Report No: 45868-SZ Swaziland’s Funding of Referrals Abroad: Assessment of the Phalala and Civil Servants’ Medical Schemes and Options for Improvement World Bank Human Development 1 Country Department 1 Africa Region Currency Equivalents Currency Unit = Swazi emalangeni US$1 = 7 Fiscal Year April 1 - March 30 ACRONYMS AND ABBREVIATIONS CTA Central Transport Authority ENT Ears, Nose and Throat GEMS Government Employee Medical Aid Scheme MGH Mbabane Government Hospital MOF Ministry of Finance MOHSW Ministry of Health and Social Welfare MOPSI Ministry of Public Service and Information MVA Motor Vehicle Accident PO Planning Office (of MOHSW) SE Swazi Emalangeni (national currency of Swaziland) SHI Social Health Insurance SMO Senior Medical Officer SNAT Swaziland National Association of Teachers SNACS Swaziland National Association of Civil Servants TA Technical Assistance TPA Third-Party Administrator WTO World Trade Organization Vice President : Obiageli Ezekwesili Country Director : Ruth Kagia Acting Sector Manager : Christopher Thomas Task Team Leader : Oscar Picazo ii PREFACE This report was prepared by Dr. Barry Kistnasamy, a consultant funded under the South African Trust Fund of the World Bank, and Oscar F. Picazo, senior economist, AFTH1, World Bank, Pretoria, South Africa. The authors acknowledge the support of the following individuals in the completion of this study, but do not incriminate them for any errors: • Mr. Dumisani Shongwe, economist from the Planning Office of the Ministry of Health and Social Welfare (MOHSW), Government of Swaziland, who was the driving force behind the study and made all the appropriate arrangements; • Dr. Austin Ezeogu, the senior medical officer (SMO) at the Mbabane Government Hospital (MGH) who is the main clinical contact point for all referrals to South Africa; • The dedicated staff of the Phalala Fund office and the Accountant General’s office that assisted with all aspects of the study: Ncamsile Masango, Mumcy Khumalo, Emily Nzimande, Lindiwe Motsa, Bongiwe Vilakati and Mcebo Dlamini; and • Dr. Moreshnee Govender, who assisted with the data entry and proof reading. The study was peer reviewed by Eva Jarawan, sector manager, AFTH2; Catherine O'Farrel, senior investment officer, Advisory Services Department of the International Finance Corp.; and Dr. Sagie Pillay, chief executive officer of the Johannesburg Academic Hospital. The authors wish to thank Sheila Dutta, senior health specialist and Christopher Walker, lead health specialist and cluster leader for HNP, AFTH1. Cristina Romero did the formatting and copy editing. The study was completed under the leadership of Ritva Reinikka and Ruth Kagia, former and current country director for CD1, and Dzingai Mutumbuka and Christopher Thomas, former and current sector mamanger of AFTH1. This study reflects the views of the authors and not those of the World Bank management and staff, nor of the Kingdom and Government of Swaziland. iii TABLE OF CONTENTS Executive Summary ........................................................................................................................................ vii A. Key Findings ................................................................................................................................. vii B. Options for Improvement ................................................................................................................ x I. Introduction................................................................................................................................................... 1 A. Background ..................................................................................................................................... 1 B. Objectives and Organization of the Study....................................................................................... 1 II. Assessment of the Administrative Framework, Management and Process of the Two Schemes............................................................................................................................................................. 3 A. Legal Framework ............................................................................................................................ 3 B. Administration of the Phalala Fund................................................................................................. 3 C. Administration of the Medical Referrals Scheme for Civil Servants.............................................. 6 D. Management and Staffing of the Two Schemes.............................................................................. 8 E. Process of Referrals and Payments................................................................................................ 10 III. Analysis of the Medical Referral Process and Statistical Data ................................................................ 11 A. Management of the Medical Referral Process .............................................................................. 11 B. Review of Patient and Provider Data ............................................................................................ 14 C. Analysis of the Indirect Costs of Referral ..................................................................................... 31 IV. Options for Improvement of the Two Schemes ....................................................................................... 32 A. The Assessment Findings as Basis for Change ............................................................................. 32 B. Principles to Govern Choice of Options for Improvement............................................................ 33 C. Needed Actions.............................................................................................................................. 36 ANNEXES ANNEX 1: LIST OF PERSONS INTERVIEWED ................................................................................................................ 40 ANNEX 2: AVERAGE COST PER CLAIM OF TOP TEN HEALTH PROVIDERS ......................................................... 41 ANNEX 3: TOTAL NUMBER OF CLAIMS AND TOTAL COST OF CLAIMS OF TOP TWENTY HEALTH PRACTITIONERS ....................................................................................................................................... 