not explicitly the major exception that beginning in 1993/94 recognized in training programs or in health midwives were to be hired, not as civil servants, manpower policy. One indication of the extent but on a contractual basis through the Pegawi of private practice is the large share of deliveries 55 Table 3.12: DISTRBUTION OF HEALTH CENTER (HC) NURSES AND MIDWIVES (NMW) BY AREA, POPULATION DENSITY AND NUMBER OF VILLAGES, BY REGION Region /a Total Area Population No. Total No. Population Total sq. Total No. Population Density Villages HC Nurses per NMW km. per Villages 000/sq. (all types) NMW per NMW km. Java/Bali 93,670,928 190,764 491 25,188 18,334 5,109 12.5 1.4 Sumatra 39,000,268 481,780 81 18,137 12,453 3,132 38.7 1.5 Kalimantan 12,664,293 613,585 21 4,346 4,078 3,106 150.5 1.1 Sulawesi 11,664,160 224,274 52 3,758 5,688 2,051 39.4 0.7 Eastern Islands /b 12,433,751 587,545 21 4,269 4,637 2,681 126.7 0.9 ToWal 169,433,400 2.097,948 81 55,698 45,190 3,749 46.4 1.2 Source: Derived from Ministry of Health data. /a Data presented are regional weighted averages. Province-specific datea are in Annex Table 3D. /b Includes NTB, NTT, Maluku, Irian Jaya and Timor Timur. assisted by dukun bayi and private midwives. In access to hospital services. Nurses assigned to 1991, dukun bayi took part in 64 percent of all hospitals apparently do not engage in extensive births. The role of these village providers was private practice. However, PK nurses employed particularly high in West and Central Java and in health centers frequently develop private in several Outer Island provinces (Annex Table practices to supplement their income. For 31). Midwives assisted in over 29 percent of instance, about 90 percent of the PKs deliveries, with the role of bidans looming interviewed in an anthropological study in especially large in Sumatra and Kalimantan. If Central Java maintained active private practices 75 percent of such assistance was provided as an (Sciortino, 1992). These practices were known unofficial, paid service, then the combined share to be illegal by both provider and consumers, of private practitioners (midwives and dukun but were nevertheless pervasive in that province bayi) in assisting births would be over 85 and elsewhere. Practices can vary from a few percent. informal contacts to more highly structured activities involving 30-40 patients per day and 3.26 The scope of private nursing and established office hours. Sciortino attributed the midwifery practice is also illustrated in the 1992 popularity of private nursing care to the Household Health Survey's finding that nurses, informality and familiarity of the setting, and the midwives and dukun bayi account for the bulk of personalized services which were rendered. privately provided services in rural settings and Many nurses had spent years in the same area are second only to physicians as sources of care and had built up a loyal and regular client base. in urban areas. For example, about 43 percent Survey data are revealing on this point. Almost of all private care in Sumatra (which accounted half of the nurses and midwives in private for 36 percent of all consultations) was provided practice in NTB had lived in the locality for by nurses. In Kalimantan, nurses were visited more than five years as compared to 11 percent by 50 percent of those consulting a private of the doctors (Rand, 1992). In addition, PKs practitioner (Table 1 .4).2 lived nearby and were available in the evenings and early morning hours which were more 3.27 Nurses and Private Practice. The role convenient for working women. Thus, patients of nurses as private providers of curative care is could avoid the rushed and impersonal context significant, particularly in areas which lack easy of the puskesmas. The physical conditions 56 where PKs received patients likely played a role instruments, contraceptive means and drugs are as well. The Kaltim-NTB survey found that the of lesser quality. Unlike nurses, midwives tend "offices" of PKs and midwives generally ranked not to be involved with the provision of above health center and subcenter facilities in medications other than those required for terms of cleanliness, ventilation and availability delivery and family planning. These activities of running water and electricity (Rand, 1992). are sufficiently renumerative to make the risks of illegally practicing other forns of medicine 3.28 Finally, the low costs charged by nurses uncomfortably high, and their private roles are were a key factor attracting patients. Estimated set out and acknowledged, though not closely per visit fees in the Central Java study were in monitored, in health law. Reports of the fees the Rp. 1,500-2,500 (US$ 0.71-1.18) range, a earned by midwives are highly variable. third or less of what doctors typically received However, the Central Java study reported that (Table 2.4). Based on these figures, it can be fees for delivery were about Rp 30,000, while estimated that monthly earnings for PKs in those for contraceptive services varied by type private practice might vary from Rp 150,000 of method from a low of Rp 3,000 for the three (assuming a 25 day working month and three month injectable (DepoProvera) to Rp 30,000 patients per day at Rp 2000 per patient) to more for the insertion of an imnplant (Sciortino, 1992). than one million rupiah (US$ 71474), for Definitive figures on fee levels are not available. practices which see 30 patients per day. However, these estimates were generally confirmed by the 1991 Demographic and Health 3.29 The medical treatments provided in Survey. As discussed in more depth below, private settings were observed to be similar to these figures suggest that private midwifery those employed by PKs within the puskesmas. practice can produce incomes which may be With the health center doctor often absent due to higher than those earned by private PK administrative duties or to attend private practitioners. practice, many PKs offered simple curative care in the puskesmas without much effective Issues and Reconmmendations supervision. The same pattern held in private practice. Nurses did not tend to base their 3.31 The nursing scene and government policy treatments on a carefully completed medical towards nurses and midwives have been in flux history or diagnosis of the presenting condition, for a number of years. Major steps and changes which would require skills beyond their level of in direction have included the 1975 training. Observation suggested that treatment reclassification of nursing categories and training emphasized use of injections and antibiotics, and arrangements, delineation of a community health that needles and other instruments were often not role for nurses as part of the primary health care sterilized. strategy embraced during Repelita III (1979-84), the 1982 decision to phase out SPKs and 3.30 Bidans and Private Practice. Midwives upgrade nursing skills to the D3 level, the have a long tradition of legally sanctioned elimination in 1980 of a separate midwifery private practice in much of Indonesia. Bidans training stream and the initiation of a are involved in assisting at childbirth, often in baccalaureate program in nursing in 1985. addition to the services of a dukun bayi, Accordingly, the 1993 Acceleration Program, providing contraceptive services, and other basic which represented an intensification of the 1989 maternal and child health services. Midwives village midwife scheme and which envisaged the tend to base their practices at home, using an training and placement of over 34,000 additional extra bedroom as a small clinic, which clients midwives, is only the latest major attempt to visit. As with nurses, bidans are sought out in reshape the field. part because of the perception that standards of care, including the attention given by providers, 3.32 Nevertheless, this stepped up training in the health centers are lower, while and deployment initiative represents a good point 57 of departure for a broad review of nursing Java, Bali and Sumatra. On the other hand, issues. This is because the Acceleration considerable surpluses relative to personnel Program will have ramifications for the standards exist in some regions in C and D allocation and use of nurses and midwives facilities. There appears to be a surplus of over throughout the health system. For instance, 1800 nurses at the class D level. over the next five years the bidan di desa scheme promises to divert over 34,000 SPK 3.35 An even clearer picture emerges if graduates from jobs in hospitals and health nursing needs are calibrated in terms of average centers. How these facilities will cope and what patient loads. For instance, among class C adjustments will be required are topics of hospitals, the numbers of outpatients and interest. Of course, these concerns need to be inpatients accounted for only 28 percent of the addressed in respect of how effectively nurses inter-facility variation in the number of PKs. and midwives are being used at the moment. The same variables accounted for 34 percent of the variation in class D hospitals. Similar 3.33 Another implication pertains to the findings applied to the number of midwives and design of training efforts. Questions relate to D3 nurses (results are reported in Annex Table how realistic is the plan to produce 34,000 3J). By contrast, the number of PKs and village midwives, what refinements are required midwives in private hospitals was closely related in the bidan di desa course, what should be done to the patient load-78 percent of the cross- with the SPK curriculum and whether the goal facility variation in staff numbers was of phasing out SPKs should be maintained, and attributable to inpatient and outpatient levels. how and according to what criteria should nurse training be regulated. Finally, the viability and 3.36 What these findings suggest is that sustainability of the village midwife scheme need staffing numbers are only weakly related to to be assessed. Concerns include how well patient demand, leaving considerable scope for communities will respond to this new category productivity gains. To illustrate, these results of worker, what the client base and fee structure were used to identify more efficient hospitals, will need to be to assure a satisfactory income defined as those for which the linkage between flow, how the bidan di desa will link up with staff and patient numbers was particularly staff based on health centers and subcenters and close.3 For instance, there were 22 class D who will regulate and/or supervise village hospitals (17 percent of the country total) which midwives. used fewer PKs than the average facility but handled a larger patient volume and achieved 3.34 Availability of PKs and Government higher bed-occupancy ratios (Annex Table 3K). Midwives in Hospitals. One implication of the Similarly, there were 12 class C facilities which bidan acceleration program is that many PKs seemed to have accomplished more with fewer will continue on to midwifery training and will PKs (Annex Table 3K). then be placed, not in a hospital or health center, but directly in villages which require their 3.37 Finally, a related way of gauging services. An immediate question is whether the efficiency and the availability of nursing staff is level and quality of hospital services will be to focus on inefficient hospitals. For example, affected if the flow of graduating nurses from there were 47 class D and five class C hospitals SPKs is sharply diminished. In this regard, in December 1992 with average bed occupancy there is some basis for concern in that there are rates of 30 percent or less. Presumably, the discrepancies in some regions and in some 1834 nurses and midwives assigned to those hospital classes between the number of PKs and facilities were underutilized to a considerable midwives available and established staffing extent. standards (Table 3.10). As discussed, there are "shortages" relative to norms at the class A and 3.38 In short, there are indications that there B facility levels, and in class C hospitals in may be an excessive number of nurses and 58 midwives in government hospitals. Two leaving less preferred public health, preventive inferences can be drawn. First, there is some and educational functions to lower status scope to "divert" PK graduates from hospital paramedical workers. Several factors had assignments to village service on bidan di desa allowed this unofficial task and role allocation to contracts. Second, hospital administrators need emerge. Doctors devoted themselves to to be encouraged to use nursing staff more administrative matters and their practices, while efficiently. Here, a possible approach is to look technical direction, monitoring and supervision for lessons in those hospitals which are handling from outside the puskesmas were ineffective (see above-average volumes of work per available para. 1.30). Nurses were more comfortable nurse. The Lembaga Swadana process may with technical-curative work and at best prove to be a useful mechanism in this regard. indifferent to outreach and community-based activities. In any event, the outcome has been a 3.39 Can PKs and Bidans be Freed up at particular pattern of nursing services the Puskesmas Level? As discussed in para. characterized by a concentration on curative 3.20, there appear to be too few health center services that are provided at unacceptable levels nurses and midwives when actual numbers are of quality. For instance, health center nurses compared with official staffing norns (Table were found to spend too little time and to limit 3.11). Nevertheless, there are no strong their interactions with patients, to conduct only indications that this "shortage" interferes with superficial and rudimentary examinations, and to service levels and quality. On the contrary, as produce inappropriate diagnoses, treatments and with hospitals there appears to be considerable prescriptions, relying excessively on drugs and misallocation and underutilization of nursing injections.5 staff within health centers. For instance, the number of nurses and midwives had little or no 3.41 This discussion then points to possible impact as explanatory variables in regression- "slack" in health centers which should make it based efforts to understand observed average possible for the Accelerated Program for village daily puskesmas visit levels in a sample of midwives to proceed without adverse impacts on kabupatens (see also para 2.46). On a univariate puskesmas services. The discussion also draws basis, the number of PKs and bidans accounted attention to the desirability of rethinking health for 12 percent of the variation in average daily center activities, in particular the appropriate visits. This contribution disappeared entirely role of PKs and bidans. Chapter Two suggested when other variables, i.e., the average number that puskesmas-based health efforts be focused of doctors per health center, and indicators for on the 750-1500 high IMR localities, with the availability of instruments and population experienced or promising facility administrators density in the vicinity were included in the given the responsibility, subject to local and regression.4 kabupaten review and assessment, of developing an effective mix of staff roles and skills. The 3.40 These findings can be buttressed and required number of PKs and bidans and actual interpreted through reference to anthropological task assignments could vary from facility to investigations of activities in a typical facility, depending on local epidemiological puskesmas. One such account, previously cited characteristics and community expectations. A in Box 1.1, offers a picture of staff misallocation larger public health and outreach role for PKs and inefficiency in several facilities in Central and bidans could be anticipated. However, their Java, and may help explain why utilization levels curative inclinations and their leadership abilities are not closely linked to the number of personnel and local credibility should be harnessed rather available (Sciortino, 1992). The principal than denied. Under appropriate supervision and finding was that an informal division of labor with additional training, nurses could be had developed enabling nurses and bidans to authorized to handle the bulk of curative confine their attention to curative care in the contacts, while taking on broader formal health center or occasionally in mobile clinics, 59 responsibilities and becoming accountable for 3.44 What is clear though is that there may be specific preventive health and community a discrepancy between what government outreach activities, including helping village agencies expect village midwives to accomplish midwives get established. This would allow and the prerequisites for establishing a facility administrators to concentrate on overall sustainable livelihood as a village-level worker. program duration and management. Able health It is anticipated by MOH and BKKBN that the center nurses should be given the opportunity to bidan di desa will spend considerable time become puskesmas managers. educating and persuading women as to the benefits of family planning and practices 3.42 A similar flexible approach could shape conducive to the health of mothers and children. hiring decisions and task assignments in respect In addition, this field worker will be expected to of nurses and midwives in health centers in train and support dukun bayi and kaders (village somewhat more prosperous areas. Actual roles health volunteers), and to assist in various data would need to be contingent on local needs and collection exercises. The village midwife is also conditions. A core set of public health and expected to deliver some babies, and to provide outreach activities would almost always need to family planning and prenatal services, and care be attended to. But there would be scope for for newborn and young children. nurses' involvement in curative care, especially in localities with concentrations of poor 3.45 At the same time, government policy households. Alternatively, puskesmas PKs and expects the bidans to become self-supporting. bidans (and doctors) might play supportive and This is not seen as occurring inmediately. quasi-supervisory roles in areas with well- Provision is being made for two three-year established private nursing and midwifery contracts per midwife, as compared to the single practices. For instance, health center staff might three year contract offered to PTT doctors. offer periodic hands-on training and guidance in Nevertheless, it remains unclear how many regard to difficult or unusual cases. Moreover, village midwives can be "absorbed" on a self- locally based nurses and midwives could be sustaining basis. In this regard, estimates were asked to participate in health center clinics or developed of the monthly income that newly other programs. trained midwives might expect to earn from provision of delivery and family planning 3.43 Village Midwives-Economic services, outside of public settings. Estimated Conditions for Success. The availability of income from these sources is expressed as a possible candidates for bidan di desa positions percent of the current monthly salary that could will not guarantee that Accelerated Program be expected by a government midwife in the graduates will be able to establish themselves at third year of her service, or about Rp 130,000 the village level. Success will depend in part on (US$ 62) per month.6 As discussed, plans call the effectiveness of training and rigorous for hiring village midwives on contractual terms adherence to selection rules which assure that at considerably higher per monthly salaries than bidans are trained in their home province. But those now paid to health center bidans. training itself must be designed with a view to Nevertheless, Rp 130,000 is a useful benchmark the skills and personality qualities which are of what midwifery students would be expecting crucial, and the concerns, decisions and trade- as their base monthly income. offs which are likely to arise for individual midwives. Unfortunately, little is firmly known 3.46 The estimates which are presented reflect on the latter questions. This quasi-experimental alternative assumptions about how quickly the program is still very new, with the first batch of new bidan would become established in the bidans having gone to villages during the last community and be able to earn income from the two or three years. provision of delivery and family planning services (Table 3.13). Potential earnings are derived as a function of the percent of births 60 Table 3.13: Total Estimated Income from Family Planning (FP) and Delivery Services, by Province, 1993 Est. Income as EstUmated Income as Estrmated Income as Esiimated Incrome as Estimated Income as Estimated Income as Esbmated Income as Percent of Expected Percent of Expected Percent of Expected Percent of Expected Percent of Expected Percent of Expected Percent of Expected Monthly Salary Monthly Salary Monthly Salary Monthty Salary Monthly Salary Monthly Salary Monthly Salary Delivery Services Delivery Services Delivery Services Delivery Services Delivery Services Province No Change In Market All othee + 25% TBA All oer + 60% TBA All other + 75% TBA New FP Use Minumum Maximum (3) (4) (5) (6) (7) DTAceh 3.84% 6.69% 9.12% 11.54% 1.25% 5.09% 12.79% Sumatera Utara 15.47% 22.86% 26.53% 30.21% 4.25% 19.72% 34.45% Sumatera Barat 10.58% 14.48% 17.73% 20.98% 4.48% . 15.05% 25.45% Riau 20.64% 32.05% 42.68% 53.30% 9.32% 29.95% 62.62% Jambi 6.72% 14.57% 21.43% 28.30% 4.81% 11.53% 33.11% Sumatera Selatan 17.12% 27.81% 36.32% 44.83% 8.46% 25.58% 53.29% Bengkulu 9.23% 15.63% 20.08% 24.54% 6.08% 15.31% 30.62% Lampung 22.50% 40.90% 58.00% 75.11% 18.53% 41.03% 93.63% Regional 11.77% 18.78% 24.15% 29.52% 5.37% 17.14% 34.89% Jawa Barat 14.26% 35.53% 56.38% 77.23% 21.64% 35.89% 98.87% JawaTengah 10.54% 24.97% 38.61% 52.25% 16.51% 27.05% 68.76% Jawa Timur 17.39% 31.18% 44.41% 57.64% 20.81% 38.21% 78.45% Bali 45.41% 62.14% 67.38% 72.61% 46.21% 91.62% 118.82% RegionalAve. 14.66% 30.90% 46.30% 61.71% 19.91% 34.57% 81.62% Kalimantan Barat 323% 7.77% 10.59% 13.41% 2.35% 5.57% 15.75% Kalimantan Tengah 5.06% 15.13% 20.95% 26.78% 4.53% 9.59% 31.31% Kalimantan Selatan 4.60% 9.02% 12.89% 16.75% 2.81% 7.40% 19.56% Kalimantan Timur 16.47% 20.57% 23.00% 25.42% 6.92% 23.39% 32.35% Regional Ave. 5.00% 10.10% 13.48% 16.87% 3.15% 8.16% 20.02% O Sulrvesi Utara 11.42% 18.54% 23.97% 29.39% 14.36% 25.78% 43.75% Sulawesi Tengah 5.09% 15.21% 21.07% 26.94% 5.14% 10.23% 32.08% Sulawesi Selatan 38.06% 103.26% 122.03% 140.80% 19.80% 57.86% 160.60% Sulawesi Tenggara 6.38% 21.65% 34.26% 46.87% 5.46% 11.84% 52.33% RegionalAve. 13.92% 35.57% 45.18% 54.80% 10.86% 24.78% 6S.66% NTB 16.35% 75.61% 123.98% 172.34% 27.94% 44.29% 200.29% NTT 6.13% 24.38% 34.56% 44.74% 6.89% 13.02% 51.63% Maluku 5.36% 12.07% 16.40% 20.73% 4.06% 9.42% 24.78% Irian Jaya 7.89% 38.91% 43.57% 48.22% 3.68% 11.57% 51.90% TimorTimur 6.11% 50.22% 54.08% 57.94% 5.86% 11.96% 63.80% Regional Ave. 7.04% 28.67% 38.89% 49.11% 7.25% 14.29% 56.36% OverallAve. 2.80% 23.58% 32.83% 42.07% 10.61% 13.41% 52.68% delivered by midwives, traditional birth would not be sufficient to assure sustainable attendants or others (relatives or no-one at all) as living arrangements for newly place midwives. measured by the 1991 DHS; an average per Once the midwife is released from the delivery fee of Rp 15,000 is assumed. Estimates govenmment contract and expected to work reflect differences in these measures as well as exclusively as a private provider, she is likely to provincial variations in the crude birth rate and experience a significant loss of income relative average village size. If there is no change in the to current civil service salaries, and even more current market for delivery services at the significant decreases relative to the contract- village level (i.e., the percent of deliveries based remunerative currently under discussion. attended by non-midwives does not change), the This exercise suggests that further thought needs newly trained midwife can expect to earn about to be given to the demographic and economic three percent of the current government salary constraints bearing on the demand for the through private delivery services (see column 3 services of the newly trained midwife. In of Table 3.13). However, this estimate varies essence, the calculations show that there simply significantly by province, with higher possible may not be enough births, or enough family eamings in Sulawesi Selatan and parts of planning users in any one village to assure that Sumatra relative to the Eastern Islands. the bidan di desa can earn what government Prospects under alternative scenarios are workers are paid. One possible outcome is for somewhat more promising (columns 4-6). But the prices of these services to increase, with even in the "best case" scenario, that is, if the consequent negative effects on affordability, bidan di desa "took over" all 75 percent of the particularly for the poor. Another possible delivery care now provided by traditional birth consequence, is that many midwives would attendants and was also successful in attracting move to urban areas where social and cultural all those who currently use other sources of care conditions are less difficult, and somewhat (including family members), midwives could higher patient charges can be expected (at least expect to earn on average about 42 percent (or initially). Rp. 18,000) of their gross monthly salary from private delivery services. Estimates (Annex 3.48 Altematives to the "one bidan per Table 31) of the additional income that the village" goal of the Acceleration Program may midwife might be able to eam through the need to be explored to improve eamings provision of family planning are quite low--such prospects for village midwives, and to assure work might add only ten percent of monthly that their services are affordable. One option civil service earnings on average. Estimated would be to look for ways to broaden the service total future earnings from midwifery and family territory for midwives, thereby bringing demand planning services range from a minimum of to a level that yields a sustainable income. This about 13 percent of current salary (assuming no would require greater specificity in the change in current market share) to a high of identification of areas which should have these about 53 percent total, if the newly trained services than is currently envisioned. Moreover, midwife is able, on arrival at the village level, to widening service areas could also carry take over 75 percent of all deliveries as well as implications for transportation and full coverage of family planning users. communications-some capital and recurrent expenditures may be entailed as part of the 3.47 Estimated earnings vary by province, policy package. Reexamination of the needfor with the potential income from family planning the program on the basis of accurate services generally higher in Java, Bali and information on local health conditions and Sumatra where contraceptive prevalence rates population density, as well as local demand and are quite high. Potential earnings are lower in supply of these services would be a first step. Eastern Indonesia, reflecting low current levels Achieving the degree of accuracy, however, of contraceptive prevalence. It seems reasonable would not likely be possible through a top down, to conclude that these levels of potential income 62 normative approach. Mechanisms which could a mechanism for articulation of overall encourage villages to apply for the services of a midwifery policy and assures that the objectives new bidan might be necessary, but again, of the program evolve in conjunction with other presume a level of demand for midwifery health policies. Last, and although peripheral to services that is currently not well articulated nor the central focus of this paper, even a well- easily assessed by government agencies. At the trained and supported midwife can do very little least, this would imply that considerable efforts to achieve sustained improvements in maternal to build demand for delivery attendance by and child health in the absence of the midwives will be necessary in the short term. improvements in referral structures and Providing practical evidence of the potential behavior, formal and informal, that the midwife effectiveness of this worker, through, for must rely on to manage emergencies and example, placement of partially or fully conditions beyond the capacity of village subsidized, appropriately trained midwives on a facilities. Improving the internal efficiency, demonstration basis might build community accessibility (especially of emergency care interest, and could also be useful in estimating services on a 24 hour basis) and quality of willingness to pay in different settings. services at the Type D hospital level and among selected health centers would be key elements of 3.49 Prerequisites for Success at the Village such a policy. Each of these areas is discussed Level. The earnings estimates presented above below. assume a degree of acceptance by the community of the bidan di desa. In this respect, 3.50 Entry Requirements and Selection a number of policy steps may be taken to ease Procedures. Entry requirements and the entry of midwives into village settings, and recruitment procedures filter the human to improve their performance and impact on resources that flow into the nursing and health status. First, recruitment strategy and midwifery pool. Modification in two aspects of entry requirements for training need to be these procedures would likely influence the supportive of broad health policy objectives, and potential effectiveness of the bidan di desa structured to maximize willingness of recruits to program. Current policy specifies that entrants take up and maintain careers at the village level. to SPKs are junior secondary school (SMP) Second, pre-service training strategies and graduates, a pragmatic choice given the procedures must be able to transfer clinical, relatively larger pool of female graduates at this managerial and interpersonal skills which the level relative to graduates of senior secondary bidan di desa will need to employ to work school. However, two aspects of this policy are successfully in village settings. Third, these problematic. First, setting entry requirements at skills need to be reinforced, refreshed and this level, in combination with the decision to brought up to date on a regular basis through dispense with the requirement that SPK effective supervision and continuing education. graduates have at least two years of practical The supervisory system can also provide a experience prior to the entry into the one year valuable mechanism for updating policy and midwifery program, has resulted in trainees who program perspectives among service providers lack the academic skills needed to be able to and, when well-designed, can provide a valuable absorb the basic science elements in a channel for feedback on program and training curriculum that would be appropriate to the policies. Fourth, developing each of these actual work of both PKs and bidans (in public or elements, and assuring consistency among them private settings). Moreover, this entry requires the establishment of a clear policy requirement creates an essentially framework which, in conjunction with insurmountable barrier between the PK/bidan appropriate institutional links through the stream of nursing and midwifery and the D3 professional midwives' association, training level (for which a secondary school degree is a organizations and health policymakers, provides precondition).7 The policy reinforces the lack 63 of complementarity between the two basic sub- the acceptability and effectiveness of the bidan di professions, relegates PK and bidan training to desa program would be strengthened through an a secondary non-professional status, and early reconsideration of the entry requirements undermines the potential to create opportunities for the training programs. However, decisions for career development for this major element of on this matter will need to be taken with due the health labor force. As Indonesia continues consideration of the desired future linkages to record improvements in educational between D3, SPK and midwifery programs, as attainment, this situation will likely become discussed further below. increasingly inappropriate from demand and supply side perspectives. 3.53 On the other hand, even the junior secondary school requirement creates problems 3.51 In addition, this entry requirement for the rapid development of midwives in many involves the preparation of midwives who are settings, particularly in Eastern Indonesia, where not well-suited sociologically or culturally for female educational attainment remains low. As their future employment. On average, graduates a consequence, an exception has been made to of the bidan di desa program are about 18 years the basic requirement for Irian Jaya, which is old, typically unmarried. Observations during developing a separate approach to the the last three years of experience with the bidan recruitment of trainees for the bidan di desa di desa suggest that young, single women are program. Irian Jaya plans, under an initiative not perceived by potential consumers as having called Program C, to accept trainees directly sufficient experience to provide credible advice from junior secondary school without the and counseling to often older clientele. Relative expected three years of preliminary SPK youth and lack of experience also undermine the training. In addition, plans call for basing capacity of the bidan di desa to establish an midwife training in field settings such as health effective working relationship with the dukun centers, and for the involvement of doctors and bayi, who is often an older woman who has other staff in on-the-job training. The decision completed her childbearing years. When to develop entry requirements specific to combined with the potential for significantly particular provincial needs is appropriate. But above average incomes relative to the general as with the larger program, the program calls population (particularly if current plans on the for a very rapid increase in the numbers of contract payment scheme come to fruition), trainees in the next three years, and thus may many young bidan di desa may face complex not be achievable. Moreover, there is doubt sociological difficulties, at a vulnerable age, in whether this approach to training will be feasible remote and poor villages. in Irian Jaya, since geographic and, more significantly, cultural factors limit the utilization 3.52 Upgrading the entry requirement for SPK of health centers for any type of health service, and bidan training to the senior secondary school including delivery and family planning. level is strongly advocated by the PPNI (the Indonesian National Nurses' Organization) and 3.54 A second problematic dimension of the IBI (the professional midwives' association) current recruitment policy relates to how to and is recognized as desirable by concerned assure that bidan recruits are selected from the policymakers in the MOH. In 1983, the same areas to which they will return. Local Consortium of Health Sciences reached a policy selection rules are recognized internationally to decision to upgrade all SPK level training to be a valuable way of minimizing social and professional status, which would include cultural distance between nurses and midwives increasing the entry requirements for SPK and their clientele. However, the acceleration training to senior secondary school graduates. program has already resulted in some departures To date, however, pressures to proceed with the from this principle. Thus, current plans acceleration program have delayed follow-up on anticipate that as many as 866 entrants to the these recommendations. It is highly likely that 1993/94 pool will be trained outside of the 64 originating province due to school capacity Moreover, since about 78 percent of all births constraints;8 400 of these recruits will be occur at home, schools need to get students into trained in Jakarta, where it is especially difficult the community and home settings that the bidan to access practical, field-based experiences. di desa will experience upon entering the village. Although stricter adherence to the local Other factors, including relatively high student/ recruitment principle would require a slow down teacher ratios, and the relative lack of in the annual pace of production, the benefit to experienced midwives in leadership and teaching both graduate and her clientele of being trained positions further inhibit the development of in social and cultural settings as close as possible experiential, hands on training under the current to the future employment location are system. These problems are compounded by the considerable. difficulties of securing sufficient client loads to allow for practical experience in providing 3.55 Pre-service Training. Public health family planning methods and information. experience provides ample evidence that perceived quality is a major determinant of 3.57 Additional work is needed to clarify the utilization, particularly as concems services for clinical tasks expected to be performed by the which there is low community and client bidan di desa. For instance, the capacity and recognition of the potential benefits. The role of midwives in handling abnormal and presence of a culturally acceptable alternative to emergency deliveries and abortions, managing professional midwifery services in the form of specific contraceptive technologies, especially the traditional birth attendant underscores the Norplant, and responding to sexually transmitted need to assure that any new provider is diseases in general and HIV/AIDS in particular, perceived to be competent and able to improve all need to be resolved. maternal and child health. Assuring the quality and appropriateness of pre-service training is 3.58 Training must also help recruits develop therefore fundamental to easing the entry of the the interpersonal and communication skills that bidan di desa into the village "market" for will enable them to work effectively with other matemal and child health services. providers, particularly traditional birth attendants, as well as consumers of health 3.56 Three aspects of the content of current services. The bidan di desa, if properly preservice training deserve further attention in prepared, can be an effective channel of this context. First, sufficient hands on, practical information on various aspects of matemal and learning experiences are needed since the bidan child health. Training which helps the bidan di di desa role requires extensive clinical skills for desa to understand the social and cultural needs both obstetric services and some contraceptive of the very poor, especially the diverse methods including, possibly, providing and indigenous populations in the Eastem Islands removing Norplant insertions. This in turn may require focused ethnographic work and requires that trainees have opportunities to experimentation. Developing realistic case participate in deliveries and provide family materials on the challenges of working with planning services, ideally in settings similar to dukun bayi, who may perceive the new bidans those in the eventual placement villages.9 as competitors in a limited market, could lead to Demographic patterns need to be considered in cooperative relationships and an effective assessing the adequacy of practical experiences informal referral system. Lessons being learned for training. For example, the number of births in demonstration projects now underway in that might be available on a per hospital, per Lombok, and in West Java could be of particular student basis, needs to be noted (Table 3.14). interest in this regard. Even under favorable assumptions about the number of births and place of delivery, there 3.59 These aspects of the bidan di desa may be insufficient births at the hospital level to curricula will require changing from a lecture sustain an effective experience-based curriculum. and classroom based mode of training to one 65 Table 3.14 Estimated Number of Births per Student at Hospital Level Est No. SPK Student Estimated Estiratod Number Number Births %-Births in all Delivered in Current Accelratd Births/Student Bitih/Student Province Population of (CBR=25) Hospltl Hospial Enrollment Enrollment Current Accelerated DTAceh 3,855,532 9,639 3.8 366 360 720 1 1 Sumatera Utara 11,453,652 28.634 9.2 2,634 800 1,400 3 2 Sumatera Barat 4,342,269 10,856 14.3 1,552 280 480 6 3 Riau 3,558,711 8,897 6.2 552 80 120 7 5 Jambi 2,433,769 6,084 6 365 160 160 2 2 Sumatara Selatan 6,799,640 16,999 14.2 2,414 360 520 7 5 Bengkulu 1,185,038 2,963 5.2 154 160 160 1 1 Lampung 7,591,265 18,978 n.a. Average 8.41 1,148 314 509 4 3 DKT Jakarta 11,388,513 28,471 33.7 9,595 400 440 24 22 Java Barat 27,453,447 68,634 5.4 3,706 800 1260 5 3 Java Tengah 31,616,987 79,042 8.3 6,561 800 1080 8 6 Yogyakarta 4,233,817 10,585 27.5 2,911 160 200 18 15 Jawa Twnur 34,099,095 85,248 14 11,935 1000 1160 12 10 Bali 2,651,370 6,628 22.5 1,491 80 120 19 12 a, Average 18.57 6,033 540.00 710.00 14 11 Kalimantan Barat 4,327,322 10,818 11.6 1,255 80 0 16 Kalimantan Tengah 2,705,898 6,765 3.3 223 80 85 3 3 Kalimantan Selatan 3,317,733 8,294 5.9 489 200 160 2 3 Kalimantan rimur 1,723,545 4,309 15.7 676 80 120 8 6 Average 9.13 661 110.00 91.25 7 3 Sulawesi Utara 2,506,195 6,265 22.4 1,403 160 240 9 6 Sulawesi Tengah 1,934,038 4,835 10.6 513 80 6 Sulawesi Selatan 7,538,298 18,846 16 3,015 360 480 8 6 Sulawesi Tenggara 561,783 1,404 8.7 122 80 120 2 1 Average 14.43 1,263 170.00 280.00 6 4 NTB 2,567,276 6,418 3.6 231 80 160 3 1 NTT 4,228,177 10,570 9 951 120 120 8 8 Maluku 1,965,343 4,913 14.4 708 120 240 6 3 Irian Jaya 3,139,144 7,848 14.5 1,138 200 240 6 5 Timor Timur 1,042,590 2,606 9 235 40 40 6 6 Average 10.1 652 112 160 6 5 Overall TotaUAverage 190,220,447 475,551 12.13 1951.63 249.26 349.96 7 5 which is field-oriented and experiential. been identified, the work of translating these However, the 1989 Assessment of Midwifery desires into specific job competencies remains Education and Practices found that school-level incomplete. This in turn has limited the constraints to achieving the desired degree of development of specific clinical supervision and field practice were considerable. For instance, quality control tools and procedures which could there were severe shortages of appropriately support and guide the on-the-job performance of trained midwifery teachers and clinical bidan di desa. There is an absence of clearly instructors, guru bidans.'0 Moreover, schools articulated vision of how the bidan di desa is officials reported that while it was relatively supposed to relate to the formal health system, easy to establish working relationships with including the health center and its associated nearby hospitals for field and clinical outreach activities. experiences, establishing workable relationships with local health centers was more difficult, with 3.61 To date, Pusdiklat has been able to responsibilities on the part of the health center provide in-service training to only 1000 bidan di and school administrators for organizing field desa, through programs based in 13 provincial training unclear. Even in settings where health training centers. Major constraints to center staff are able to assist, limited strengthening Pusdiklat's contribution to in- transportation and housing at field sites can service technical training include the lack of severely limit the amount of time students spend professional trainers in midwifery, limited at the health center level and the quality of their facilities and materials and the lack of the job experience. Moreover, field training in the competencies necessary to develop appropriate health centers may not adequately provide in-service instruction. experience in village and home-based settings. Another finding of the assessment is that the 3.62 Several policy directions will need to be reliance on part-time teachers and guest lecturers pursued to strengthen in-service training and may undermine efforts to provide clinical and supervision systems. First, improved field experiences during training." supervision capacity will be needed, with responsibility for providing clinical and 3.60 Inservice Training and Supervision. administrative guidance to the bidan di desa Pre-service training must be reinforced through placed primarily at the health center, and backed inservice training and supervision mechanisms if up at the kabupaten level. Evaluations of health service quality is to reach acceptable levels. programs in a number of settings suggest that Some guidelines on the role and work routines ratios of one supervisor to seven to ten direct for the bidan di desa have been developed. But care providers are necessary for effective service actual supervisory capacity and procedures are delivery. Creating positions for senior bidans at weak, while potential for on-the-job training is the kabupaten level would help achieve the undermined by the many administrative tasks necessary staffing intensity, and would also faced by health center and kabupaten level health establish a career ladder for nurse midwives. staff. At present, the health center doctor and bidan staff are responsible for the supervision of 3.63 A second option, which fits well with bidan di desa; no-one is responsible for plans to set up the bidan di desa as a private midwifery supervision at the district level. practitioner, would be to encourage greater Tools and materials which could assist the involvement of the Indonesian Midwives' supervision process, such as detailed guidance Association (IBI) in providing continuing on how to monitor the quality of service education for the bidan di desa, as well as for provided, or standard protocols to guide specific local supervision. To date, IBI's membership, clinical procedures are not widely available organized in provincial chapters, has played only outside of limited demonstration project areas. an indirect role in determining midwifery policy. In addition, although the types of tasks and Yet their capacity to shape behavior could be services to be provided by bidan di desa have considerable, as recently evidenced in the 67 establishment of a peer-to-peer system to train the completion of the village midwives training private bidans in family planning skills. program in 1996/97. Decisions will be required Professional associations are naturally concerned relating to at least two dimensions of nursing with the development of high practice standards, and midwifery policy. First, it will be necessary and can, if given sufficient voice in licensing to clarify roles and performance expectations for matters and the certification of service sites, different levels of nurses, from the basic D3 become a major influence on behavior. This level through specializations in fields such as could also provide a way to assure that older, midwifery, pediatrics and public health nursing. more established private bidans could contribute Second, policy decisions will be needed to create to the development of the profession. a career path through various levels of nursing and midwifery and to define career expectations 3.64 Developing clear approaches to and professional competencies (with associated evaluating midwife performance will be examination and licensure requirements) for each fundamental to strengthening private and public stage and level of profession. Although there is supervisory support to the bidan di desa. This room for debate on the pace at which the SPK in turn requires further work to improve the level training should be upgraded, there is little means of assessing maternal and infant health doubt that efforts to upgrade the SPK should outcomes at village, health center and kabupaten begin by the end of Repelita VI. To accomplish levels. Experimentation with medical audit these steps, it will be important for CHS, PPNI techniques and verbal autopsies for measuring and IBI to collaborate on developing the roles trends in mortality are positive steps in this and functions of each level of nursing and direction. The development of performance midwifery. measures will also need to consider how to incorporate client and consumer views into the 3.66 Several observations seem relevant to the evaluative framework. Indonesia could make task of developing a broad policy framework for considerable use of innovations in the quality nurses and midwives. First, different elements assurance field in developing ways of assessing in the overall demand for nursing services need bidan di desa performance. to be identified. Private curative care comprises one such component-this is seen in the heavy 3.65 Policy as Regards Basic and Academy- reliance on PKs, especially those based in health Level Nurses. The recent focus on training and centers, for curative care. Recourse to nurses deploying village midwives has postponed for such services may be especially widespread resolution of various policy questions in respect in poorer areas which may be less remunerative of PK and D3 nurses. One issue relates to the for physicians. There is also demand for nurses role and training envisioned for basic nurses. In for hospital-level care. The strength of this this regard, there has been little follow-up to the component is seen in the rapid growth of 1982 decision in principle to eliminate SPKs and private, often hospital-linked D3 training raise entry level requirements for basic nurses. programs. A final demand element relates to the The process of upgrading SPKs to D3 level need for improved maternal and child care at the nurses has been constrained by limited financial community level. resources, and by the lack of faculty with sufficient experience to sustain a larger scale D3 3.67 A second observation relates to the likely program or to develop a larger program for close relationship between the utilization of producing S1 nurses. In addition, there is a health services and their quality. Policy may widespread perception that D3 graduates are have erred in fostering large increases in the reluctant to practice at the health center level, number of nurses and midwives, without especially in rural or poor areas. It will be considering quality impacts. A third point of important that plans to eliminate SPK-level reference is the recent decision to severely training and upgrade selected SPKs to D3 level restrict civil service growth during Repelita VI. institutions be prepared for implementation upon This decision reinforces the need to more 68 systematically encourage and assure quality (according to explicit accreditation criteria) control in the development of private training public and private D3 Academies. institutions. A careful review of numbers and capacities of individual SPKs, regardless of 3.69 Clarifying Roles and Competencies for ownership, will be necessary to assure that the PKs. Currently, there is confusion concerning upgrading process focuses on selective PSKs, the desired role and competency of the PK consistent with future demand levels, and nurse. For example, the PK curriculum focuses identifies which SPKs would be closed. on a vision of nursing that anticipates a major Eliminating some of the SPKs and D3s, and role in community health activities. Training making recruitment and selection more does not adequately reflect the heavy competitive would also be consistent with the involvement of the PK in provision of curative zero-formasi policy, and the broader concern services in the health center, in substitution for that educational policies for the the often busy or missing health center doctor nursing/midwifery professions focus on quality and in private practice. The curriculum also of training, rather than on norm-driven presumes that the nurse works in tandem with a production targets. In this regard, the health physician who is responsible for diagnosis and system in Indonesia would perhaps be best treatment. But the more typical situation is for served, in the medium to long term, through the the PK to operate as an "independent" provider. development of self-governing nursing and Also, as the bidan di desa program evolves, it midwifery professions. Hence, helping will become increasingly important to distinguish professional associations to improve the between the village midwife's role and that of regulation and competence of their membership, the PK in community health. and the quality and responsiveness of their services, could itself be a potent policy 3.70 The present situation presents an instrument. opportunity to use training and deployment to develop the PK as a care provider, with a legal 3.68 Career Paths for Nurses and mandate to provide specified curative services. Midwives. The different dimension of demand This would require reorienting the basic cited above will need to be addressed through curriculum with a view to turning PKs into role and task differentiation within the nursing nursepractitioners (sometimes called physician- profession. The pyramidical structure extenders), with all of the clinical training in recommended in several expert reports (Roemer, medical diagnosis and treatment that this would et al, 1990; Reid, 1993) may provide a useful imply. Such a policy could help remedy framework. Entry-level professional nurses with problems of attracting doctors to some remote D3-level credentials would form the base, with areas. Moving in this direction would require nursing specialists in various functional and that entry qualifications for SPK training be clinical areas filling out the rest of the pyramid. upgraded from junior secondary school This structure would allow specialization and completion to at least senior secondary school career development, building on common basic attainment. MOH would also need to develop a nurse training, in response to demand for more explicit regulatory stance governing private hospital and private curative care, community care by PKs, accompanied by meaningful health nursing and midwifery. Bringing about examination, licensure and certification such an adjustment in the structure of the practices. In this fashion, the PK could profession, and more specifically, finding complement the specialized maternal and child nursing/midwifery faculty and managers who health role to be played by the bidan di desa. can lead the development of the upgrading process, will depend in turn on an early decision 3.71 A strong case can be made to raise entry to enhance the number and quality of training requirements, whatever subsequent role is programs for S I level nursing graduates, projected for PKs. The reasons are analogous to perhaps through upgrading selected, qualified 69 those cited in the case of the bidan di desa the response to these issues be guided and program. Unless entry requirements for PK managed? training are elevated to senior secondary school completion, it will be impossible to develop a 3.74 Policy Formulation. Policy on the D3 career ladder which would allow nurses to curriculum is the responsibility of a working progress from PK level jobs to more senior group convened under the auspices of the positions as D3 nurses. Furthermore, PK Consortium of Health Sciences, which is training suffers from the same limited exposure administered by MOEC. Responsibility for the to hands-on training that is an issue with the PK and bidan di desa curriculum as well as bidan di desa program. A range of problems monitoring the qualifications and allocation of including small hospital case loads, inadequate resources to SPKs and Academies rests with faculty numbers and skills, and difficulties of Pusdiknakes.'2 Pusdiklat is concerned with in- arranging for field practice limit the degree of service training. The Directorate of Family hands on training for PK students. Health within MOH is also actively involved in the provision of support to the midwife program. 3.72 Directions for D3-Level Training. Recently, MOH formed a steering committee on Nursing academies are the major source of bidan di desa training that includes the nursing staff in class A and B, specialty and Indonesian Midwives Association (IBI). While private hospitals and a major supplier of teachers certainly valuable for ad hoc coordination, this for both PK and bidan training. This will latter group appears to lack the authority to remain the case in the absence of a significant provide policy direction concerning broader increase in the production capacity of the nursing questions. Baccalaureate nursing program Therefore, upgrading D3 training, to promote a large pool 3.75 School Accreditation and Management. of recruits for nurse and midwifery education Current arrangements for school accreditation and managements positions, should be the are limited in scope. Though schools prepare central objective for public spending at this draft work plans, budgeting remains highly level. Specifically, public resources should be centralized; there is little variation in allocations used to develop well qualified nursing faculty by location or setting. Pusdiknakes monitors and leaders, in sufficient numbers, through basic aspects of school quality, including the limited and focused subsidies at the D3 level. number of faculty, the availability of equipment At the same time, the involvement of private and transport, and, to a lesser degree, hospitals in developing and operating D3 level availability of training materials, and also is facilities should be carefully monitored and responsible for revision and distribution of regulated. There will also be a need for curricula for PK and bidan di desa training. subsidies to ensure that students from poor areas However, site visits are ordinarily not can benefit from training opportunities. undertaken, and the information submitted allows for little evaluation of teaching processes 3.73 Oversight of Nursing and Midwifery or outcomes, and does not provide for self- Policies. Some challenges lie ahead in respect assessment by teachers or school managers. A of nursing and midwifery policies. There is an common complaint among schools is that most inunediate need to revise plans for the of the limited teaching materials are available acceleration of the bidan di desa program, and to only in English, which limits their utility, address the widely felt problems of insufficient particularly for junior secondary school nursing skill and quality of care in both hospital graduates. Resources for translation and and health center settings. And there is a need duplication of materials are inadequate. to recognize and regulate nurses' involvement in Moreover, Pusdiknakes lacks the funds and providing private curative services. How should capacity to routinely assess actual PK or bidan di desa performance or conduct research which 70 would contribute to either curriculum revision or examinations and establish registration and better targeting of budget resources. licensing requirements, including specification of continuing education or refresher training needs. 3.76 Licensing and Certification. PK and This body could be made up of members drawn D3 nurses are not currently licensed, and from the midwifery profession and from the systems for routine reexamination of clinical or Ministry of Health, and should include midwives administrative skills for PKs, D3s and bidans holding teaching, managerial and clinical have not been developed. A rudimentary system positions, representatives of at least two for licensing bidans exists, with authority for Provincial Boards, medical officers from issuing a license to practice resting with relevant disciplines (especially obstetrics/ provincial health authorities. The work of PKs gynecology and pediatrics), a lawyer, and in curative care and treatment at the health representatives from consumer organizations. center level is also not formally sanctioned. However, health center doctors routinely issue a surat wewenang which informally authorizes the PK to follow the physician's treatment orders. In practice, the surat wewenang is broadly interpreted, and provides PK with a measure of legal protection against malpractice. Licenses for the bidan di desa are issued upon placement at the village level, or civil service post, but there are now no requirements for periodic recertification. As discussed, private practices managed by health center level PKs are not legally recognized or sanctioned, nor are they monitored formally or informally for quality. 3.77 In sum, regulatory and quality control mechanisms covering training and actual clinical practice for both nurses and midwives are embryonic at best. Although professional associations exist for nurses and midwives, their role has been limited to participating in selected task forces and working groups through the Consortium of Health Services and Pusdiknakes. Both organizations maintain small central offices in Jakarta and a network of provincial and district level chapters. 3.78 Achieving qualitative improvements in training and practice will require a more coherent and effective structure for nursing and midwifery policy. For midwives, establishment of a statutory National Board, with associated Provincial Boards of Midwifery, is a promising intiative.'3 Such an entity should have broad oversight of midwifery education including the authority to accredit public and private training schools, assess midwifery students through 71 3.79 The establishment of analogous structures and processes for nursing overall would also be desirable. A Nursing Board would provide a mechanism to assure that governmental authorities as well as representatives of the profession associations were actively involved in determining policy directions, and would provide a means of enhancing the accountability of staff work on nursing and midwifery policy within the central and provincial departments of health. Endnotes 1. Province-specific information, presented in Annex Tables 3C, 3F, 3G, and 3H reveals shortfalls relative to government norms in most kabupatens, in a variety of provincial settings. 2. These data do not distinguish between nurses and midwives. Since the results referred to those who were ill, it is unlikely that these indicators covered routine consultations related to pregnancy and family planning. Thus, the extent of private visits to midwives and dukun bayi was probably underestimated in this survey. 3. Hospitals whose staff numbers relative to patient loads fell one standard deviation or more below predicted levels were considered to be more "efficient". 4. In the first regression, the visit rate (V) was found to vary directly with the number of nurses and midwives (NM). V = 29.5 + 20.4 NM R2 = 0.12 (3.1) (3.6) I t - statistics in parentheses; R-squared = 0.12; number of observations = 941 In the second specification, NM was included along with the a number of doctors per health center (D), the proportion of facilities with mother and child health kits (M), the population density in the area served by the puskesmas (P) and a variable indicating location outside Java and Bali (0). V = 1.1 + 15.1 D + 42.2 M + 0.68 P + 3.4 NM - 15.3 0 (0.07) (1.5) (5.6) (2.8) (0.7) (-1.8) [t - statistics in parentheses; R-squared = 0.54; number of observations = 94]. 5. Katherine Neilan arrived at similar findings in her 1993 study of service delivery in ten health centers in Lombok. Neilan observed that "When the nurse or paramedic gave care there seemed to be no consistent diagnostic rationale. Patients were often seen in rapid progression and the health care worker began writing out the prescription almost as soon as the patient sat down. Often, the prescriptions did not make sense" (Neilan, 1993, p.6). 72 6. Annex Table 3K works through these calculations. 7. Under current conditions, an SPK graduate (PK or bidan) wishing to advance further in the nursing profession must withdraw from the labor force to apply for admission to nursing academy. This creates a major disincentive for further career or professional development. 8. Affected provinces are Kalimantan Barat (174), Kalimantan Selatan (160), Sulawesi Tengah (54), Riau (52), Java Tengah (160), Nusa Tenggara Timur (80), Maluku (40), and Timor Timur (40). 9. Members of IBI have observed that newly placed village midwives are referring "too many' cases, i.e., clients of not abnormal risk whom they are hesitant to handle. 10. As reported in the Assessment, "There is a significant shortage of guru bidans in almost all schools.. .Some have never done such important procedures as IUD insertion and 90 (of 137 guru bidans surveyed) have no experience working in a health center. Thirty of the 137 have not had any midwifery practice for more than five years (range 5 to 23 years). Some midwifery programs have no midwives at all among their teaching staff... Theoretical teaching is most frequently conducted by the lecture method... Students in most programs have no opportunity to contact community leaders, do assessments, teach TBAs, kaders, pregnant mothers or the community.. .Teaching equipment and facilities are used by both the SPK and PPB (midwifery program). In most schools there is not sufficient quantity for both programs. Also, the existing books and references are not sufficient and many are outdated" (pp. vii-viii). 11. Some findings of this Assessment have already been reflected in government policy. For example, the 1993/94 GOI budget for the Acceleration Program calls for establishment of a bidan pengawas, or midwifery school supervisor to be assigned at the provincial level, at the rate of one for each 10 schools in a province. The bidan pengawas will, as they are put in place, be responsible for helping schools to develop links with field training sites, and will manage the collection and reporting of data on school characteristics for accreditation purposes. 12. This division of responsibilities dates back to a 1983 decision to establish nursing, at the D3 level, as a profession, rather than a paramedical vocation (PK level). A decision was also taken to upgrade all SPKs to Academy status. Meanwhile, the nursing working force continues to be divided into separate professional and vocational streams. 13. This proposal is under consideration within MOH. 73 CHAPTER FOUR PARAMEDICAL WORKERS and 7000 drivers and other support staff 4.1 The remaining group of health-related assigned to health centers-these workers are staff are often referred to as paramedical mentioned here because of their potential workers. This catch-all category includes those, involvement in specific circumstances, e.g., the e.g., pharmacists, laboratory and x-ray absence of designated staff or in emergencies, in technicians and so forth, who work in hospitals health service delivery. At the hospital level, and health centers in auxiliary and supportive there are 3000 public health nurses who perform functions. The activities of this subgroup are a wide range of tasks. determined typically by instructions specified on a case by case basis by a doctor or nurse-there Training Initiatives were roughly 12,000 such staff employed in hospitals and health centers in 1992 (Table 4.1). 4.4 During the last fifteen years, GOI used training and other interventions to increase the 4.2 A second group of paramedical workers number and availability of paramedical staff includes those such as sanitation and nutrition working in government health facilities. New technicians and assistants with previously schools were opened, and significant increases determined duties and tasks. Activities are occurred in the number of paramedics who were specified, typically in generic terms, in program hired in government hospitals and health centers. handbooks and guidelines or in written plans And with government encouragement, an which may be region or even facility specific. expansion occurred in privately run training. As These staff operate as part of a larger team, with doctors, nurses and midwives, this broad e.g., a communicable disease group within a approach was driven by expected rapid increases puskesmas, and are subject to supervision by in the demand for such employees. It needs to doctors and others. However, they may enjoy be examined whether these large scale training considerable autonomy in regard to their own and deployment efforts were effective, and technically defined agenda. There were whether existing policies should be retained or approximately 16,000 such workers employed in modified in some fashion. hospitals and health centers and subcenters in 1992 (Table 4.1). 4.5 Training Facilities. Paramedics are trained at the high school, academy (involving 4.3 Finally, there is a large residual group of one to three years beyond secondary school) and hospital and health center workers who do not university levels in Indonesia. High school have clearly defined tasks and responsibilities graduates include dental assistants and laboratory (Table 4.1). This category includes 35,000 analysts. Academy products include those, e.g., puskesmas health auxiliaries. These are workers assistant sanitarians, laboratory workers, and with limited educational attainment (usually nutritionists who earn a one year (DI) certificate primary or middle school) who receive four and others, e.g., laboratory technicians and months of orientation training and then perform nutritionists, who receive a three year (D3) various health and administrative functions. degree. University degrees (Si) are awarded to There are also over 8000 administrative workers nutrition and public health workers who have 74 Table 4.1 Distribution of Paramedical Staff In Indonesia, 1992 Hospitals Health Staff Category A _ C D Private Total Centers Sanitarians D3 218 531 DI 40 101 205 188 31 565 4,346 Pharmacists Si 693 DO 195 587 504 314 1,242 2,842 1,001 Laboratory/Chemical Analysts D3 475 DO 165 454 429 284 930 2,262 1,594 Nutritionists Si 12 D3 58 127 135 66 149 535 89 Dl 82 189 269 217 177 934 1,604 Dental Techniclans/Assistants D3 DI 140 DO 60 187 256 283 180 966 2,672 X-Ray Technicians 03 66 153 121 40 210 590 Other Physiotherapists (03) 814 Anesthesiologists (D3) 560 Electromedical Technicians (D3) 265 Optic Refractionist (D3) 6 Orthotic Prosthetics Technicians (D3) 18 Speech Therapists (D3) 17 Social Workers (DI) 197 Occupational Therapists (D3) 10 Public Health Nurses (S1) 124 Public Health Nurses (D3) 304 742 849 881 1,589 4,365 Medical Technicians 12 Psychologists 90 Community Medicine Specialists 36 Auxiliary HeaKth Workers 9,020 (Primary School) Auxilliary Health Workers 26,079 (Junior High School) Administrators 13 8,310 Messengers 53 Driver/other 7,045 Total Non-Clinical Staff 970 2,540 2,768 2,273 4,508 16.759 62,344 Source: Ministry of Health 75 completed two years of study after earning a D3 considerable variation in quality from school to certificate. school and region to region. Those who want to open a school are supposed to provide 4.6 There are 181 paramedic schools, 99 (57 information to Pusdiknakes on the number of percent) of which are publicly owned and teachers and the availability of buildings and managed. In 1983, there were 107 schools, 64 other physical facilities. Pusdiknakes examines of which were publicly run. Five staffing ratios and the adequacy of facilities in categories-assistantpharmacists, sanitarians and deciding whether to allow a school to begin nutritionists, laboratory and dental operations. Examinations are, in theory, set by assistants-account for 70 percent of the schools. teachers appointed by the provincial education Most schools (75 percent) are located in Java officer. Pusdiknakes is also responsible for and Bali, and Sumatra and most students (75 evaluating these schools annually to make sure percent) come from these regions (Table 4.2). that standards are maintained. Somewhat The shares of private schools (43 percent) and different evaluation and quality control (45 percent) accounted for by Java and Bali, and arrangements have evolved for nutrition schools. Sumatra are even higher. The Indonesian Nutrition Association (INA) plays an important role in ensuring that high 4.7 Private sector training programs now standards are maintained in this discipline. The exist for most paramedical skills (Table 4.3). INA uses teachers from reputable government Only a few fields, e.g., anesthesiology, dental schools, including the University of Indonesia, technicians and assistants, lack a strong private to assess each nutrition school periodically; sector training presence, while private schools schools which need help may ask for technical have achieved a dominant position in respect of assistance from the Association. The INA, pharmacists, speech therapists and which sees its role as reviewing and assisting electromedical technicians, and are virtually at rather than passing judgement on schools, parity with the public sector in other fields. occasionally provides instructors to schools with specific teaching needs. Government nutrition 4.8 Government schools have more full time schools are allowed to give their own teachers than private schools. In 1993, there examinations. Private schools must link their were over 2,000 full time teachers, of which 36 tests to those of nearby government schools. percent were in private schools. The teacher- student ratios for government and private 4.10 Enrollment Levels and Distribution. schools were 1:11 and 1:15 respectively. In Each school is allowed normally to admit a addition, there were over 2300 part-time maximum of 40 paramedical students per class. teachers nationwide, 43 percent of whom Beginning in 1984/85, double shifting, i.e., 80 worked in private schools. Overall, the teacher- students per class, was mandated in selected student ratios were close to the mandated 1:10 schools in order to increase the flow of for academic institutions. Government schools graduates. have come closer than private schools to achieving the required ratios. It is important to 4.11 Nevertheless, enrollments and the note, however, that private schools rely more on number of graduates grew rapidly in the 1989-93 part-time teachers who are also full time interval (Tables 4.3 and 4.4). This increase was government teachers in some cases. These especially large in private schools. The number teachers may hold other part-time teaching jobs of students rose from 6558 in 1989 to 10,976 in in several schools. 1993, an average increase of 13 percent per year, putting private enrollment within striking 4.9 Pusdiknakes is responsible for ensuring distance of that in public schools. The that required standards of education in categories most affected were sanitarians (798 paramedic schools are maintained. However, percent increase) and laboratory analysts (288 standards are not well specified, and there is percent increase). The 1974 private school 76 Table 4.2 Regional Distribution of Paramedical Students and Teachers in Indonesia, 1993 Government Schools Private Schools Total Full Time Part Time Student/ StudenV StudenU Student/ Student/ Student/ Full Time Teacher Part Time Teacher Full Time Teacher Part Time Teacher Teacher Teacher Region Students Teachers Ratio Teachers Ratio tudents Teachers Ratio Teachers Ratio Total Ratio Ratio Regionl 4,040 376 10.7 302 13.4 2,635 206 12.8 186 14.2 6,687 11.5 13.7 Region2 5,864 608 9.6 612 9.6 7,524 464 16.2 615 12.2 13,483 12.6 11.0 Region3 1,084 85 12.8 81 13.4 322 13 24.8 28 11.5 1,406 14.3 12.9 Region4 1,640 148 -11.1 163 10.1 375 26 14.4 71 5.3 2,088 12.0 8.9 Region5 639 35 18.3 65 9.8 120 5 24.0 12 10.0 759 19.0 9.9 Total 13,267 1,252 10.6 1,223 10.8 10,976 714 15.4 912 12.0 24,423 12.4 11.4 Region1 30% 30% 25% 24% 29% 20% 27% Region2 44% 49% 50% 69% 65% 67% 55% Region3 8% 7% 7% 3% 2% 3% 6% Region4 12% 12% 13% 3% 4% 8% 9% Region5 5% 3% 5% 1% 1% 1% 3% Total 100% 100% 100% 100% 100% 100% 100% Key Region 1: DKI Jakarta, West Java, Central Java, Dl Yogyakarta, East Java, Bali Region2: Dl Aceh. West Sumatera, Riau, Jambi, South Sumatera, North Sumatera, Bengkulu, Lampung Region3: Central, South, East, and West Kalimantan Region4: North, South Central and South East Sulawesi Region5: NTT, NTB, Maluku, East Timor, and Irian Jaya Source: Ministry of Health Table 4.3 Percentage Change of Paramedical Students In Public and Private Schools by Category, 1989-93 Public Schools Private Schools Percentage Percentage Percentage Percentage Percentage Percentage Students of total Students of total Change Students of total Students of total Change Catego 1989 1989 1993 1993 1989-93 1989 1989 1993 1993 1989-93 Sanitarbans D3 1195 12.3% 1928 14.5% 61.3% 164 2.5% 1472 13.4% 797.6% DI 608 6.3% 886 6.7% 45.7% Pharnacists 03 213 3.2% 456 4.2% 114.1% DI 20 0.2% 40 0.3% 100.0% DO 1722 17.7% 1941 14.6% 12.7% 4339 66.2% 5091 46.4% 17.3% Laboratory Analysts D3 158 1.6% 325 2.4% 105.7% 311 4.7% 1208 11.0% 288.4% DO 1677 17.3% 1549 11.7% -7.6% 689 10.5% 1079 9.8% 56.6% Nutritionists D3 893 9.2% 2146 16.2% 140.3% 169 2.6% 440 4.0% 160.4% DI 398 4.1% 914 6.9% 129.6% 40 0.6% 0 0.0% -100.0% Dental Technicians/Asslstants D3 160 1.5% DI 21 0.2% 40 0.3% 90.5% DO 1979 20.4% 2538 19.1% 28.2% 40 0.6% 0 0.0% -100.0% X-tay Technicians D3 229 2.4% -100.0% Other Physiotherapists (D3) 304 3.1% 369 2.8% 21.4% 221 3.4% 355 3.2% 60.6% Anesthesiologists (D3) 370 3.8% 386 2.9% 4.3% Electromedical Technicians (D3) 298 4.5% 562 5.1% 88.6% Optic Refractionists (D3) 139 1.4% 205 1.5% 47.5% 25 0.4% 74 0.7% 196.0% Speech Therapists (D3) 49 0.7% 79 0.7% 61.2% 9713 100% 13267 100% 36.6% 6558 100% 10976 100% 67.4% Note Public Schools include Anny and Provincial Schools Source: Ministry of Health Table 4.4 Percentage Change of Paramedical Graduates From Public and Private Schools By Category, 1989-93 Public Schools Privabt Schools Percerita Pomnn Pomne ec Parcota Pment ngs Graduate of total Graduates Of totl Change rduates of totl Graduts of tobl chango lCdw~ 1 1989 1989 193 193 198993 18 1989 1"93 1993 1989-93 Sanitarlans D3 278 8.4% 952 20.0% 242.4% 60 3.0% 503 13.4% 738.3% Di 413 12.5% 446 9.4% 8.0% Pharmacists D3 152 4.0% Di 18 0.5% 18 0.4% 12.5% DO 503 15.3% 540 11.4% 7.4% 1451 73.5% 1897 50.5% 30.7% Laborao AnalystB D3 38 1.2% 117 2.5% 207.9% 95 4.8% 275 7.3% 189.5% DO 547 16.6% 590 12.4% 7.9% 157 8.0% 338 9.0% 115.3% Nutrtlonlste D3 222 6.7% 518 10.9% 133.3% 43 2.2% 173 4.6% 302.3% Di 378 11.5% 397 8.4% 5.0% 39 2.0% 77 2.1% 97.4% Denbtl TechnicanhAssldn < Di 17 0.5% 22 0.5% 29.4% DO 598 18.2% 761 16.0% 27.3% 0 60 1.6% X-Ray Technicians 03 75 2.3% -100.0% 0 37 1.0% Physioferapists (D3) 62 1.9% 188 4.0% 203.2% 30 1.5% 103 2.7% 243.3% Anesthesiologists (03) 121 3.7% 129 2.7% 6.6% Elcomedical Technicans (D3) 24 0.7% 75 1.6% 212.5% 0 21 0.6% Opfic Refractfonists (D3) 99 5.0% 99 2.6% 0.0% Speech Therapists (D3) 19 0.5% Total 3292 100% 4763 100% 44.4% 1974 100% 3754 100% 90.2% Source: Ministry of Healfth graduates in 1989 grew to 3755 in 1993, a 90 paramedic positions were filled each year using percent increase and comprising 91 percent of resources transferred through routine budget and public sector output. Sanitarians, assistant INPRES channels. Formasis assigned to the nutritionists, physiotherapists and laboratory Ministry of Home Affairs to build local analysts were among the fast growing categories. government capacity have been used in some Java, Bali and Sumatra continued to provide the instances to hire paramedics. These resources, bulk of both students and graduates in the which are typically allocated at the kabupaten private sector despite the fact that some new level, have often financed employment of health schools have been opened in other regions. center drivers, administrative workers, and cleaners, but also some paramedics. In 1992, 4.12 The increase in the number of students approximately 10,000 formasi were assigned by and graduates in public sector schools was less local government authorities-many can be dramatic compared to what occurred in the presumed to have been used to fund position in private sector (Tables 4.3 and 4.4). Public other, non-health sectors. Finally, some school enrollments grew by 37 percent in the puskesmas paramedic or quasi-paramedic staff 1989-93 interval, while graduates increased by may be taken on as "volunteers", i.e., on a 44 percent to 4138. Major increases occurred in temporary basis, until a permanent position the following categories--laboratory analysts becomes available. These workers are not paid (D3), physiotherapists and assistant nutritionists. salaries, but usually receive a regular "bonus" financed apparently from the fees paid by Employment Patterns and Trends patients or from informal, additional service charges paid by health center visitors. In East 4.13 Staffing Norms and Funding Kalimantan, 48 percent of retained earnings Mechanisms. Graduating paramedical students from fees were distributed to paramedics, have typically looked to the government for administrative workers and to a category called employment opportunities. As with doctors and "honoraria" in 1992/93. nurses, public sector needs reflect staffing norms specific to each class of hospital and to health 4.15 Employment Outcomes. These funding centers. For instance, the four class A hospitals mechanisms have resulted in distinctive are each supposed to have 867 paramedics, employment patterns for paramedical workers including 30 anesthetists, 20 pharmacists, 20 X- (Tables 4.1 and 4.5): ray technicians, 16 nutritionists, 15 physiotherapists and 678 auxiliary health 0 On the whole, health centers have fared workers with broadly defined functions. better than government hospitals in Similarly, each class B hospital is expected to regard to attaining facility-specific have 385 paramedics, including 304 auxiliaries. staffing standards. On average, each Standardized "demand" at the C and D facility puskesmas has 10.1 paramedics level is lower at 96 (69 auxiliary staff) workers (including 2.7 administrative workers, and 29 (20 auxiliaries) workers respectively in drivers and others) as compared to the every hospital. Finally, the "standard" health expected figure of eight or nine. All center is expected to have a complement of told, over 62,000 paramedics work in seven paramedical workers along with two health centers, as compared to the administrative workers. approximately 13,000 and 4,500 employed in government and private 4.14 As with other health worker categories, hospitals respectively. staffing needs in hospitals and health centers have been addressed, in part, through the * Class D hospitals have come closest to creation of new "formasi", i.e., authorized and attaining facility specific staffing norms fully funded new posts. During the 1989/90 - for paramedics. Dr. Cipto 1993/94 interval, an average of 2438 such Mangunkusomo, Jakarta's class A 80 Table 4.5 Average Number of Paramedics by Type of Facility Paramedics Hopial* enl Hoeqpfs per HospeW Paramnedks Povne A e c D P A B C Prvat AN HoepItb per HC BALI 164.0 40.3 23.5 1.5 14.0 2.0 17.8 11.0 CENTRALJAVA 117.3 24.4 15.0 12.3 26.7 8.0 0.6 17.1 11.9 D.L YOGYAKARTA 194.0 31.5 26.3 13.2 8.0 10.0 23.6 13.6 DIU JAKARTA 558.0 93.0 48.5 22.3 18.0 3.7 35.5 4.5 EASTJAVA 223.0 204.5 22.2 11.8 10.2 30.0 0.0 19.3 13.5 WEST JAVA 161.5 26.9 16.0 17.7 21.5 5.0 21.4 10.3 Re*M lAverar 390.5 155.7 32.3 18.5 12.8 22.0 7.0 3.3 22.4 11.3 ACEH 90.0 21.3 7.0 5.9 7.0 13.0 8.9 BENGKULU 52.0 16.0 6.0 10.0 18.9 7.7 JAMBI 28.0 23.2 2.8 13.0 15.0 7.0 LAVPUNG 66.0 36.0 14.5 3.4 7.0 13.0 12.5 NORTH SUATRA 107.0 151.0 17.7 7.1 5.9 18.0 3.0 11.4 13.0 RIAU 40.0 34.0 11.4 9.0 6.0 10.9 7.6 SOUTH SUMATRA 108.0 26.0 17.3 7.5 14.0 1.0 13.3 10.4 7.2 Rqg_nelAvwag 107.0 76.4 27.0 13.8 5.8 18.0 7.3 3.0 13.4 10.5 CENTRAL KALIAANTAN 44.0 11.4 13.4 EAST KALIMANTAN 24.0 23.5 14.8 8.8 2.0 9.7 8.9 SOUTH KALIMANTAN 74.0 17.6 7.4 18.0 11.0 14.8 4.7 WEST KALIMANTAN 55.0 44.0 13.1 4.2 3.0 2.0 12.5 8.7 7.9 ReginalAwre 51.0 37.2 14.2 6.1 3.0 7.3 11.0 12.6 7.9 CENTRAL SUIAWESI 20.3 10.2 4.6 7.0 9.6 NORTH SULAWESI 37.5 13.7 11.5 2.6 8.0 8.3 11.2 SE. SULAWESI 30.0 14.8 2.0 9.0 8.9 14.2 SOUTH SUCAWESI 100.0 34.0 12.7 6.1 49.0 12.6 5.9 8.4 Reonal Aveae 100.0 37.5 24.5 12.3 3.8 28.5 8.0 9.9 9.9 IRtANSYA 41.0 13.6 4.0 5.0 10.6 IALUKU 46.0 31.0 6.0 0.2 3.0 6.9 6.6 N.T.B. 60.0 20.4 1.8 11.0 15.3 7.5 N.T.I 20.0 8.9 6.3 8.0 8.6 TIVOR TIMUR 23.0 5.7 6.7 7.5 WESTSUUATRA 122.5 23.0 13.2 4.7 11.0 1.0 14.8 6.2 RenlAverg 76.2 27.6 11.3 3.4 11.0 3.0 8.0 1.0 10.4 7.4 GrandbW 247.0 102.6 25.7 13.3 10.5 21.2 7.1 9.0 2.3 16.71 10.5 Soutme: Mindby of He facility, has almost half of the instance, Kalimantan has the highest paramedics specified by MOH, while the number (1.04) of sanitarians per center other three class A units have far fewer. as compared to .56 in Sumatra. Class B, C and D facilities all far well Similarly, Java and Bali had the largest short of the targeted staffing levels. In number of auxiliary health workers all, class A and B general hospitals (five), while facilities in the Eastern absorb 44 percent of the paramedics Islands had fewer than 3.2 on average. working in government hospitals. Private hospitals account for a quarter of 0 There is also considerable intra-province paramedical employment in hospitals. variation in the distribution of paramedics. For example, the average * Java and Bali account for 55 percent of number of paramedical workers per the country's public hospital-based facility varied from five in Bogor paramedics; another 22 percent work in Kabupaten to 17 in Majalengka (Chart Sumatra. Class A, B, C and D facilities 4.1). Several districts appeared to be in Java and Bali have been more "overendowed" with sanitarians, dental successful than those elsewhere in workers, auxiliaries and administrators attracting paramedical staff. (Annex Charts 4A-4G). * Private hospitals employ fewer Issues and Suggestions paramedics on average than public hospitals of any class. Java and Bali 4.16 As discussed, government formasis were account for three quarters of paramedical made available to hire 2438 paramedics on employment in private facilities, average annually, during the 1989/90 - 1992/93 followed by 16 percent in Sumatra. interval. This was less than the number of new paramedical graduates during this period e The paramedical work force in hospitals (approximately 7000 per year), and far less than differs from that in health centers in the number being produced annually in publicly terms of skills mix and educational run schools (4,753). Looking ahead, it is characteristics. Hospitals have attracted doubtful whether new paramedical workers can more educated workers 62 percent of the continue to be added to the government payroll paramedic labor force has at least DI at this rate. Indeed, budgetary constraints may level skills training as compared to 20 result in sharp reductions in the number of percent for health center workers (not authorized new positions. This would further including administrative workers). A aggravate an emerging absorption problem for large share (75 percent) of puskesmas new graduates as well as existing workers, and paramedics has below-high school levels would raise questions about the capacity and of school attainment. focus of training facilities. The possible ramifications of sharp reductions in paramedical * There are province and region-wise formasis need to be reviewed. Two issues need differences in numbers of paramedics per to be addressed. The first concerns the number puskesmas. Java and Bali, with 11.3 of paramedical workers actually required in paramedics per health center, account for different facilities, and the second relates to the 54 percent of such workers, while sort of training needed to underpin emerging Sumatra, with 10 per puskesmas, staffing patterns. employed 38 percent of such workers in the early 1990s. In addition, there are pronounced differences between 4.17 Staffing Norms and Needs in provinces and regions in number of Government Hospitals. Are paramedics in specific paramedical categories. For short supply in public hospitals? A comparison 82 Chart 4.1 Total Paramedical Staff per Health Center, West Java, Indonesia, 1992 18 16 - 14- 12 C.2 8 8 E IL 6- 4- 2- 0- i l I I I - I - I - I I - I I -....I. - I --..I I I I - I I - I I I I I I - I Kod. Kod. Kab. Lebak Kab. Kab. Kab. Bekasi Kab. Subang Kab. Garut Average Bandung Sukabumi Sukabumi Kuningan District of actual staffing levels with norms established 4.20 These calculations suggest that by MOH indicates there are significant paramedics are not in short supply in shortages. For instance, on average each class government hospitals. In fact, many workers C facility had only 26 percent of the expected would be redundant if higher productivity number of paramedics in 1992. This suggests standards were applied. For instance, many of that any reductions in the number of new the 750 paramedics employed in class C and D formasis could threaten the level and quality of hospitals with low (below 30 percent) bed services provided in government hospitals. occupancy rates could be considered to be dispensable. Private hospitals seem to operate at 4.18 Before this inference is accepted, some higher efficiency standards than public hospitals additional perspectives need to be considered. in terms of their use of paramedics. And within One issue concerns the robustness of the staffing the public sector, there are hospitals in each standards that have shaped MOHs work force class which have succeeded in handling large initiatives. Why, for instance, are staffing patient volumes with below-average norms so high for paramedics at the class A complements of staff. For instance, there were (867 workers), B (385) and C (96) hospital 29 class D hospitals which used fewer levels? If paramedical skills are so important, paramedics than the average facility but handled then why do auxiliary health workers, who a comparable patient volume and achieved typically have a high school degree or less, loom slightly higher bed occupancy levels (Table 4.6). so large, e.g., 70 percent or higher, in the prescribed staffing lists? What is the rationale 4.21 In short, this discussion suggests that for the assignment, in terms of staffing reductions in government demand for standards, of 44 percent of the paramedical work paramedics, e.g., through cuts in the number of force to A and B hospitals? These questions formasi, are unlikely to undermine hospital suggest that a review of the basis for staffing services. Indeed, many facilities do not seem to norms may be timely. It may be that existing be making effective use of their current standards are too generous, and were framed complement of paramedics. Ways need to be without taking full account of budgetary found to lessen underutilization of paramedics. constraints, the high costs, included personnel This will likely require improvements in expenses, of operating hospitals and the allocation within and between facilities. In this possibility of assigning a number of tasks to respect, the process of giving hospitals greater workers with multiple skills. financial and managerial autonomy through Lembaga Swadana provides a promising 4.19 A related set of issues pertains to how opportunity. Hospital administrators need to be well existing paramedical workers are allocated. held accountable for service strategies and As was done for nurses and midwives, results, but need the latitude to build a facility regression analysis was used to determine to work force (including paramedics) through what extent the number of paramedics in appropriate hiring, firing and deployment different hospital classes was linked to outpatient decisions. Managers need to experiment, to try and inpatient case loads. Results indicate that different ways of assigning tasks, motivating patient volume at the C and D levels had only a workers and obtaining scarce skills. Contracting weak impact on the total number of paramedics out pharmacy, X-ray, laboratory and some other (Annex Table 4A). Patient "demand" played a paramedic services may be an attractive way of much larger role in determining the number of cutting overheads and improving service paramedics in private hospitals. These management. These trials should be guided by observations also applied to individual close analysis of existing experience, including paramedical worker categories, e.g., laboratory lessons from public and private facilities which analysts, X-ray technicians and assistant appear to be making more effective use of their pharmacists and nutritionists. paramedic staff (Table 4.6). 84 Table 4.6: USE OF PARAMEDICS: CHARACTERISTICS OF "EFFICiENT" /a AND "INEFFICIENT" HOSPITALS Class C Class D Pfivate Efficient Average Inefficient Efficient Average Inefficient Efficient Average Inefficient Number of beds 116 155 211 54 62 60 270 85 244 Number of 14 26 49 6 16 27 11 9 78 paramedics Number of 12 18 3 1 S 6 9 10 5 24 doctors Number of 68 88 140 27 39 47 151 48 196 nurses Outpatients per 105 168 256 87 88 96 292 101 316 day Inpatients per 157 222 300 68 67 72 354 85 254 day Bed occupancy 50 59 59 44 43 42 67 47 60 ratio Number of 21 88 18 29 116 25 16 309 32 hospitals Source: Derived from MOH data. /a Hospitals whose paramedic numbers relative to patient loads fell one standard deviation below predicted levels were considered to be more "efficient"; facilities whose staff numbers were more than one standard deviation above predicted levels were defined to be "inefficient". 4.22 Paramedical Needs in Health Centers. unsupervised settings, providing preventive, There also appear to be opportunities to health education and public health services rationalize paramedic staffing and use at the which may not be wanted or appreciated. The puskesmas level. For instance, many health tasks assigned to workers are characterized by centers are already over stocked with their variety and overlapping quality. For paramedics, at least from the perspective of instance, sanitarians are expected to provide prescribed staffing norms (Table 4.5). guidance, training and advice to communities as Moreover, those facilities with above average well as businesses, schools and other public paramedic strength have not achieved facilities in regard to environmental cleanliness, significantly higher utilization rates-this is seen obtaining and using clean water, and in the lack of significant statistical associations construction and maintenance of wells, latrines between average daily visits and the total and housing; assist the puskesmas doctor by number of paramedics, and between daily visits leading the communicable disease control team, and the number of sanitarians, auxiliary workers taking on other management functions, and or other individual categories.' contributing to other activities; support the health center's nutrition program and operations 4.23 It needs to be mentioned, though, that of the mobile clinic; and report on the work of assessment of staffing levels and allocations is village-based health and sanitation groups. hampered by the nature of the work performed Assistant nutritionists (Dls) are supposed to which makes it difficult to devise robust assess individual, family and community efficiency indicators. Most paramedical workers nutrition situations; respond to these conditions are involved in outreach work, often in by providing education and training (including 85 demonstration activities) on an individual or paramedical jobs within the health center are group basis; and record, assess and report on largely overlapping and even interchangeable. results. 4.26 In this setting, experience-based 4.24 Not only are there numerous, often open- performance standards have been slow to ended formal tasks assigned to health center emerge. This may explain why supervision paramedics, but there are also other activities methods have remained cumbersome and which fall to them as a result of the informal mechanistic, relying on targets and easily division of labor and pecking order in the observable or quantifiable indicators (see para. puskesmas. Thus, Sciortino observed that 1.30). A further complication is the dependence malaria workers helped transport other staff, the on inadequately trained auxiliary workers and sanitarian acted as the organizer and coordinator "volunteers" for many paramedical tasks. for drugs and equipment, while nurses passed along inconvenient or unpalatable outreach tasks 4.27 Improving Paramedic Performance in to paramedics including auxiliary workers and Health Centers. In principle, the response to puskesmas "volunteers" (see Box 1.1 and also possible underutilization and misallocation of para. 3.40). paramedics at the puskesmas level should be the same as that suggested for hospitals (para. 4.21). 4.25 The multiplicity of tasks, formal and Facility managers need to be held responsible infornal, performed by paramedics may for service strategy and results, but also need to account, in part, for the wide inter and intra- have the authority, resources and positive provincial variations in the availability of incentives to assemble staff with the requisite different field worker categories (Chart 4.1). capabilities. For example, health center Why would some districts and provinces be administrators should not be obligated to hire overendowed with sanitarians or laboratory sanitarians, assistant nutritionists or other technicians, and so forth, and how would these workers who happen to be in excess supply differences in staffing levels come about? The locally or because of instructions based on above discussion suggests that where there are uniform staffing norms. excessive numbers of one paramedic category, e.g., sanitarians, these workers are most likely 4.28 Some preparatory work would clearly be handling other puskesmas tasks as well, many needed before this staffing approach could be without any relationship to their formal job broadly applied. Some suggested steps include: description or training. The concentration of sanitarians or other workers in particular * Rewriting terms of reference and districts is attributable to two factors. First, qualifications for paramedic jobs to there could be a paramedical school in the underscore qualities such as flexibility, vicinity producing yearly batches of a specific good judgement, interpersonal skills, type of worker, not all of whom can find jobs in ability to handle multiple tasks and health centers in other provinces or districts2. trainability; Second, some formasi, i.e., those financed from the routine budget, must be filled during the 0 With a focus on the country's 750-1500 same fiscal year, otherwise the funds are high IMR localities, compiling and withdrawn. It is not unlikely that authorities, disseminating case studies of successful when faced with the prospect of losing positions team building and sustained and effective and perhaps informal pressures to find jobs for teamwork amongst paramedics, with unemployed but not suitably qualified graduates, special attention to how leadership was would choose to hire "extra" workers from a exercised and the use of motivational particular paramedic category. Further, this factors; practice might be abetted by a perception that 86 * Experimenting, again targeting remote underutilized. Yet there are some promising areas, with the use of special pump- examples, e.g., the activities of INA, of how a priming and troubleshooting specialist broad and constructive oversight role could be teams who can be called in by facility fashioned. Indeed, INA's approach to managers to fill skill gaps temporarily, maintaining nutrition training standards can be and to provide advice and training extended to other paramedical categories, e.g., assistance; laboratory and X-ray technicians, especially those working in more technical fields in * Developing supervision arrangements, hospitals. particularly in target areas, which in place of the checklist approach focus on 4.31 The schools producing these workers helping puskesmas staff to define work could be required to satisfy standards worked priorities and diagnose and remedy key out and upheld by relevant professional bodies, delivery problems; e.g., those with an overview of laboratory work in hospitals. Moreover, students would need to * Refashioning supervision to include pass standardized examinations and to satisfy much greater technical support and on- other criteria before being recommended by the the-job coaching; professional association for licensing by city or provincial authorities. Government agencies and * Recording and sharing experiences in professional bodies should not try to determine regard to defining measurable and the number of schools in operation, enrollment operationally useful health center goals; or tuition levels for different worker categories- and these are matters which would be resolved through market pressures operating on schools, i Encouraging health center managers to employers (hospitals), and actual and potential apply for suggested in-service training students. Different levels of government can grants for particular staff, with the intervene by offering scholarships or loans to training subsidy linked to attendance and support paramedic training for needy students. attainment of specified levels of Alternatively, provincial and district competence. governments may want to use contractual arrangements or conditional grant schemes (see 4.29 Policy Towards Paramedical Training. para. 4.28) to attract workers with particular This approach to staffing public facilities with paramedical skills. paramedics needs to be supported through a revised government stance as regards training. 4.32 Another target for public intervention Rapid growth (see Tables 4.3 and 4.4) in private should be curriculum and training process for school enrollment and graduates suggests that those workers, e.g., sanitarians and nutritionists, the government can now withdraw from the role likely to be candidates for health center jobs. of operating and directly financing schools. Earlier discussion (para 4.28) emphasized the Instead, policy should focus on assuring that importance of preparing puskesmas workers for private schools produce graduates of requisite multifaceted and open-ended assignments in quality and competence, and that able students community settings. Because of this, from poor areas and low backgrounds can afford prospective paramedic workers need to be given to enter paramedic training programs. a thorough introduction to laboratory science, statistics and various public health concerns and 4.30 To date, the government has not played options, including issues relating to nutrition and a strong regulatory role as regards paramedic the environment. These workers also need to be training. Mechanisms such as school prepared to interact effectively with local accreditation, curriculum review, examinations, community members, and to grasp and respond licensing and scholarships have been to local problems. This means that training 87 needs to not only impart a range of technical and examination and entry level qualifications for communication skills, but should do so through paramedics to bring about the needed changes in appropriate mechanisms and experiences. curriculum and the training process. Clearly, Training needs to be problem-directed and to the activities of Pusdiknakes would need to be concentrate on communication skills, and how to intensified and recast to deal with this enhanced identify what needs to be done in particular regulatory role. For instance, funds and trained localities and how to bring about behavioral staff are needed to carry out regular, thorough change (Berg, 1992). The best modality for school visits, to evaluate teaching-learning such training is field work, first as an observer processes and to systematically use feedback and subsequently as an assistant and intern, from worker performance on-the-job to redesign working with more experienced staff members. training materials and procedures. Because of the changes entailed, it would be advisable to 4.33 The government could use its regulatory begin this process in two or three provinces, instruments, e.g., accreditation of new and using lessons learned to fashion replicable existing schools, certification of students through procedures. Endnotes 1. The number per health center in each of several paramedic categories (laboratory assistants, sanitarians, nutritionists, health auxiliaries and administrative workers) was included in a specification together with doctors and population per puskesmas, location outside Java and Bali and the number of mother and child health kits per facility. The total number of paramedics was also evaluated in a similar specification. None of the paramedic variables proved to have even a marginally significant association with utilization in this formulation. In fact, a negative (but statistically insignificant) coefficient was estimated for several of the paramedical categories. 2. The overall number of paramedical workers in hospitals and health centers is strongly associated with the number of paramedic students in the same province. This relationship becomes even stronger when variables are weighted by population size. The relationship holds as well for individual worker categories (see Annex Charts 4H and 41 and Annex Table 4B). 88 CHAPTER FIVE HEALTH WORK FORCE CHALLENGES AND OPTIONS subcenters was outpacing supply. I For 5.1 Previous chapters touched on key example, MOH's authoritative manpower plan, features of Indonesia's health scene, and finalized in 1983, predicted that personnel reviewed issues specific to each major health requirements would increase by 68 percent work force category. The present chapter during Repelita IV (1984/5-1988/9), and by a provides a summary assessment of the further 76 percent and 40 percent respectively government's health work force initiatives, and during Repelitas V (1989/90 - 1993/4) and suggests ways of reformulating the goals and Repelita VI (1994/5 - 1998/9). means of work force policy. 5.4 MOH's response to this perceived 5.2 A Target-Driven Approach. To date, manpower challenge was decisive and far- GOI has pursued what has been virtually a reaching. An overarching manpower planning textbook example of manpower planning in the process was established involving different units health sector. In this approach, the "production" of MOH as well as BAKN, MENPAN and of health has been visualized in almost Bappenas. New units, i.e., Pusdiklat and mechanical, fixed input-output terms, with Pusdiknakes, to oversee training were set up specified quantities of distinctive, non- within MOH, and CHS was given greater substitutable skills, e.g., GPs, basic nurses, X- authority over medical education. New ray technicians, nutrition assistants, seen as manpower planning techniques were introduced, limiting factors in the improvement of health including the Indicator of Staffing Needs (ISN) status. Accordingly, health work force methodology. And facility-level staffing norms development has been conceptualized in and standard workloads were developed and technical rather than economic terms, with the refined, along with new allocation, deployment, central concern that of eliminating staffing gaps. supervision and performance assessment The agenda has been to put into place training procedures. Finally, as discussed in previous programs and other mechanisms that would sections, huge increases were engendered in the alleviate anticipated skill shortages and supply number of doctors (including specialists), nurses, the personnel thought necessary to operate the midwives and paramedics who were trained (or country's publicly run health delivery system. re-trained), and then hired, directly as civil servants or on contractual terms. 5.3 The voracious appetite for manpower associated the government's multi-tier, multi- 5.5 This manpower development strategy function service network became apparent in the succeeded in greatly increasing access to trained 1970s. Projected to become the principal source government health staff. In 1993, each health of health care in rural Indonesia, it quickly center had, on average, over 15 trained health became axiomatic to planners and participating workers. With over 95,000 tertiary level donors that demand for staff to serve in the workers, there was one government health emerging array of hospitals, health centers and functionary for every 2000 Indonesian. This 89 ratio did not include the 26,000 trained staff other facility-level characteristics. Specifically, working in class C and D referral hospitals and the average number of daily patient visits for all the 24,000 workers employed in class A and B health centers in a kabupaten in 1991 served as teaching hospitals. the dependent variable and various indicators of staffing, availability of instruments and location 5.6 As access improved, use of public were included as explanatory variables. Results services rose over levels prevailing in the late indicated that utilization levels varied directly 1970s. In 1992, 26 percent of respondents in a with the number of doctors working in each national household survey reported that they facility. However, the responsiveness of visited government facilities when ill-the utilization to additional physicians was found to proportion going to government hospitals and be rather low. Utilization was also positively health centers was as high as 38 percent in the associated with the availability of different Eastern Islands (Table 1.4). Moreover, daily medical instruments, with the population density use of individual facilities had reached high in the area in reach of the puskesmas, and with levels in some areas, e.g., over 80 outpatient location in Java or Bali. No relationship was visits a day to health centers in some kabupatens detected with other locational characteristics, in Java and Bali. e.g., whether the puskesmas was situated in a remote, transmigration, or in "normal" rural or 5.7 Nevertheless, despite increased access to urban areas.2 trained staff, the performance of the health system has been disappointing (see Box 5.1). 5.9 These findings were derived from cross- Utilization of public facilities and staff has been sectional data and thus may be contaminated by low, especially in areas outside Java and Bali, estimation biases due to what econometricians the quality of service provision has caused call the non-randomness of program placement. concern and it can be surmised that the health Despite this possible distortion, the results seem impacts associated with the public system have reasonable. We would expect that a second been fairly low as well. Government service doctor or additional nurses would bring modest outlets do not seem to have gained ground in the increases in puskesmas utilization since the menu last five or six years relative to other providers. of services could be expanded. But while Meanwhile, average daily patient contacts in plausible, these results do not provide much many health centers and public hospitals remain policy guidance in part because estimated levels extremely low. The large numbers of trained of responsiveness are low, but also because it is doctors, nurses, midwives and paramedics not feasible in budgetary terms to expand the staffing these underutilized facilities do not already substantial staff contingents present in appear to be fully occupied. virtually all health centers. Indeed, the full costs (including educational subsidies) to GOI of Accounting for Low Utilization and Efficiency adding a second or third doctor to the puskesmas team and/or assigning additional nurses and 5.8 These findings suggest that past health paramedics would be very high. work force initiatives have not worked as expected. What has been the problem? As 5.10 Factors Internal to the Puskesmas. discussed (paras. 1.26-1.30, 2.55-2.56), one The more pertinent policy question relates to possible explanation lies in funding shortfalls why utilization and the quality of services that have kept staff deployment to levels below remain low in facilities which are accessible and what was seen as necessary in the early 1980s well endowed with staff. In this regard, and below what is codified in facility-specific regression analysis which captures intrasample staffing norms and other parameters. This variations in utilization, staffing and so forth is proposition was explored by looking for unlikely to provide helpful clues. Instead of systematic linkages between utilization levels and dwelling on differences between facilities, 90 Box 5.1: IMPACT OF GOVERNMENT HEALTH SPENDING What has been the payoff to Indonesia's sizable investment in health delivery and training facilities and large recurrent outlays on staff salaries, drugs and other items? This question is highly pertinent in the present context of tightening budgetary constraints, sluggish IMR trends and growing recognition that past policies need to be reconsidered. Answers to this question need, however, to be framed carefully. For instance, the health consequences of urbanization, rising per capita income and educational attainment, and public policy in other sectors would need to be factored into the analysis as well. So should the possibility that some segments of the health system, e.g., health centers, have had greater impact than other components and that expenditures have inevitably taken some time to yield results. And once the underlying proposition has been framed appropriately, there remain huge challenges in regard to measuring health impacts and assembling the data needed for a credible statistical assessment. Not surprisingly, no study seems to have investigated the health effects of cumulative government spending in a comprehensive way, and few researchers have taken up the matter even in a partial or superficial fashion. In fact, the present review found only one instance in which the question was pursued using a defensible formulation and methodology.l/ This exercise combined data from the mid 1970s and 1980s on village characteristics, infrastructure and government program activities (the 1976-77 Fasilitas Desa and 1980 and 1986 Potensi Desa surveys) with 1980 census and 1985 population survey information on households. The dependent variable used in different formulations of the impacts question was the cumulative mortality rate of children born to women aged 25-29. Between 1980 and 1985, this rate fell by over 40 percent--different variable specifications were used to identify the factors, including health facilities, that may have led to this decline. The final versions, which covered 2,856 sub-districts and merged the 1980 and 1985 data sets, took account of household characteristics, e.g., the schooling of mothers and amount of land owned, as well as the proportions of villages which had health centers and other facilities. This study found that the amount of land owned per household (a wealth indicator) and the schooling level attained by mothers, aged 25-29, had significant and positive effects on child survival, while urban location had no bearing on health outcomes. Surprisingly, the presence in the village of a health center did not have the expected negative impact on child mortality. Indeed, in all specifications (including several taking account of the presence of schools and farmily planning clinics), results, although statistically insignificant, pointed to a perverse, direct relationship between child mortality and availability of a health center. This might have been expected in a single data cross-section because of the possibility that health centers were introduced first in high mortality settings. But what is noteworthy is that the result was sustained even after rigorous econometric techniques removed possible bias due to program placement. These results should not be taken as a definitive assessment of what has been the impact of expenditures on health centers. The data set needs to be expanded to include household and village information from recent years. Moreover, it would be desirable to enrich the variable list, adding further dimensions of health experience and introducing income, occupation, and private health spending, and capturing other processes and trends that lie outside the formal health system. Non-linear and/or lagged specifications also need to be explored. At the same time, these results should not be discarded. For they represent a preliminary indication, using rigorous techniques, a large data set, and a systematic approach, that expenditures on health centers have not brought substantial benefits. 1/ M.Pitt, M. Rosenzweig, and D. Gibbons, "The Determinants and Consequences of the Placement of Government Programs in Indonesia," The World Bank Economic Review, Vol. 7, No. 3, 1993, pp. 319-348. 91 attention should focus on the human resources the handbook, operational guidelines and related available normally in nearly all health centers. materials seem to describe tasks to an excessive Thus, most facilities, including those at the degree and with misplaced precision. But the lower end of the staffing spectrum, have on pedoman and petunjuk do not provide needed hand a large and diverse enough team to provide guidance on how to rank or choose between a wide range of useful services. As an example, activities, and how to prioritize tasks and use 99 percent of the health centers surveyed in available time most effectively. For example, NTB had a doctor, 91 percent had four or more the performance of nurses in numerous, separate paramedics, 84 percent had two or more curative, preventive and health education tasks is auxiliary nurses, and over 60 percent had one or carefully spelled out. Yet the amount of more administrative or other non-health workers. attention and energy a nurse should devote to each task or to each cluster of activities is not 5.11 It follows then that utilization and addressed. In practice, this has given individual effectiveness issues should not be approached workers the opportunity to carve out some primarily in terms of staffing gaps or other discretionary control over their work scope and missing elements. Instead, it is the range and flow. For instance, nurses seem to have some quality of available services and the effectiveness autonomy in deciding whether to emphasize of the associated staff tasks and activities which some activities, e.g., curative care, at the should be scrutinized. In short, the problem expense of other, possibly more important tasks likely lies within the puskesmas, in the manner and obligations. in which available inputs and skills are combined and managed, and tasks and responsibilities 5.14 These "zones" of autonomy and discharged. discretionary judgement do not necessarily undermine service delivery. Indeed, given the 5.12 This proposition that utilization and large number of tasks to be attended to, workers related manpower efficiency problems stem from need to be able (and should) be encouraged to factors internal to the health center is somewhat exercise some judgement over priority actions. surprising. This is because detailed operational However, personal interests, motives and goals, guidelines have been developed and disseminated including the opportunity to earn income through for most aspects of service delivery. The private practice, seem especially likely to come ubiquitous pedoman (handbook) and petunjuk into play when opportunities for job (operational guidelines) explicate organizational interpretation and discretionary use of time arrangements and procedures, and include job present themselves. The risk here is not just descriptions for all staff and instructions for that health center personnel reserve a large part conducting monthly staff workshops and other of their day for private work, but that decisions meetings. Technical guidelines are available for and behavior in their public duties may be individual programs, e.g., immunization, and influenced by the aim of enhancing their performance standards and targets have been unofficial income. In the process, the stated and quantified. Various team and performance of tasks and activities on public individual assessment procedures have also been account may be neglected. In this regard, instituted. Indeed, supervision events occur previous chapters have drawn attention to how frequently, while staff appear to prepare widespread and entrenched private practice is diligently for evaluations according to criteria and to what extent this phenomenon is that are well understood. intertwined with health center operations. Indeed, Chapter Two suggests that opportunities 5.13 Yet these carefully delineated instructions for private practice income have long been a and procedures have not produced the desired crucial consideration in the continuing results. Several factors seems to account for "bargaining" process between doctors and MOH observed outcomes and behavior patterns. First, 92 over the terms of employment and location of comprises only one of eight dimensions assignments. in the personal assessment, limiting its value as an evaluation tool. 5.15 Another problem is that established supervision and assessment mechanisms do not * While quantitative targets encourage seem to have provided the feedback needed to wider coverage, they may also distract improve puskesmas performance. This is likely providers from objectives which may not due to flaws in existing evaluation arrangements: be so easily observed or measured. Targets, for example, might give * Supervision and technical support visits providers an incentive to inoculate can be blunt and disruptive instruments children who do not require which are oriented to quantitative immunization or to put pressure on indicators and based on checklists kaders (village volunteers) to increase dominated by administrative and process attendance at posyandus. They may also categories. For instance, "integrated" discourage attention to important but team visits can take up to eight hours, uncounted cases or to the less noticeable most of which is spent reviewing aspects of client care, e.g., patiently registers, reports and statistics, and communicating health education interactions intended for technical messages, especially if they take extra supervision may be used primarily for time. administrative purpose. Also, when these visits are preannounced, they may * Instruments which rely on self- lead to unproductive, orchestrated, assessment, quantitative indicators and responses by staff. checklists all work to encourage staff passivity and to discourage overt * Supervision comments may dwell on initiative taking and problem solving. how to best reach specific targets, and This is unfortunate, because technical support, in the form of on-the- unanticipated problems arise frequently job training and routine appraisal of staff in new and complex health delivery performance of key tasks, may be systems. In this regard, the specificity limited. Moreover, qualitative feedback of operational manuals and guidelines may be provided verbally,3 if at all, but may actually inhibit local flexibility and is not usually conveyed in logbooks, adaptiveness. supervision records or other written forms. * Little information is collected on the quality of services provided by * The stratification and personal evaluation puskesmas staff. For instance, exercises rely heavily on self assessment, assessments ought to take note of and give too little weight to performance compliance with standard case indicators. The competitive nature of management protocols, whether these mechanisms may lead to inappropriate defacto service standards, misreporting, while self-assessment tools staff roles and work norms have may make it easier to deny the existence emerged, whether tasks and activities are of problems and the need to seek being avoided, how effectively health assistance. A health center's center personnel work in teams, and stratification score is based on a dozen patients' perceptions and reactions. or more indicators, which makes it Without such information, puskesmas unhelpful for problem diagnosis and doctors have little basis to take specific prescription. Similarly, job performance actions to improve the process of service 93 provision, and to improve worker of hospital, puskesmas and district level efficiency. innovations. For instance, there are doctors who have initiated special orientation and * Activity-specific costs, and resource use training sessions for employees, reorganized and efficiency indices are generally not work teams and reporting relationships, calculated, nor are measures of facility energized and motivated staff and brought closer level productivity and resource use. ties to local communities. Moreover, MOH This interferes with the identification of records show there are numerous facilities which activities or services which are have attracted a larger number of daily patient underutilized and thus overendowed with visits. It would be useful to determine which resources. It also makes it difficult to factors lie behind and accompany high average work out which combinations of staff and/or rising utilization rates, focusing on and other resources are more cost replicable features of management and effective. operational arrangements within these and other facilities. What management and community * The information system is over- outreach mechanisms seem to work? Which developed. Different and sometimes activities and services are most popular with overlapping reporting requirements exist local residents? What sort of people do well as for a number of services and activities. hospital, health center and district Data collection and transmittal absorb a administrators? Which puskesmas staff perform considerable amount of staff time. well as acting directors in the absence of the doctor-in-charge? What is the quality of patient- Moving to a Facility-Centered Approach staff interactions and the medical assistance that are conveyed? How efficiently and energetically 5.16 The sense of the above discussion is that do staff perform their tasks individually and as health work force policy needs to attend to teams? Which staff use discretionary powers facility-level processes. In particular, revisions effectively? Which workers are overloaded and seem to be needed in the priority setting which staff categories are underutilized? What procedures, incentives and assessment signals and incentives do staff respond to? What mechanisms, which affect the range and quality means of engaging local communities and of services and facility and staff utilization engendering interest, trust and involvement seem levels. Revised rules of the game need to be to work? What sort of staff have the best developed which give staff more formal rapport with local residents? What are examples authority and autonomy (acknowledging that of effective supervision and technical support they have already claimed informally a from staff based at the kabupaten and province considerable zone of discretionary authority), levels? while also making use of conventional motivational factors, e.g., income adequacy and 5.18 Undoubtedly, there are numerous security, and career advancement (which may examples which can be cited and documented in currently be a disruptive element), and holding response to these and related questions. For employees accountable for worthwhile but instance, the initial results of Lembaga Swadana realistic performance levels and standards. in the hospital sector may offer interesting raw material. Gathering and collating such case 5.17 The building blocks for the sort of studies on a continuing basis would be a demand-sensitive approach that is envisaged can worthwhile exercise. Examination of the be drawn from several sources. First, it would experiences of Yayasan Kesejahteraan Muslimat, be desirable to scrutinize successful district, an Islamic NGO which runs 30 clinics and hospital and health center programs within hospitals, and the accomplishments and problems Indonesia. There are many anecdotal accounts of other NGO and private ventures may also 94 yield interesting lessons. This portfolio of case districts with lagging health indicators. It is histories would provide practical guidance useful such localities which have experienced the most in guiding and training staff. And from this unsatisfactory aspects-high turnover, erratic collection, it is likely that some home-grown commitment and significant numbers of unfilled proposals for puskesmas reform would emerge. positions-of MOH's doctor allocation policy, Thus, it should be possible to gamer useful and which often register low levels of facility remedial suggestions regarding the composition, utilization. It is these high IMR (and likely high focus and management of puskesmas staff and MMR) areas which need a strong and focused their support and supervision at the district level. health service delivery effort, featuring decisive These proposed solutions could be tested in puskesmas (and hospital) leadership based on different settings. problem solving, careful priority setting and good community development work. 5.19 Along with systematic efforts to cull and build on local best practices, it is important that 5.21 The number of health centers which Indonesia looks abroad for appropriate lessons. require such special treatment will need to be After all, there are numerous other countries determined through the application of robust which are trying to develop means of sustaining selection principles. Inclusion in the targeted delivery of high quality, demanded health group should depend on epidemiological, services through dispersed facilities and staff. economic and social criteria and the availability Included here are attempts to improve service of private providers. The health centers provision in public facilities by linking all or designated as "remote or very remote" for PTT part of staff income to performance (as deployment purposes might be considered the measured by technical quality, productivity and minimum number of priority facilities.4 There patient satisfaction), and by giving user are 758 such health centers, including 92 in Java constituencies a formal role in health center and Bali, 118 in Sumatra and 548, representing planning and decision making (Frenk, 1993). 30 percent of all remaining facilities, located in Important lessons are also available from Kalimantan, Sulawesi and other Outer Islands. programs which foster private service delivery A possible upper bound estimate of the number arrangements (see Box 5.2). These NGO and of localities requiring special attention can be community-sponsored, and occasionally derived by combining the 210 remote sites in university or hospital-assisted efforts typically Java, Bali and Sumatra with a figure (1248) involve use of subsidies and per patient representing 70 percent of all Outer Island payments to encourage private providers to centers-a total of 1458 facilities. deliver services in deprived areas. Arrangements often include explicit quality 5.22 A more selective approach can be assurance mechanisms as well as innovative pursued in other, more prosperous areas. In organizational and personnel measures. these localities, greater reliance can be placed on private providers, including NGO-run facilities, 5.20 The Locus of Facility-Level Reforms. while publicly run centers concentrate on The approach just outlined amounts to a major preventive and classic public health services and experimental endeavor, aimed at gaining a larger service delivery to those without insurance payoff from the substantial investment Indonesia coverage. has made in health infrastructure and manpower. The coverage of this suggested exercise needs to 5.23 This would amount to a two-track be discussed. For example, it would be difficult approach in more dynamic and prosperous areas. and probably unnecessary to implement this First, private service provision could be approach in all of the country's more than 6000 supported and regulated through appropriate health centers. Accordingly, it seems best to initiatives. Policy measures might include focus on facilities located in poor and remote making available information on market 95 opportunities to medical students and PTT Health Work Force Implications doctors; offering loans, concessions on local taxes and registration fees, free equipment, 5.25 In summary, this paper argues that training and/or rent supplements to help Indonesia needs to enter an era of opportunistic physicians establish their practices, especially in eclecticism in its health work force initiatives. poorer localities; involving professional With public spending likely to remain associations in licensing, certification and constrained, government resources need to be continuing education matters; and paying private used, in the first instance, to make available physicians fees to perform specific tasks, e.g., trained health workers and high impact services inserting IUDs, and to conduct, daily curative in areas with lagging health indicators. The clinics at the puskesmas. In addition, insurance organizing principles for this effort should be companies can be encouraged to organize care effectiveness and sustainability. An intensive provision through networks, e.g., preferred and continuing search process is needed to providers or health maintenance organizations identify administratively tractable and financially (HMOs), of private doctors who work for affordable worker allocation and service delivery salaries and/or fees per patient or per service mechanisms, which catch on in local contract. communities and achieve lasting results. 5.24 At the same time, existing health centers 5.26 As discussed, this exploratory process would need to be reoriented so as to complement would likely lead to major departures from the and support what was being done by private rigid, uniform and not demonstrably effective actors including NGO-run clinics. The typical delivery system now in place. This ongoing puskesmas could focus on delivering a scaled- exercise could also be expected to result in a down set of curative services to the very poor, more flexible view of staffing in health centers using means testing or other selection and hospitals. It is worth noting that micro level mechanisms. Health centers should also provide health reforms often have implications for the a range of preventive health education and number of staff, the mix of skills required and public health services, with the mix varying the way personnel are allocated and according to local circumstances. There would managed-reductions in overall demand for staff probably need to be fewer health centers, and and changes in the needed skills profile are they would be located in or near poor common outcomes (see Box 5.2).6 neighborhoods and in settings well-placed to This is not surprising in view of the large carry out high impact public health initiatives. share of health budgets accounted for typically The administrators of these facilities could be by salaries, benefits and training costs. selected through an open and competitive hiring process. Those chosen would be expected to 5.27 Chapters Two, Three and Four have articulate and address an appropriate set of drawn attention to likely work force- priorities and objectives for which they would be consequences and implications of facility level held accountable. As in more remote areas, it health reforms in Indonesia. To recapitulate: would be advisable for provincial, district and municipal health authorities to experiment with * Demand for doctors in the public system alternative management and delivery may grow less rapidly as new staffing arrangements, including turning over facilities to mixes and levels are worked out within NGOs, using non-physicians as administrators, health centers and hospitals, and if and subcontracting specific services to private delivery efforts are targeted on areas practitioners.' with unfavorable health and social indicators. 96 * An easing of physician needs would * Requirements for specialist doctors in allow sharp cuts to be made in current government facilities appear to have large subsidies to public and private been overestimated because demand has medical education. This would require been tied to norms rather than utilization students to finance a much larger part of levels. Compared to projected levels, training costs. But large increases in fewer specialists are likely to be needed. tuition could affect equity and regional Hospitals need to decide on realistic and balance adversely. These concems could affordable specialist skills requirements be addressed through increases in on a per facility basis. In many cases, scholarships and student loans, with GPs, with appropriate training, should financial supports linked possibly to be able to perform tasks now assigned to career decisions, e.g., to set up medical specialists. Also, some specialist skills practice in remote areas. can be obtained on a part-time basis through fee-sharing contracts with * Lower demand for physician private doctors. Compulsory service for requirements and some focusing of specialists should be ended, and efforts on impoverished, high IMR appropriate incentive packages should be localities, would further vitiate the used to attract specialists to jobs in usefulness of centrally orchestrated remote areas. allocation arrangements, including the contract doctor scheme, as currently 0 As government demand for GPs and designed. Compulsory service for some or all specialists levels off, public doctors should be scaled back. and private universities may need to Physicians should be drawn to decrease the number of students accepted puskesmas and other jobs through a for medical training. Of course, market combination of monetary and non- signals such as tuition increases and monetary incentives. Facility managers, reductions in income streams accruing to under the supervision of provincial and doctors relative to other professions will district officials and local communities, engender some reduction in demand for should play a leading role in this medical school places. But medical process. Central support could be schools should make anticipatory channeled through the professional rural adjustments to avoid the sort of health corps, available for assignment physician oversupply that has been exclusively in peripheral areas. observed in Mexico, Egypt and several other countries. * Along with financial and other inducements to publicly hired staff, * Doctors need to be better prepared for NGOs may provide mechanisms to their managerial, public health, improve access to high quality health community development and quality services in targeted areas. Different assurance roles, and also for their likely delivery arrangements, including entry into private practice. This will networks of clinics operated by contract require enhanced mechanisms for doctors or even by doctor substitutes assuring medical education quality, such as nurse practitioners, should be including strengthening medical school tested. Demand for doctors in the public assessment, curriculum review and system would decline if these alternative student examination processes. In delivery arrangements proved effective. addition to CHS's initial assessment, medical schools need to be evaluated via ongoing accreditation mechanisms which 97 Box 5.2: PROSALUD AND THE COMMUNITy DOCTORS SCHEME Efforts to devise alternatives to public sector delivery of health services have flourished in Latin America. In Bolivia, PROSALUD, a private non-profit network of health centers has successfully delivered high quality curative and preventive services in low income urban and rural communities. Eleven of the 16 activities undertaken by PROSALUD are preventive services which are provided free of charge. Facilities are subsidized through grants from the health ministry and through free of use of buildings that belong to the Church, local communities, the Red Cross and the Social Emergency Fund. Nevertheless, PROSALUD is 92 percent self-financing thanks to fees charged for curative care and for maternal and dental services. PROSALUD is operated by a central unit which oversees planning, marketing, hiring and training personnel and the purchase and distribution of drugs. Each clinic develops an annual plan based on 55 service indicators. Clinic directors convene monthly to review service delivery and financial performance. In preparation for these meetings, the central unit makes available graphs showing progress to date as regards each facility's service goals. Clinic directors also undertake selectively a three year planning exercise each July. PROSALUD has been especially innovative in personnel matters. The employee incentive scheme includes a guaranteed base salary and a variable bonus that is paid when the clinic surpasses performance levels in the previous two quarters. Revenues generated in ten of the eleven PROSALUD cost centers are used to measure performance-the pharmacy is excluded to avoid overprescription of medicines. The clinic doctor receives 30 percent of the revenue surplus designated for bonuses, and clinic staff divide up the remaining 70 percent of the funds. Bonuses usually comprise about 10 percent of compensation. This approach encourages staff to expand the clientele served by the clinic and avoids automatic pay increases unrelated to productivity. PROSALUD has also used other means to hold down personnel costs, including fee/risk-sharing with part-time dentists, pediatricians and obstetricians/gynecologists, requiring all staff to handle a range of tasks and functions, sharing staff between clinics and altering the division of labor between GPs and other staff. Another Latin American example which deserves mention is the community doctor model which is being tested in 300 clinics by the Mexican Family Planning Foundation (MEXFAM) and by a similar entity in 40 clinics in the Dominican Republic (ADOPLAFAM). In this approach, physicians are recruited, trained and established in private clinics which offer MCH and family planning care in addition to other health services. Support includes furniture and equipment on a loan basis, technical assistance, follow-up training and supervision, materials and payment of fees per client and/or a base salary and rent supplements. Over a two year period, doctors are expected, and in most cases (90 percent for MEXFAM) have been able to generate revenues sufficient to repay loans wholly or partially, cover operating expenses and meet personal income targets. The community doctors model is managed by a central NGO unit which selects sites based on community requests, market studies and proposals from doctors, health officials and technical staff. Doctors are chosen from candidates who have applied or been recommended-preference is given to those who fit a profile of characteristics including adaptability, service orientation, optimism, self assurance, initiative, creative and entrepreneurial ability. Successful candidates sign a two year contract conveying the terms of financial, technical and material support, and the services to be provided, fees to be charged and the procedures to be followed by the doctor. After expiration of the contract, doctors can be integrated into the program permanently-they then receive regular technical support visits, subsidized materials and invitations to training activities. The NGO uses different means to promote, supervise and support clinical activities and community relations. Several variations of the basic model have emerged. Some MEXFAM clinics have established ongoing ties to medical schools and research institution-this has provided opportunities for medical students to gain valuable experience in poor communities. ADOPLAFAM has signed cooperative agreements with doctors with an established patient base under which equipment and supplies are made available, and in return clients are charged lower fees for particular services. Sources: J.L. Fiedler, "Organizational Development and Privatization: A. Bolivian Success Story," International Journal of Health Planning and Management 5(1990): 167-186; The Enterprise Program, Community Doctors Handbook, 1991, John Snow, Inc., in collaboration with Birch and Davis International, Inc., Coverdale Organization, Inc.; and John Short and Associates, Inc.; and John Snow, Inc., "Enterprise Program follow-up study," draft, 1994. 98 could also cover the effectiveness of the sufficiently tested supervision curriculum. A common national procedures. examination for all students, public and private, should be instituted after each * A National Midwifery Board, with wide phase of study-this would help establish representation and associated Provincial uniform standards. And various ways of Boards, should be established. This upgrading physician's skills after they body would provide oversight for leave medical school need to be midwifery training and overall policy strengthened. guidance. * Government demand for nurses, * A national and provincial board structure midwives and paramedics may decline as for nurses is required as well, to provide predetermined, possibly inflated staffing policy guidance on upgrading the norms are replaced by a facility-specific, profession and rationalizing its career budget-constrained determination of ladder. Nurses need to be better trained personnel needs. In the short run, some for their large curative role in public and diversion of basic nurses to become private settings. Their curative activities village midwives may be possible. But should be sanctioned legally, but also the sustainability and effectiveness of this effectively supervised and regulated. initiative, as currently designed, are in Entry level requirements for nursing and doubt. There may not be enough midwifery should be raised to the senior income-earning opportunities to support secondary degree, and SPKs should be one bidan in each village--the midwife selectively upgraded to the D3 level. may need to cover a larger area. In Curricula and learning processes should addition, those bidans placed in remote be revised for each category with a view localities may need a continuing subsidy, to preparing some students to become linked perhaps to performance or nurse practitioners and facility managers, coverage indicators. Financing such and at the S 1 level, nursing teachers and services could be part of the suggested managers. The government's direct focus on the 750-1,500 high IMR involvement in nurse training should be localities. confined to funding scholarships and loans for students from poor For individual midwives, success will backgrounds (conditional on agreement depend on perceived competence, to work for specified periods in remote judgement and interpersonal skills. But areas). In addition, limited subsidies to these qualities depend, in turn, on support D3-level and S1 training of developing effective recruitment, pre- future nursing teachers and leaders are service and in-service training and warranted. supervision. Current arrangements in this regard, which involve the * Qualities such as versatility, good compromises and exceptions usually judgement, leadership and interpersonal found in crash programs, could skills and further trainability should be undermine the credibility of the entire cultivated in training programs for initiative. A more gradual and selective paramedics likely to be employed in approach, targeted on high IMR areas, health centers. Greater attention should would allow recruitment of local be given to in-service training, including candidates, instruction by experienced on-the-job coaching, of paramedic teachers and the implementation of workers. Subsidies should be provided to health center managers conditional on 99 review of training proposals and 5.30 According primacy to labor market workers' completion of specified signals, broadly construed, implies considerable activities at acceptable levels of change in the profile and focus of government competence. work force interventions. Policy needs no longer be shaped by a command-and-control * As with nurses, the government should reflex nor driven by a view of personnel withdraw from the ownership and shortages that was in many ways overblown. In management of paramedic training retrospect, the staffing gaps which have been the programs. It should develop and use its target of nearly two decades of manpower policy regulatory powers to assure that private were an artificial phenomenon, the result of a schools produce graduates who meet superimposed, rigid and exaggerated sense of the specified standards. Public resources numbers and mix of staff required to produce could be used to provide training good health, and an inability to fund the opportunities for the poor. budgetary resources needed to fully implement this approach. The Focus of Work Force Policy 5.31 This gap-filling stance can be dispensed 5.28 Many of these predicted or recommended with. It is not necessary to rely on extensive health work force adjustments will be market- subsidization of staff training, conscription and driven. As discussed, market forces are a key compulsory job assignment, and complicated and mechanism governing the supply of health largely ineffectual facility assessment practices. workers in Indonesia. The major decisions and Likewise, detailed staffing norms for different daily behavior of health staff are profoundly facility classes are not required. In this regard, influenced by income earning opportunities. calculation of minimum performance standards This needs to be recognized and fully and by staff category based on observed service systematically exploited in public policy. A start levels, using the well known indicators of has been made in this regard in the PTT scheme staffing needs (ISN) methodology, seems at first for doctors and the bidan di desa initiative. sight to represent an advance-facility Still, much greater use could be made of administrators could base their staffing plans on incentive packages to attract staff to specific ISN formulas and the number of patients. But facilities and to improve on-the-job performance. this approach has some potential disadvantages.7 For instance, it creates pressures to overstate 5.29 Similarly, district health officials and patient loads, includes no incentives to use more health center and hospital managers should be cost-effective staffing mixes (including sub- induced to take account of price and budgetary contracting tasks to private suppliers), and does constraints, e.g., the costs of hiring staff with not allow for differences in geography and different skills and productivity, and the infrastructure that may affect staff performance. availability of subsidies, in making decisions Lastly, the current individual staff evaluation about what services to offer and staff to employ. system (based on eight or more attributes and And tuition and fees in public training facilities indicators) seems too cumbersome and arbitrary should be raised considerably. Students need to to bring about deserved changes in behavior. make education and training decisions in light of net private benefits that are in line with net 5.32 Instead of trying to force-feed and micro- social returns to pursuing health-related careers. manage health development, policy needs to be Meanwhile, universities and other training demand-responsive, selective and supportive. institutions, public and private, should be guided Interventions should aim to raise the by a realistic sense of the costs and benefits of effectiveness of the large existing health work operating programs for different health workers. force, public and private, and tackle continuing and emerging challenges. Enumerated below are 100 tasks and issues which stand out in the emerging and capacity to carry out their expanded tasks health work force policy agenda. and responsibilities. 5.33 Prepare Doctors for Private Sector 5.36 Develop Alternatives to Current Employment. In early 1995, the first batch of Delivery Arrangements. Indonesian best 924 PTT physicians will complete their practices and international experience need to be contractual assignments. Other batches will then distilled into approaches which seem worth follow at six month intervals. These "graduates" testing, at least on a pilot basis. The aim would need to be prepared to enter careers in private be to greatly improve facility operations and practice. Advice should focus on opportunities impact by examining management techniques for doctors in different locations in the country and different combinations of incentives, skills and on how to pursue possible jobs. Briefings and roles that work well. In this regard, some should be based on periodic surveys of private variants of the NGO or privately sponsored medical work in different regions, and on community doctors scheme, which is being information obtained from large private tested in the Philippines, could be evaluated. companies, insurance agencies, private hospitals And in each pilot scheme, it would be desirable and HMOs. Some needed policy actions would to gauge to what extent health workers can also promote private sector absorption of substitute for one another. Pilot exercises physicians. The first is increased cost recovery should be closely monitored, with provincial and in public facilities, which may make private district officials playing a growing role in the providers more attractive to patients and could design and implementation phases. enlarge work opportunities. The second is the orderly expansion of the coverage of health 5.37 Ensure Appropriate Representation of insurance, with networks of carefully selected Individuals from Low Income and Outer private doctors (and nurses and midwives) Island Backgrounds. To this end, various designated as the main service suppliers. initiatives, e.g., scholarships, some special preparatory courses, may be needed to widen 5.34 Identify Priority Areas for Government access to medical education and health training Health Initiatives. Available data need to be opportunities. However, these equity-enhancing analyzed to determine regions and localities measures need to be designed carefully, with an which require intensive health interventions. eye to costs and sustainability. Crash-programs Geographic targeting should be based on various should be avoided as should investment in criteria including the epidemiological profile, training facilities that may be underutilized or income levels and distribution patterns, which are expensive to run. ecological and cultural characteristics, and the availability of private practitioners in the area. 5.38 Monitor the Distribution of Staff Across Worker Categories. "Balance" is also 5.35 Require and Enable Provincial and needed in the numbers who become GPs, District Governments to Play a Larger and general or consulting specialists, or who opt for More Effective Role in Work Force Policy and other fields. Of course, market signals will play Service Provision. Local governments need to a major role in attracting or discouraging entry be held responsible for decisions and actions into different job categories, and in encouraging relating to the coverage and performance of educational institutions to make adjustments in public facilities; allocation and deployment of training capacity. But this process may take too different worker categories in the public sector; long. Market signals, i.e., net income earning provision of drugs, means of transport and staff opportunities, may be difficult to discern, housing (as needed); and supervision and distorted or counterbalanced by special interests assessment of workers and facility performance. and social valuations, e.g., the prestige Local governments will need enhanced authority associated with becoming a doctor or a 101 specialist, and may eventuate in overproduction innovative approaches to health service delivery, in different fields with unexpected and costly to put together effective accreditation and outcomes. One possible consequence, which is licensing procedures and to work out other well known in high income countries, is policies regarding health education opportunities spiraling health care costs due, in part, to the and quality, and to use various means to inform phenomenon of physician-induced demand. A the public of opportunities and prospects in the second possible result is unemployment or health field and to stimulate debate on policy underemployment of doctors. The latter options. This entity should include consequence is not only politically explosive, but representatives from concerned government can take attention away from the possibility of agencies as well as the suggested Nursing and using cheaper non-doctor staff to provide many Midwifery Boards, various professional health health services (Abel-Smith, 1986). worker associations, private universities and hospitals, insurance companies, NGOs and 5.39 Initiate a Phased Withdrawal from distinguished private citizens. The mandate of Public Ownership and Management of the NHC or equivalent oversight body should Medical Education and Health Worker include physicians, but also all other health Training Institutions. Initially, public schools workers. The NHC should largely rely on should be granted extensive autonomy in persuasive powers, and strong moral and financial, administrative and academic matters, technical credibility. However, it may need and also held accountable, through market tests statutory authority in a limited number of and budgetary and accreditation processes, for outcomes. Meanwhile, government agencies should strengthen their regulatory capacities and instruments, preparing for the handover of most if not all training establishments to private management. The possible advantages, e.g., production of workers and research in strategic fields, of direct government involvement in training should be evaluated within a larger review of the public role in post-secondary education. 5.40 Strengthen Institutional Capacity for Health Work Force Policy Making. Policy formulation and execution are currently diffused across and within several entities, i.e., MOH, MOEC, BAPPENAS, CHS, IDI and so forth. Some functions, especially those of developing and maintaining an overview of the manpower scene, are not getting sufficient attention. A National Health Council (NHC) or equivalent body, with a supporting administrative and technical secretariat, should be created to carry out strategic planning functions, to monitor and react to public and private sector employment conditions and consequences, to foster 102 matters. It should, for example, be authorized to determine whether existing medical programs are financially viable and sustainable, whether additional medical schools, public or private, are needed, and whether an appropriate number of medical specialists is being produced in different fields. 5.41 Encourage Further Development of Professional Associations. The NHC or equivalent body could help bring about the further maturation and involvement of professional health worker associations, guiding and inducing these entities to play a larger role in setting standards of work and care, exercising peer control and providing oversight and performance review functions, designing and conducting in-service and preservice training schemes and handling malpractice cases in a credible fashion. The NHC or equivalent would also need to help draw boundaries between associations, define the domain of self-regulation and provide a forum for the resolution of inter- profession disputes. Endnotes 1. The World Bank and other involved donors (USAID, WHO) also visualized manpower issues in these terms. The Bank's Staff Appraisal Report for the Second Health (Manpower Development) Project, approved in April 1985, stated that 'despite substantial increases in health manpower (albeit from a very low base), it was clear by the early 1 980s that the demand of the public health system.. .was outstripping the supply' (page 8). 2. A similar exercise with analogous results was conducted using 1991 data for 264 health centers and subcenters in Kaltim and NTB (see S. Indradjaya, work in progress). Average daily visit rates, in this sample, were positively associated with the number of doctors and other staff, and with the ratio of doctors to other facility personnel. Utilization also varied directly with the number of years the facility had been in operation and with the number of days per week that staff were available for consultation. 3. Some provinces sponsor various health center contests. In East Java, average puskesmas performance is one factor in determining the overall district health score in the competition for the Governor's trophy for health development. 4. A remote location is defined as one that is hard to reach (without regular transportation services and requiring travel of at least six hours from the nearest sub-district town) and in which health and education status, income levels and commnunity development are considerably below 103 average. Usually, the provincial governor designated areas as remote or very remote, and conveys this classification to MOH. 5. Here, it is worth noting that some provinces, e.g., East Kalimantan, have made arrangements with specialists who are paid an incentive fee of Rp. 20,000 to make a weekly four hour visit to a specified puskesmas. 6. As an example, one study of physician utilization in HMOs found that doctor requirements in the USA would fall by 40-60 percent by 2000 if the entire population was enrolled in prepaid plans, see R. Mulhausen and J. McGee, 'Physician Need: An Alternative Projection from a Study of Large, Prepaid Group Practices," Journal of the American Medical Association 1989, 261(13): 1930-1934. 7. MOH is in the process of testing a more refined form of the ISN methodology. 104 Bibliography Abel-Smith, B. "Economic Implications of Health Manpower Imbalances,' Council for International Organizations of Medical Sciences, Conference of 'Health Manpower Out of Balance: Conflicts and Prospects. Acapulco, Mexico - 7-12 September 1986. Berg, A. "The Challenges of Upgrading Quality and Increasing Relevance: Some Thoughts on Nutrition in Indonesia,' trip report, World Bank, 1992. The Enterprise Program, John Snow, Inc. in collaboration with Birch and Davis Inrnational, InC., Coverdale Organization, Inc., and John Short and Associates, Inc. Communi Doctors Handboo . Arlington, Virginia, October 1991. Fiedler, J.L. 'Organizational Development and Privatization: A Bolivian Success Story,' Intemational Journal of Health Planning and Management, Vol. 5, 1990, 167-186. Francisco, M. and W. Stinson. wAssessment of Manpower Management at the Health Center Level in Indonesia," University Research Corporation, October 1993, mimeo. Frenk, J. "The Public/Private Mix and Human Resources for Health," Health Policy and Planning; 8(4): 315-326, 1993. Gani, A., N. Kumara Rai and P. Tengko. "Multi-Professional Planning of Human Resources for Health; The Case of Indonesia." Paper prepared for World Health Organization Consultation on Methodologies for Multi-professional Policy and Planning Development of HRH, Jakarta, 21-25 June 1993. Gopalakrishna Kumar, B. "Low Mortality and High Morbidity in Kerala Reconsidered," Polation and DeveloDment Review. Vol. 19, No. 1, March 1993, pp. 103-201. Government of Indonesia, Central Bureau of Statistics, National Family Planning Coordinating Board and Ministry of Health. Demographic and Health Survey 1991, Macro International Inc., Maryland, U.S.A.: October 1992. Government of Indonesia, Ministry of Health. Basic Data on Health Centers, three volumes. Jakarta, 1993. Government of Indonesia, Ministry of Health. "Stratification Approach in Health Management System in Indonesia," Surabaya: June 1991. Government of Indonesia, Ministry of Health, Center for Education and Training. "Health Manpower Projections: The Indonesian Experience," pp. 240-255. Government of Indonesia, Ministry of Health, Center for Health Information. Indonesian Health Profile 1990. Jakarta: 1991. 105 Government of Indonesia, Ministry of Health, Center for Health Information. Indonesian Health Profile 1991. Jakarta: 1992. Hapsara. 'Health Manpower Projections: the Indonesian Experience," World Health Statistics Ouarterly 37, 1984, pp. 240-255. Heaver, R. 'Managing Primary Health Care: Implications of the Health Transition". Washington D.C.: World Bank, June 25, 1993. Indradjaya, S. Ph.D. work in progress. International Council of Nurses, "Guidelines for National Nurses' Associations and Others," Geneva, 1990. John Snow, Inc., "Enterprise Program Follow-up Study," draft, October, 1993. Kwast, B., S. Miller and C. Conroy, Management of Life Threatening Obstetrical Emergencies. Mother Care, Arlington, Virginia, n.d. Mulhausen, R., J McGee. "Physician Need - An Alternative Projection From a Study of Large, Prepaid Group Practices," Journal of the American Medical Association, April 7, 1980 - Vol. 261, No. 13, pp. 1930-1934. Neilan, K.H. "Case Study Survey of Ten Puskesmas in Lombok, NTB, Indonesia - August 18 - September 9, 1993," Project Overview and Preliminary Report, Stanford University School of Medicine: September 11, 1993. Pitt, M.M., M.R. Rosenzweig and D.M. Gibbons. "The Determination and Consequences of the Placement of Government Programs in Indonesia," The World Bank Economic Review, Vol 7, No. 3, September 1993. pp. 319-348. Rand Corporation (with Lembaga Demografi, UI). 'Indonesian Resource Mobilization Study Report on Utilization and Medical Care Facilities," August 1992. Reid, U.V. "Fifteen Year Strategic Plan for Nursing Education Indonesia, Assignment Report," World Health Organization Project: INO HRH 102, November 26, 1992 - February 28, 1993. Riley, J.C. "The Risk of Being Sick: Morbidity Trends in Four Countries," Population and Development Review, Vol. 16, No. 3, September 1990, pp. 403-429. Roemer, L., S. Wuntu and T. Heryanti. "Nursing Education in Indonesia," Development Discussion Paper No. 362. Harvard University, November 1990. Samekto, U. and K.L. Clark-Trumbull. "Safe Motherhood," Assessment of Midwifery Education and Practices. Ministry of Health; United Nations Development Programme; World Health Organization, 17 December 1991. 106 Saxenian, H. 'Indonesia - Physician and Specialist Physician Supply Issues,' Background paper, Washington D.C.: World Bank, November 1993. Schroeder, S. 'Physician Supply and the U.S. Medical Marketplace," Health Affairs, Spring 1992, pp. 235-243. Sciortino, R. Care-takers of Cure. A Study of Health Centre Nurses in Rural Central Java, Amsterdam, Jolly Publishers, 1992. Selby Smith, Dr. C. 'Economic Feasibility of Methodologies for Multiprofessional Policy and Planning Development of Human Resources for Health," World Health Organization, Consultation on Methodologies for Multiprofessional Policy and Planning Development of Human Resources for Health, Jakarta, Indonesia 21-25 June 1993. Background Paper. Shipp, P.J. "Workload Indicators of Staffing Need (WISN)," World Health Organization, Consultation on Methodologies for Multiprofessional Policy and Planning Development of Human Resources for Health, Jakarta, Indonesia 21-25 June 1993. Background Paper. Simmons, R. and C. Elias. "The Study of Client-Provider Interactions: A Review of Methodological Issues," The Population Council, Programs Division Working Papers, No. 7, 1993. Simmons, G.B., S. Koerber and R. Simmons. "Techniques for Improving Client Relations in Family Planning Programs." World Bank Staff Working Paper. Washington D.C.: World Bank, September 1992. Soepardi, J. "Assessment of Personal Medical Private Practice (Study No 3)," Indonesian Medical Association: 1993, Jakcarta, mimeo. van de Walle, D. "The Distribution of the Benefits from Social Services in Indonesia, 1978-87," World Bank Staff Working Paper. Washington D.C.: World Bank, March 1992. Wensing, M., R. Grol and A. Smits. "Quality Judgements by Patients on General Practice Care: A Literature Analysis," Social Science and Medicine, Vol. 38, No. 1, pp. 45-53, 1994. The World Bank. Indonesia - Issues in Health Planning and Budgeting. Report No. 7291-IND. February 28, 1989. The World Bank. Indonesia - Public Expenditures. Prices and the Poor. Report No. 11293-IND. August 31, 1993. The World Bank. "Staff Appraisal Report, Indonesia, Second Health (Manpower Development) Project,' Report No. 5442-IND. April 24, 1985. The World Bank. World Development Report 1993: Investing in Health, 1993. 107 Annex Table 2A Distribution of Doctors by Kabupaten in Selected Provinces, 1992 E' th IIuCant=SDoa1th 3ent1s Cod. PKalng In STir0 6 1.28 Kod. KapumV 29 4 0.83 Lb. Baito KUa 14 10 0.71 Lab. BAS 13 13 1.00 Lb. Kuta Wun&a lat 14 9 0.64 Lb. Kota WuIgI arm 30 26 0.87 Lb. Boajarmu 25 34 1.36 Kb B_fit K_al s 15 6 1.07 Lab. Banjar 23 O 0.00 Kab. Mlu Sunpi Tenph 13 11 0.Q5 Xib. HluW Sunpi Sdlt 17 9QS.3 Kab. Hulu Sunapi Uta 20 19 0.Q9 Kab. Kabdru 26 8 0.31 Lb. Taa 13 1 0.62 Kab. T*n 13 7 0.53 Kb. Tabalos 12 10 0.93 Lsb. KoaukiLUa 2210.73 LCb. MariT 2 213 0.97 12 7 ~~~~~~~~0.583 Ksb. Muni ~~1712 0.71 Kab. Buto 35~1 129 0.63 KLd. Ambo-F 16 15 0.94 Kab. HEma1m Tonlph1 13 l.OS Kab. Msluku lta 24 21 0.58 Kab. Mluku Teap1 43 35 0.81 Kab. Mdukcu Togem 28 13 0Q46 Kab. Jaysu 30 20 0.67 lC^b. Nomnfor Bart 12 7 O.S8 Kab. Manvuwar 17 12 0.71 K -.Srn 19 12 10.63 Kab. Fakc-Fak 10 8 0.80 YAb. Mfak 21 11 0.52 108 Annex Table 2A: continued Heath Cente Docto DoctorHeath Centes Lab. Jayawijaya 19 1S 0.79 Y.Ab. Paniaa 25 13 0.52 Kab. Jax WaropaI 8 6 0.75 ~~ Ilmurag _ _ _ __ _ '"" '-'': LKb. Dlii 6 7 1.17 Lib. Bob.nzu 6 4 0.67 LKb. DaEsa 6 6 1.00 KLb. Lautm 7 8 1.14 Kib. Viquequs 6 4 0.67 iLb. Maaato 6 5 0.83 Lab. Alia 4 3 0.75 Kib.Aimm S 3 0.60 iLb. Maufahi 4 2 0.50 KLb. Kovaim 6 4 0.67 rLb. ELmm 6 7 1.17 Kb. Laiduia 3 4 1.33 Lab. Ambmo 1 4 I 4 1.00 Suwrce.w Mniy of rHrIth, 1992 Halith Center Survey. 109 Annex Table 2B Doctors Employed in Hospitals, by Province, Decenber 1992. Total Gns Rdden Sp -J ava a nd B al _ _ _ _ _ _ _ _ _ _ _ _ -_ _ _ __:_ _ _ _ _ _ _ _ Jakart 3339 795 875 1669 W Java 1977 977 265 735 cAntml Java 1501 500 364 637 YogyakA 654 103 344 207 FAA Java 20M3 635 570 848 393 187 76 130 MAb it1 58 60 North SumYmU 980 359 227 394 Wen d rasz 342 153 55 134 Riau 154 81 73 Jambi 71 47 24 South Sumaua 416 142 106 16 B!ankulu 75 51 24 LA. ung 145 76 69 weld1 Kinuan 99 60 39 Cane1 KajiiU n 47 35 12 South KAlimauan 123 37 36 East KaIian| 174 101 73 Neth Sul wai 246 119 32 Cona1l SWla-us 87 59 23 Soth suvlaw m27 1t8 173 166 SE Sula_ 36 27 9 NTB 8O 52 23 NTT 75 61 14 E n rim 56 37 19 Mahau sO 53 27 irin Jaya 81 56 25 Total 13,929 5,099 3,017 5,743 scum: Mmiuy of1 110 Annex Table 2C: Fifth Round of Contract Placements, October 1993. provnce Requestd Placed Diffenaea A_cb 59 27 32 N. SUmm 60 59 1 Su__u Sa m 55 32 23 Rim 27 28 l1 Jmwai 31 26 5 M" Salman 33 32 1 Bmgu__ 23 14 9 T u. a 40 41 -1 DKl jak1a 27 42 -15 Jaws Ban 67 77 -10 Jaws Talph 110 II1 l1 Yoyakm 19 19 0 Jaws Turir 59 61 -2 Klirmnu Bam 31 24 7 Kali. mnu Tenpgh 34 19 11 Kalinuntan Selam 24 24 0 K _liaIUaz ru= 41 41 0 Nodt Suli 25 25 0 . Ias. Tazaph 19 16 3 Sul^Am SeEMAn so 49 I Sulawesi Tengpm 29 10 19 lhli 35 35 0 NTB 42 40 2 NIT 36 25 11 Mab_ __ 23 20 3 Irisn Jaya 57 22 35 ruwr ruour 59 36 22 TOW 1114 955 159 Soum: Miniauy of Health. 111 Annex Table 3A Trends in Nursing Work Force in Class A, B, C and D Hospitals, 1986-1992 NofIbwd Haab I WK I dmm Tow ~~~~~~~~~~~~~~., .. ,,,, ,,,41a ,. .. .. , .-y0 .StET....... - ,: .f.70f .: 1936 2 127 1357 235 2219 1987 2 129 1-30 224 2183 19U 2 2 144 2024 212 2380 1989 2 4 151 2059 221 2431 1990 3 11 196 2203 235 2639 192 4 13 252 2613 269 3147 8 -g.. @ : i g . -; : . -.- 1956 15 253 6106 953 7317 1937 23 353 7231 1344 391m 1913 23 491 7475 1403 9371 1939 23 642 7762 1392 9800 1990 23 747 7760 1405 9923 1992 30 19 933 $413 1537 10952 1936 s0 213 6473 1479 3165 1937 99 204 6959 1554 S717 1911 120 334 3475 1683 10542 1989 120 551 3996 1703 11251 1990 121 17 658 9245 I12 11647 1992 127 2 741 9359 1734 II13 !... >: ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.:.r .-E.: ..... ........ . - 1956 222 156 610 1403 3359 193 19S 164 5476 1153 6793 19U 153 210 5117 937 6264 1939 1 =4 264 5555 995 6814 1990 14 2 314 5741 1051 7103 1992 171 3 406 5129 1043 6536 1936 319 754 21236 4070 26060 1937 322 35 21546 4275 26676 19U 327 2 1229 23091 4235 28557 199 329 5 160Q 24372 4311 30296 1990 331 30 1915 24954 4418 31317 92 332 37 2332 25514 46U 32571 aigeMini- -- d lth- 112 Annex Table 38: Comparison of Average Number of Nures per Facility to Government Staff Norms, by Hospital Class and Province, 199211993. Diftfrence cls B Ditffernce Class C Didffrence Class D Diference Cleas A NMWI from Normrk NMW/ from Norm NMW/ from Norm NiUW/ from Norm Region I/ -Hsi pal _016981 Hospital 15610 Hospit-a-l J941 Hospiota 12-8. DKT Jakarta 1.587 -11 378 -168 76 -19 Jews Bent 369 -187 102 8 40 12 Jaws Tnrgah 327 -219 91 -3 41 13 Yogyakarta 364 -192 71 -23 74 46 Jaws Timur 1,024 -574 270 -276 83 -11 27 -1 Bali 662 106 128 34 69 41 JNv&/w 1,306 -586 337 -209 90 -4 45 17 DTAcoh 133 -413 68 -26 22 -6 Sumaterc Utcrs 468 -88 87 -7 29 1 Sumatera Bat 126 -1,472 434 -112 63 -31 41 13 Riau 90 -468 67 -27 23 -6 Jambi 67 -479 28 0 Sumaterm Sebtan 311 -236 82 -12 37 9 Bengkuhi 96 -461 36 7 Lampuna 185 -361 80 -14 36 7 Sumatr 126 -1.472 245 -201 78 -18 32 4 Kslimanten Bwrat 99 -447 80 -14 21 -7 Kalnmantan Tengah 156 82 29 1 Kaldnntan Selatan 117 -429 11 -83 37 9 Kasimantan Tlmur 131 -415 129 36 32 4 KaJknantin 116 -430 101 4 2 Sulawesi Utra 203 -343 90 -4 65 37 Sublweoe Tengsh 103 9 46 18 Sulawesl Seltan 397 -1.201 126 31 41 13 Sulawesi Tenggara 130 38 63 26 Sulawesi 397 -1,201 205 -343 109 15 45 17 NTB 206 -341 42 -52 49 21 NTT 132 38 41 13 Makiku 144 -402 112 18 28 -2 Irian Jays 126 32 51 23 Timor Timur 148 54 18 -10 Eastwn 1*nd&2 174 -372 114 10 38 9 Total_ 784 -814 268 -278 91 -3 38 10 Seur.: Derved from Minritry of HeUth Anne" Tae 3C: Difference hI Average Number of Nurses per Healh Cente (HC) and Current Staffing Norms, by Type and Province, 1992 TotWl 8 of Total PK + D3 DifflHnce Total Biden Difference Total NMW HC. Per NC from Norm Per NC tram Norm per HC Region 1/ 110:11 (3:1) DAKJ. JWuwrta 313 1.458 48.32 2.64 40.36 4.32 Jews Borat 770 4.47 -6.63 2.01 -0.09 6.48 Jews Tengsh 757 3.61 -6.39 2.50 -0.50 6.11 D.l. Yogyakwat 121 5.07 -4.93 2.31 -0.69 7.38 Jaws Tlrur 843 3.92 -6.08 2.45 -0.55 6.37 BaN 98 7.20 *2.80 4.07 1.07 11.28 Jevebai 2.902 3.90 -6.10 2.42 -0.58 6.32 D.I. Acah 157 9.04 -0.96 3.14 0.14 12.18 SumnAtw UtHe 318 8.00 -2.00 3.63 0.83 11.64 Sunatars Bret 163 7.20 -2.80 2.52 -0.48 9.72 Riau 120 6.22 -3.78 2.54 -0.46 8.76 Jambi 89 7.01 -2.99 1.52 -1.48 8.53 SumatareSohaten 223 4.87 -5.13 2.26 -0.74 7.13 Lanpung 159 6.17 -3.83 1.38 -1.62 7.55 Bengkulu 82 6.49 -3.51 1.54 -1.46 8.02 Srtrea 1,311 6.95 -3.13 2.55 -0.45 9.50 Kelimantan Bwrat 165 5.03 *4.97 2.02 -0.98 7.05 Kamntan T.ngsh 103 5.99 -4.01 2.79 -0.21 8.78 Kaimanten Selatn 176 4.75 -5.25 1.83 -1.17 6.58 KWimantan Tirur 125 5.20 -4.80 1.62 -1.38 6.82 Ka&,rantan 569 5.15 -4.85 2.01 40.99 7.17 SulawesiUtera 121 5.26 -4.74 3.11 0.11 8.37 Sulawes T.ngh 80 15.66 5.66 1.73 -1.28 17.39 Sulawesl Selatan 250 6.22 -3.78 3.50 0.50 9.72 Sulawesi Tenggara 88 7.63 -2.38 2.07 -0.93 9.69 Stdawai 539 7.64 = 2.36 2.92 .0.08 10.55 Nuse Tengger Beret 92 6.22 -3.78 2.30 -0.70 8.52 NuesTenggsraTlniur 148 7.05 -2.95 1.07 -1.93 8.11 Maluku 110 5.72 -4.28 1.95 -1.05 7.66 Idan Joys 141 10.09 0.09 0.54 -2.46 10.62 Tlmor Tlmur 69 3.12 -e.88 1.39 -1.61 4.51 Es tn m,b 21 560 6.93 -3.07 1.35 -1.65 8.28 Total 5.881 5.33 -4.67 2.35 -0.65 7.68 *WWqh_t Awwa 6orc: Dwevd from Lkiaaof HoWth date. 1/ Date presented are woighted averages. 21 Incdudee NTB. NTT. Meluku. Irlen Jo" ard rTmor rimnr. Annex Table 3D: Distrbution of Health Center (HC) Nurses and Midwives (NMWsI by Area. Population Density and Numnber of VEages. by Province. | Pouttlon TotWl Number Total Number Region Total Density Nunibr of HC Nurses Population Total Sq. Km. of Villa _______________ _I Population Area 000/Sq._Km of VIllgS c an tVpuli per NMW Per NMW per NMW D.K.I. Jaat 10,900,000 55.390 197 2,261 1.352 8,062 40.97 1.67 Jewa Berat 17.500.000 44,176 396 4,517 4,992 3,506 8.86 0.90 Jaws Tongah 28.400,000 34.503 823 8,796 4,623 6,143 7.46 1.90 D.l. Yogyakana 4.048.710 3.142 1,289 300 893 4,534 3.52 0.90 Jaws Timur 30.100,000 47.921 628 8,190 5,369 5,606 8.93 1.53 eal 2.722.218 5.632 483 625 1.105 2,464 5.10 0.57 J&v~ 93.670,928 190,764 491 25,188 18,334 6.109 1Z47 1.37 D.I. Acch 3.304,265 55,390 60 5,631 1,912 1,728 28.97 2.95 Sumaters Utara 11,500.000 71.680 160 3.192 3,700 3,103 19.37 0.86 SumaterB Swat 3.964.562 42.297 94 3,162 1.584 2,503 26.70 1.99 Riau 3,179.233 94,561 34 776 1.051 3.025 89.97 0.74 Jambi 1.730.844 53.438 32 1.173 759 2.280 70.40 1.55 Sumatere Saatan 6.410,793 109.254 59 1.373 1.589 4,034 68.76 1.24 Lampung 7,750.118 35.376 219 1.081 1.200 6,458 29.48 0.90 Benakuhl 1.160.453 19.786 69 1.159 658 1,764 30.07 1.76 Swnnb. 39,000268 481,780 81 18J137 12,453 3.132 38.69 1.46 Kanimntan Brat 4,984.227 211.400 24 128 1.163 4.286 181.77 0.11 Kalimntn Tengsh 3.581.197 153.800 23 1.128 904 36912 170.13 1.26 KamantnSaldatn 2.301.682 36,985 62 2,021 1.158 1.93U 31.94 1.75 Kamentan Tmur 1 797 137 211.400 9 1,069 853 2,107 247.83 1.25 Kmmtn 12664293 613.585 21 4,346 4,078 3.106 150.46 1.07 Sulaweai Utate 3.467.686 25.786 134 987 1.013 3,423 25.46 0.97 Suwedl Tngeh 1,633780 68.003 24 1,262 1.391 1.178 48.89 0.91 Sutlweal Selaten 5.180,089 62.482 33 1.176 2.431 2.131 25.70 0.48 Sulawedl Tenggara 1.377.625 68,003 20 333 853 1,615 79.72 0.39 SWaeubw 11,664,160 224,274 262 3,758 5,688 2,061 39.43 0.66 Nuse Tenggare Beret 2.981,161 20,153 148 601 784 3.803 25.71 0.77 Nua T*nggrs Tlmur 3.408.129 47.389 72 1,274 1,201 2,838 39.46 1.06 Mdulw 1,712.230 85.728 20 809 843 2.031 101.69 0.96 Iran Jays 3.244.793 419.660 a 936 1.498 2.166 280.16 0.62 TimorTimur 1,087,438 14.615 74 649 311 3.497 46.99. 2.09 Eastrn band 12,433,751 587.545 21 4,269 4,637 2,681 126.71 0.92 Tota _ __ 169,433,400 2_097,948 81 S5,698 45,190 3,749 46.43 1.23 Source: Derved from Minstry of Heath data. Annex Table 3E: Nunber of Bidan di Desa (BdD) Graduates, (1989/90-19921931 and Remaining Target for Training, 1993194-1996/1997 NO. Vasa Eeadd found at Expected Found at NA. det we nal NO. NO. NO. Ne. at mwe No. in No. at VW* MOW No. of and wiut Tr_nd Trained Traie V_ e Lal ft..a Trinn Vage Le" Pemewt wk 8Am SdD needed I'm heeth cener 69100 01S 01192 80912 91192 0mp Oum 02193 02193 929 Dwp Out 19921199 te twech moo Tra III 121 131 11+2.31 f4J (6) 1+2+3+ 10 171 aeIleae DICTJakaa 0 21 0 0 21 100.00 0 21 0 100.00 0 0 Jew. swat e.489 672 469 070 1.600 1.77Z 1.71 640 2.640 2,61 1.17 2.012 3,467 Jews Teomeh 6.736 756 640 472 1,.37 1.504 10.44 640 2.707 2,34 13.41 flow 6.068 Yogyakat 161 124 63 0 177 77 56.50 0 177 77 56.60 74 87 Jaws ThIur 6,314 673 656 661 1.080 1.616 13.34 680 2.660 2.296 10.23 2.192 4,122 Bel 413 146 35 79 262 296 -12.60 60 342 376 -0.65 332 el Subtotal 19.003 2.303 1.743 1.680 6.016 5.272 12.37 2.440 6.456 7.712 6.60 6.276 12.861 0 DTAceh 6,220 240 234 191 666 397 40.30 400 1,066 797 26.16 783 4,437 Sumee Uter 4.030 254 634 657 1.446 1,365 0.16 020 2.365 2.276 3.76 2.194 2.638 Suater et 2.802 200 200 76 479 389 10.79 320 709 709 11.26 54o 2.334 Riu 70 96 74 40 210 196 7.14 s0 290 276 6.17 184 6s6 mb. 1.162 164 67 39 270 264 2.22 160 430 424 1.40 397 7865 Sumreter Slt_an 2.086 267 166 16l 604 6"6 6.20 320 924 885 4.11 832 1,264 Seniku*a 604 146 39 40 226 141 37.33 60 306 221 27.54 216 668 tEwpung 1.491 157 76 1l1 361 116 66.05 120 471 236 49.69 234 1,257 Subtetal 10.371 1.624 1.382 1.343 4.249 .424 19.42 2.400 6.640 6824 12.41 5,392 13.979 Kaknentan eret 4.364 177 64 34 266 276 -3.77 60 346 355 -2.90 332 4.032 Kabnata Tenah 9006 so 9 6 6 326 -243.18 120 215 444 -107.44 368 56 Kal_etun Selman 1.994 222 143 64 429 223 48.02 120 649 343 37.62 337 1.657 Kelmantln Trmw 742 106 37 37 160 44 75.56 so 260 124 62.31 Flo 632 Subteul 6.006 65 243 141 see 864 10.42 400 1.360 1.268 7.38 1,167 6.839 Sulwew aUtr 9a36 104 114 66 206 301 -6.24 120 406 421 -3.69 372 664 Sulawesi Tenuh 1.163 t6 60 30 192 169 17.19 120 312 279 10.60 242 911 Sulewee Ser m 706 260 152 152 664 369 36.82 200 784 669 27.42 642 163 Sul_weal Tlenawe 597 32 32 34 38 109 -11.22 60 17a 189 -6.18 178 410 Subtetn 3.391 512 356 200 1.160 93* 10.14 520 1.680 1.458 13.21 1,334 2.067 NTB 378 123 4 67 194 let 6.70 120 314 301 4.14 284 94 NTT 1.431 174 13 42 220 416 41.60 160 389 670 -46.07 560 66l Maluku 1.495 143 73 38 264 264 0.00 160 414 414 0.00 397 1.098 MIan Joy 662 161 0 0 I1I 177 -17.22 120 271 297 .9.69 290 362 TImor"ltrW 303 60 16 14 60 138 -72.60 60 160 218 -36.26 198 106 subrt.el 4.2ss 641 106 161 900 1.166 -26.41 640 1.648 f,8o0 -16.67 1,719 2.640 Total 64.120 6.665 3.832 3.816 1&.302 51.660 12.28 0.400 19.702 l8.o6a 8.29 16.0so 38.230 Soure: Keglatan PFndidikn Bidan Tahun 199311994. DNlUgkungen Puet ndiWUan Tones Keeeheten. DepKea. October. 1993 Annex Table 3F: Distribution of Health Center (HCs) Nurses and Midwives (MWs) in Central Java Total Difference Total MWs Difference Total SPK& SPK +D3 per from Norm per HC from Norm District Name D3 HC (10:1) (3:1) (3:1) Kod. Megelang 29 4.83 -5.17 1.20 -1.80 Kod. Pekalongan 25 3.13 -6.88 1.57 -1.43 Kod. Teogal 26 3.71 -6.29 3.43 0.43 Kod. Semarang 70 2.41 -7.59 3.28 0.28 Kod. Salatiga 26 5.20 4.80 2.80 -0.20 Kod. Surakarta 51 4.64 -5.36 4.36 1.36 Kab. Banvumas 106 3.12 -6.88 3.10 0.10 Kab. Purbalingga 77 3.85 -6.15 2.69 -0.31 Kab. Cilacap 135 4.82 -5.18 6.10 3.10 Kab. Banjamegara 79 2.39 -7.61 1.86 -1.14 Kab. Magelang 98 3.63 -6.37 2.09 -0.91 Kab. Temanggung 67 3.94 -6.06 3.20 0.20 Kab. Wonosobo 24 1.33 -8.67 3.78 0.78 Kab. Purworejo 111 5.05 -4.95 2.89 -0.11 Kab. Kesumen 103 3.32 -6.68 2.13 -0.87 Kab.Pekalongan 84 4.00 -6.00 4.94 1.94 Kab. Pemalang 74 2.47 -7.53 2.30 40.70 Kab. Teogal 85 3.54 -6.46 2.81 -0.19 Kab. Brebes 114 3.35 -6.65 3.79 0.79 Kab. Semarang 65 2.60 -7.40 2.85 -0.15 Kab. Kendal 90 3.91 -6.09 3.00 0.00 Kab. Demak 54 2.70 -7.30 2.11 -0.89 Kab. Grobogan 91 3.37 -6.63 2.04 -0.96 Kab. Pati 101 3.61 -6.39 1.72 -1.28 Kab. Jepara 82 4.10 -5.90 2.75 -0.25 Kab. Rembang 60 4.00 -6.00 2.27 -0.73 Kab. Blora B8 3.09 -6.91 2.31 -0.69 Kab. Kudus 61 4.36 -5.64 3.33 0.33 Kab. Klaten 110 3.67 -6.33 3.92 0.92 Kab. Boyolali 96 4.00 -6.00 3.48 0.48 Kab. Sragen 108 4.32 -5.68 3.52 0.52 Kab. Sukoharjo 90 4.74 -5.26 2.78 -0.22 Kab. Karanganyar 79 3.95 -6.05 5.46 2.46 Kab. Wonogiri 112 3.61 -6.39 1.89 -1.11 Kab. Batang 80 4.00 -6.00 2.41 -0.59 Source: Ministry of Health 117 Annex Table 3G: Distribution of Health Center (HC) Nurses and Midwives (MWs) In Kalimantan Barat Province Total SPK Difference Total NWs Total SPK + 03 per from Norm per HC Difference from District Name & D3 HC (10:1) (3:1) Total MWs Norm.(3:1) Kod. Pontianak 15 51 3.00 -7.00 2.53 -0.47 Kab. Pontianak 107 141 3.92 -6.08 2.61 -0.39 Kab. Sambas 127 150 6.25 -3.75 3.33 0.33 Kab. Ketapang 75 124 5.64 -4.36 0.77 -2.23 Kab. Sanggau 104 146 5.41 -4.59 1.78 -1.22 Kab. Sintang 108 117 4.68 -5.32 1.68 -1.32 Kab. Kapuas Hul 70 101 5.32 4.68 0.47 -2.53 Source: Ministry of Health 118 Annex Table 3H: Distribution of Health Center (HCs) Nurses and Midwives (MWs) in Sumatra Barat Total SPK + D3 Difference from Total MWs per Difference from District Name Total SPK & D3 per HC Norm 110:11 HC (3:1) Norm (3:11 Kod. Padang 153 9.00 -1.00 4.94 1.94 Kod. Padangpanjang 17 8.50 -1.50 3.00 0.00 Kod. Bukittinggl 32 8.00 -2.00 1.75 -1.25 Kod. Payakumbuh 17 4.25 -5.75 4.25 1.25 Kod. Solok 29 14.50 4.50 1.50 -1.50 Kod. Sawahlunto 36 7.20 -2.80 0.80 -2.20 Kab. Pasaman 97 5.71 -4.29 1.41 -1.59 Kab. Pariaman 138 6.57 -3.43 2.14 -0.86 Kab. Agam 120 6.67 -3.33 2.67 -0.33 Kab. Unapuluh Koto 41 2.73 -7.27 1.93 -1.07 Kab. Solok 159 6.91 -3.09 1.78 -1.22 Kab. Tanahoatar 87 4.58 -5.42 2.21 -0.79 Kab. Sawahlunto Sajunjung 101 5.94 -4.06 1.35 -1.65 Kab. Pesisir Selatan 146 12.17 2.17 3.17 0.17 Source: Ministry of Health Annex Ta 24: .rtthmd __F_ _ ___ k________ 5k .W A *. M gm M A, . MU MA 121 121 141 16 161 DTA..h 1 34 20D03 4% is 6% 12% sumats.. Ut.s. 360 52.400 18% 23% 27% 30% sm.smeeSa,t 2.46 3e.611 11% 14% Is% 21% Riau 6.16 1 2.424 21% 32% 43% 53% .iemb4 3.26 4,1H6 7% f6% 21% 2% Suma.. .lten 6.26 76.721 17% 21% 38% 46% 6.ng&s*u 2.96 42,764 6% 16% 20% 26% Lampuna 6.67 130.006 23% 41% s6% 76% RdhndAe.. 3.43 61.3s6 12% 18% 24% 30% Jw. s .t o.00 136.044 14% 36% 66% 77% JewsTagih 6.16 62.736 11% 26% 26% 62% Jaw. turn, 6.61 103,567 17% 31% 44% 5u% $84 .37 126,531 46% 62% 61% 73% RegisaaAt. 7." 10S.26 16% 31% 46% 62% K.bm.ntmn,B,rt 1.61 22e648 3% 8S IIS 13% KaI.tan TmTeA 2.6 44,717 6% 16% 21% 27% Ksb,,aIan kI 1.64 2.064 5% 6% 13% 17% 0 KuIt ebntn r 2.01 43646 161% 21% 23% 26% R.g' Am 1.61 21700 5% 10% 13% 17% 6Wiwms Utw. 3.61 62,32 11% 16% 24% 29% mT _msiTangh 3.00 46.022 6% 15% 21% 27% sdsw.. _.1st 14.96 2243607 32 103% 122% 141% 6wmoi r_.nwa 5.36 00.176 6% 22% 34% 47% Rmg' A,. 6.04 60.602 14% 36% 46% 65% NTs 16." 2ff.187 16% 76% 124% 172% NTT 4.6 74.871 S% 24% 36% 46% MAkhu 2.36 36.211 5% 12% 16% 21% klan Jo" 4.67 .o1 6% 326% 44% 46% r Tbmw 6.61 63,323 e6S 0% 64% u6% __givdAv. 6.30 60.333 7% 2% 38% 49% O waA 4.66 73.128 3% 24% 33% 42% Annex Tl 31: Potental Monthly Slden hnn orom At twne Bith.. As Pwroent of Expectd PubNo Sel8Y. 1993 7"O lseW, A-. cm I.W E. a.. see on"W. _ by t m by Z A h D _Am W. by r nwil ^PWA $#.W bub &W _o oam av, _w MA Reuse __e oil OTAge 2375.063 5.220 651 23.10 32.882 is 62.3 52.036 1.0 337 33.2 27.762 23 2.343 100 sUw,wswe Ut... 6,61J.476 4,626 1.36, 30.70 202.024 42 26.4 73.904 3.2 6.467 61.1 103.710 n.2 13.673 100 Simem am" 3.130. 64 2.332 1.10? 26.60 34.72 2n 46.3 3,366 1.3 1.e26 43.3 42.260 2.2 1.312 100 Riau &2.2.332 *70 2.567 26.30 64.327 74 33.6 31B.40 0.3 222 3S.7 24.n 1.1 7"a 100 Jamb&i 1.681327 1,182 1.334 23.40 46J4J1 36 72.1 23.752 1.1 611 2327 11.013 2.6 1.206 100 3m.tes Sedetan 4.438.453 2.066 2.129 26.60 131.376 63 56.2 73.036 2.2 3.022 27.7 4.5630 3.6 4.130 100 _ensksi * 34.766 *04 1.163 21.30 27,606 J 4 U 4.2 14,0 2.2 06 37.4 10,237 5.6 1.623 100 t _nvm 5.6.26e8.72 1.431 2.526 25.50 166.076 104 66.4 106.071 0.2 486 30.0 406523 1.3 2.016 100 M p' Am 27.120.733 12.371 1,400 29.3. 730.365 41 u 4.2 4322.31 1.7 13.737 30.7 316.051 4.2 33.269 IOD Jews swat 23.170.307 6,463 4,237 25.6o 650.843 106 0.3 474.447 1.0 o 6.01 13.3 106,124 0.4 2.363 100 Jews Tegh 20.622.247 8.736 3.0D1 24.00 4n,.734 74 76.3 262.2 2.7 13.403 16.7 3".44 1.1 6.467 100 Jews Thwn 23.671.667 4,314 3.723 22.20 523.233 63 66.4 347.464 2.7 10.3e2 29.1 162.277 0.7 32.63 100 3am 2.043 113 413 4.947 20.30 41,475 100 21.7 3,000 3.7 1,636 62.7 26.006 11.3 4.336 100 M1 1' i Am 63.607.230 16.322 3.067 23.76 t.655.341 37 12.3 1.213.203 3.3 40.210 23.2 364.354 1.0 16.469 101 Kemnienta erat 2,632.377 4.364 534 30.50 7n.067 13 64.7 61.167 0.7 663 24.7 13,530 3.3 7.323 100 Kulmetn Teieh t 1.10.62 SW 1.270 23.20 32.447 26 47.7 21.6"7 0.3 a7 16.6 6.300 12.3 3.331 100 Kanten Seleten l."32.67 1.34 360 24.50 46.236 22 63.1 32.066 O0 27o 27.4 12.712 2.6 1.1l0 100 Keenten T~ 353.603 742 1.293 27.00 25,06 36 20.6 7.4n7 0.7 III 66.4 16,044 5.0 1,26 100 Au_mdAw. 6.53e.24J 6.006 824 27.67 13J.816 22 61.3 112.670 0.6 1.111 30.2 65.646 7.6 14,274 100 SuiW.eiUte 1.13.161 932 2.044 20.60 33.411 42 53.6 21,124 4.3 1.836 37.6 14.313 4.2 1,656 100 6Sd_ T _ph 1.422.1t 1. I,3 1,233 26.20 41.52U 36 67.7 23.113 0.3 126 1.60 6.140 12.3 6.106 100 oulewe Selen 3.296.146 7 7.611 23.10 126U.54 I3 4.5 66.061 0.2 252 23.4 37.207 26.3 34.043 100 Wh_tvow Low . 1T116,643 so 1.676 34.20 3,26_2 64 61.J 31.323 0.0 0 12.6 5,204 4.3 1,647 100 Mh1_ Aw 3.160.142 23.21 2.375 25.21 246.73 72 66.2 136.613 0.6 2.072 26.6 65.446 17.3 42.4U3 100 NTI 2.73.6616 373 7.360 32.00 63.266 236 35.2 76.063 0.4 367 ".6 3.566 4.6 4.285 100 NTT 2.L66.077 1.431 2.024 2n.60 *6.712 so 70.7 0.36" 0.3 436 14.2 12.171 14.0 12.000 100 makAm 11.43.643 1.436 1.002 26.10 42.112 23 62.8 26.910 0.6 373 2.4 11,113 3. 3,706 100 mmen i 1,233.366 662 l.33 30.30 36.121 so 23.1 12.021 0.3 114 13.7 7,130 46.3 17.663 99 r6 T en933e 303 2n275 26.30 20D136 67 24.1 4.368 0.3 *t 12.7 2,656 62.6 12 664 100 A_s.Am 3.107.137 4.253 2.163 30.24 275.415 06 65.7 131,060 0.6 1.5t7 13.1 41.53 1t.4 30,7 100 Owv'aUAv. U#DUU8U 63.369 2.264 263.3 3156.726 Oa *6.7 2.076.070 1.3 53.354 27.4 863.453 3.0 167.002 Annex Table 3J Accounting for Nurse Staffing in C. D and Private Hospitals PKC = 33.01 + 0.01 OP + 0.004 IP, R.quaid U .28 (5-18) (2.2 (4.13) PKD 14.3 + 0.06 OP + 0.011 IP, R4quaed - 0.34 (7.52) (2.43) (5.71) PK -.1204 + 0.16 OP + 0.013 IP. R-equated * 0.78 (-6.46) (11.54) (21.) AC i6.0 0 .0 OP + 0.000002 IP. Rf-quatd - 0.13 (6.60 (.0.5) (3.52) AD 26 . 0.002 OP + 0.000002 IP. Requatd Q 0.12 (6.21) (40.41) (3.66 Ap 2.55 + 0.029 OP + 0.000001 IP, Requt - 0.13 (0.U) (1.87) (4.01) Bc 8.31 + 0.024 OP + 0.000003 IP, R-equaed - 0.19 (6.23) (2.92) (2.36) BD 4.08 + 0.019 OP + 0.000006 IP, Requatd - 0.11 (6.69) (2.76) (1.73) B, 3. + 0.016 OP + 0.000007 IF, R-eusamd - 0.39 (7.41) (4.3 (3.26) [t - suastis ae in parenthes. PKc, PlC0 and P1C, refer to the number of basic nurses, Ac, Am, and A. to the numbe of D3 nunse, and Be, BD and B. to the number of midwives in cla C, D and pnvae facilitie; OP and IP se outpatent visits and inpatent days respectively. 122 Annex Table 3K Use of Nurses: Characteristics of 'Efficient' and 'Inefficient' Hospitals Clias C Class D Private Efficient Avenge Inefficient Efficient Average Inefficient Efficient Average Ilefficient Numberof 163 146 204 76 54 76 169 68 229 beds Numbersof 7 5 8 2 2 4 10 2 31 D3s _ _ _ _ _ _ Nuraber of PKR 40 66 139 17 27 54 58 22 200 Number of 9 14 19 4 6 9 10 4 18 bidans NumberofGPs 9 8 13 4 3 4 5 2 11 Number of 19 18 24 7 6 7 11 3 20 doctors Parmedics 22 24 35 10 13 18 27 6 37 Outpatients per 196 159 190 108 84 85 236 71 240 day Inpeatients per 293 196 245 100 56 73 220 55 240 day _ Bed occupacy 70 57 54 46 43 37 66 43 56 rmtio Lengthof(iy 4 6 5 4 5 4 6 43 55 Numberbf 12 102 16 22 129 24 52 429 35 hospitals _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ Source: Derived from MOH dat. 123 Annex Table 4A Accounting for Paramedical Staffing in C, D and Private Hospitals Pmc = 13.84 + 0.05 OP + 0.0004 IP, R-equd 0.29 (7.30) (5.12) - (1.3) PM, = 8g8 + 0.015 OP + 0.002 IP. R-equed 0 020 (8.72) (126) (427) PM, = 2AS + 0.3SOP + 0.004 iU, R-equaed - 0.2 (-2.84) (5.39) (13.05) = = 1.36 + 0.12 OP + 0.00000004 IP, Rs"ud - 0.17 (2.59) (3.59) (0.34) LD = 2.16 - 0.0006 OP + 0.000000002 IP, R4awd O AO (7.59) (O ) (1=) L LS4 + 0.005 OP + 0.000000004 IP, Rt-equad 0 (6.70) (4.43) (14.15) PH0 = 3.0 + 0.0020P + 0.000000009U, R4quad - 0.12 (8.80) (109) (2.72) PHD = Ln + 0.003 OP + 0.000000004 P. R-eqd - 0.17 (8.70) (1.35) (3.74) PH, = L72 + 0.008 OP + 0.0000004 IP. R-equad _ 0.1 (5.77) (4.93) (10.5) XC = OJ4 + 0.002 OP + 0.00000002 IP, R-equed * 0.24 (6.10) (3.41) (1.55) X4 = L21 + 0.003 OP + 0.000000002 IF, Rquared - (4.87) (3.57) (0.7) Nc = L94 + 0.0030P - 0.000000005 P, R-equaed . 0.14 (7.96) (2.06) (-1.86) No = 1.64 - 0.001 OP + 0.000000001W P, Requad - 0.02 (9.91) (.0.64) (1.53) Np = L35 + 0.0820P - 0.ooooo00OS P, Rqud - 0.06 (11.5S) 2.45) (-4.M) [t - staista ue in parenthem. OP and IP refer to the number of outpaent visita and inpant days repcpdvely; PMc, PMO, PMp refer to the total number of paramedics, 1., Lg, and L to the number of labotory analysts, PHc, PH0 and PH, to the number of assitat phumcists, X0, XD and Xp to the number of x-ray tchnicians, and Nc, ND and Np to the number of assistant nutritionist in clam C, D and privato hospitali 124 Annex Table 4B: Accounting for Province-wise Variation in Paramedic Staffing Correlation between paramedical staff and students in each province = 0.93 PStaff = 339 + 1.062867Pstudents, R-Squared = 0.83 (10.87) D3AKZI = 7.76 + 0.514474D3, R-Squared = 0.59 t- Statistics are in parentheses. PStaff refers to all paramedical staff currently employed in public facilities in each province. Pstudents refers to all paramedical students in each province in 1993. D3AKZI refers to all the D3 level nutritionists working public hospitals and health centers in each province. D3Graduates refers to all D3 level nutritionists graduates in 1993 by province. 125 9ZI Sanitarians (D3)/HC o 0 - 0 0 9 0 0 6 6 L j k ) i O c1 0 a' 0 Cn 0 01 Kod. Bandung Kod. Cirubon _ ) Kod. Bogor Kod. Sukaburnl Kab. Pandeglang Kab. Sarong B Kab. Lobak _ Kab. Tongoorang K b. Bogora Kab. Sukaburni l K b CUonjur XIi Kab Cib rebon rr KsKb. KuMninwn.C Kob. norwnew 0 > K b. KWaingko C- ib CL Kab. BSuang Kab. KarBwang Kmb.~~~~~~~~~~~~~~~~~~~~C K b. Gorutt Kob Taoyo En -Kab CUwrnb Average_ Annex Chan 48 Sanitarlans (D11) per Health Center, West Java, Indonesia, 1992 1.20 - 1.00. 0.80D P c" 0.60- (00.40- 0.201 0.00~~~~~~~~~~a 8ZI Pharmaciats/HC . 0 .~ I.. .A .a .a .~ . C *I 0 01n 0. 0m O 0 o 0 Kod. Bandung Kod. Cirebon Kod. BooorJ_ Kod. Sukabumnl _ Kab. Pandeglang Kab. Serang O Kab. Lobsk Kab. Tangoerang Kob. Bogor Kab. Sukabumni >_ Kab. Cianur KS Kab. Cirebon , r *~~~ Kab. Kuningan S Kgb. Inoramayu m Kab. Majoalngko 0 Kgb. Bokasi _ Kab. Kwrswing__C11 Kgb. Purwakarts 0. Kgb. Subang 0 Kgb. Bandung S Kgb. Sumoddagin Kab. Garutn Kgb. Tasikmalaya Kab. Ciamis Average Annex Chart 40 Administrators per Health Center, West Java. Indonesia, 1992 8.00- 7.00 6.00- ~5.00- 0 *4.00- E 3.00- 2.00- 1.00- 0.00- LiB ooo 11'sw| w i, ! -, ,isi Distrit Lab Technicians/HC o o .0 0 0 0 0 Ti O - M ~w 0p 0 0 l ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~ I I Kod. Bandung r Kod. Cir bon 0 Kod. Bogor _ Kod. Sukabumi ___ Kab. Pmndaglang 05 Kab. Serang Kab. Labak Kab. Tanggorang Kab. Bogor Kab. Sukabumi > Kab. Cibnjur X Kab. Cirobon 0 Kab. Kunmngan Kab. Inoremayu m Kab. Maja.ngka Kab. Bekesi L Kab. Karawang * Kab. Purwakawt _ Kab. Subang 0 Kab. Bendung _ Kab. Sumedang Kab. Gwut t Kab. Tesikmalaya Keb. Ciamis Average Dental WorkerslHC o 00 0 o o -A o o* o~ o o o o o I 0 0 0 0 0 0 0 0 Kod. Bandung Kod. Cirabonin Kod. Bogor Kod. Sukabumi Kab. Pandegnngvin Kab. Serann Kab. Lebaki Kab. Tanggarang - Kab. Bogor V Kab. Sukabumi Kab. Cbnj4ur J:XI C,~~~~~~~I Kab. CirSbon t Kab. Kuning anin n S~~~~~~~~~~~~~~~~~~~~~~~~~~~~I Kab. Inoramayu 0 Kab. Majalangka C D) Kab. Bekasi i1 Kab. Karawangm Kab. Purwakart_ Kab- Subang Kab.~~~~~~~~~~~~~~~~~~~~~~~~~~ Kab. Bandung_ Kab. Sumedangimd Cn Kab. Garut Kab. Tasikmalaya - Kab. Caamism_ Average -4 Annex Chart 4G Auxiliary Health Workers per Health Center, West Java, Indonesia, 1992 t~~~~~~~~ ~ ffl i i X 3~ X~ *0 U i n U C 2 D*~ D . 05Xxg 10-ct 0-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~. District Staff/Students 0w o n 0 0o0 n 0 00 0 o0o0 0 0 0 r~~~~~~~~~~~ 8 ACEH _G N. SUMATRA _-_ CT W. SUMATRA _ RIAU O0 JAMBI En LAMPUNG _ DKI JAKARTA WEST JAVA --- CENTRAL JAVA YOGYAKARTA 19 EAST JAVA___3o 51 C, W ~~~N.T.B._ N.T.I109 p TIMOR TIMURE W. KAUMANTAN _ C. KAUMANTAN CL 0. S. KAUMANAN T N_ E. KALIMAHA NS N. SULAWESI _ C. SULAWESI to S. SULAWESI - S.E. SULAWESI MALUKU IRIANJAYA 1- a I. O 1- t 1 Staff/Graduates 0~~~~~ I I I ACEH N. SUMATRA__ W. SUMATRA RIAU JAMBI S. SUMATRA _ NGKULU LAMPUNG _ DKI JAKARTA VWST JAVA _ CENTNAL JAVA___ __ YOGAKARTA T m CD .: EAST JAVA _ __ I ~~BA U c S N.T.B. N.T.I TIMOR TIMUR w W. KALAANTAN Z C. KALAMANTAN S. KALIMANTAN 0 E. KALSAANTA N. SULAWESI C. SULAWESI S. SULAWESI S.E. SULAWESIJP MALUKU "_ IRANJAYA ______ * -L CI. =c IBRD 26570 13 -.-1 X hoiS E > , INDONESIA \ ] 3 7|s ' > \ - ~~~~~~~BRUNEI 0= P 9; VMLAYSIM A R -tXA!GO '_NDA- 12 -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ .~~~~~ % A Tt 14E S/ >a 2 Lg X oM , t a~~~~~~~ A L I . N T A ID (OZ< '%a NORT PACIFIC OA Io " _ _ _ _ I S ULAK MAA ' ERA.- 1 DXLJW,T . -/\ A 2 JAWA P.AWE 5 AWA11TM J AE Y A L JV 5UD-:W_D (Pd>,. e k . . 6 UEIRCII 6 6 7 N .A - 6 I 5. 12 gMAIUTAPA1 13 DLACEN, 14 KAMU4VW4U o %W -OL%~ I n KuINDIANOCA <) v7:- 21 -AAWRIA D40oA_ 21, SULA6E T0AWA 2M NULSAT4OGAMAL M 24 #j 4 oGAA To"I INDIAN OCEAN 25 1460TML- -- 6 27 26 wm~ ~ ~ ~ Y ~StN.C-,MITO-I