42 ANNEX 4: LOG SCALE OF HOSPITALS’ CLAIMS PROFILE ........................................................................................ 43 ANNEX 5: CLAIMS PROFILE OF HEALTH PRACTITIONERS ...................................................................................... 44 ANNEX 6: NUMBER OF REFERRED PATIENTS AND THE COST OF THEIR CLAIMS ............................................. 45 ANNEX 7: AMOUNT, NUMBER, AND AVERAGE COST PER CLAIM OF HOSPITALS, BY INDIVIDUAL HOSPITAL ........................................................................................................................................... 46 ANNEX 8: MAP OF SWAZILAND ...................................................................................................................................... 49 iv TABLES TABLE 1: SUMMARY OF THE FINDINGS OF THE ASSESSMENT OF THE PHALALA FUND AND MEDICAL REFERRALS SCHEME, 2006 ................................................................................................................. VII TABLE 2: EXPENDITURE ON THE PHALALA FUND, FY2002/03 TO FY2004/05......................................................... 5 TABLE 3: NUMBER OF PATIENTS ATTENDING VISITING SPECIALISTS AND PATIENTS REFERRED TO SOUTH AFRICA, AUGUST 2001 TO MARCH 2003....................................................................... 5 TABLE 4: COMPARATIVE EXPENDITURE BY THE PHALALA FUND AND THE MEDICAL REFERRALS SCHEME AGAINST CURATIVE MEDICINE EXPENDITURES BY THE MOHSW FOR THE GENERAL POPULATION, FY02/03 TO FY04/05 ................................................................................... 13 TABLE 5: OVERVIEW OF THE PHALALA FUND AND THE MEDICAL REFERRALS SCHEME, 2000S............................................................................................................................................................................. 14 TABLE 6: AVERAGE AMOUNTS OF CLAIM AND AVERAGE NUMBER OF CLAIMS PER PATIENT UNDER THE PHALALA FUND AND MEDICAL REFERRALS SCHEME .......................................... 15 TABLE 7: TOP 10 CIVIL SERVANTS’ AMOUNT OF CLAIMS UNDER THE MEDICAL REFERRALS SCHEME, BY MINISTRY........................................................................................................................................... 18 TABLE 8: AGE OF PATIENTS, DIAGNOSIS, COSTS AND OUTCOME OF MEDICAL INTERVENTIONS IN THE TOP 10 PATIENTS REFERRED UNDER THE PHALALA FUND ............................ 19 TABLE 9: CLAIMS PROFILE OF MINISTRIES UNDER THE MEDICAL REFERRALS SCHEME, BY MINISTRY ................................................................................................................................................................... 21 TABLE 10: CLAIMS DATA OF HEALTH PROVIDERS UNDER THE PHALALA FUND, BY TYPE OF PROVIDER OR SERVICE..................................................................................................................................... 22 TABLE 11: CLAIMS DATA OF HEALTH PROVIDERS UNDER THE MEDICAL REFERRALS SCHEME, BY TYPE OF PROVIDER OR SERVICE ................................................................................................. 22 TABLE 12: AMOUNT, NUMBER, AND AVERAGE COST PER CLAIM UNDER THE PHALALA FUND, BY TYPE OF PROVIDER............................................................................................................................... 24 TABLE 13: AMOUNT, NUMBER, AND AVERAGE COST PER CLAIM UNDER THE MEDICAL REFERRALS SCHEME, BY TYPE OF PROVIDER.................................................................................................. 25 TABLE 14: LABORATORY CLAIMS AND COSTS UNDER BOTH FUNDS.................................................................. 25 TABLE 15: TOP 10 HEALTH PROVIDERS UNDER THE PHALALA FUND, BY THEIR NUMBER OF CLAIMS AND AVERAGE COST PER CLAIM ......................................................................................................... 27 TABLE 16: TOP 10 HEALTH PROVIDERS UNDER THE MEDICAL REFERRALS SCHEME, BY THEIR NUMBER OF CLAIMS AND AVERAGE COST PER CLAIM .................................................................... 28 TABLE 17: AMOUNT, NUMBER, AND AVERAGE COST PER CLAIM OF HOSPITALS IN BOTH FUNDS, SUMMARY FOR ALL HOSPITALS ........................................................................................................... 28 TABLE 18: PROVIDER AND PATIENT REFERRAL CLAIMS PROFILES FOR THE PHALALA FUND............................................................................................................................................................................ 29 TABLE 19: PROVIDER AND PATIENT REFERRAL CLAIMS PROFILES FOR THE MEDICAL REFERRALS SCHEME ............................................................................................................................................... 30 TABLE 20: AMOUNT, NUMBER, AND AVERAGE COST PER CLAIM OF REFERRED PATIENTS FROM BOTH FUNDS.................................................................................................................................................. 30 TABLE 21: SUBSISTENCE AND TRAVEL COSTS FOR BOTH FUNDS ........................................................................ 31 TABLE 22: TIMEFRAME AND BUDGET FOR SHORT-TERM INTERVENTIONS....................................................... 38 v FIGURES FIGURE 1: ADMINISTRATIVE PROCESSES FOR THE PHALALA FUND AND THE MEDICAL REFERRALS SCHEME ................................................................................................................................................. 7 FIGURE 2: EXPENDITURE ON THE PHALALA FUND AND THE MEDICAL REFERRALS SCHEME, FY02/03 TO FY04/05 ................................................................................................................................. 12 FIGURE 3: TOP 5 DISEASE CONDITIONS AND NUMBER OF PATIENTS REFERRED UNDER THE PHALALA FUND TO SOUTH AFRICA, JANUARY TO NOVEMBER 2006 ......................................................... 16 FIGURE 4: PERCENT OF TOP 5 REFERRAL SPECIALTIES UNDER THE PHALALA FUND, MAY TO OCTOBER 2005 ..................................................................................................................................................... 16 FIGURE 5: TOP 10 REFERRED PATIENTS IN PHALALA FUND, THEIR NUMBER OF CLAIMS, AND AMOUNTS OF CLAIM...................................................................................................................................... 17 FIGURE 6: TOP 10 REFERRED CIVIL SERVANTS, THEIR NUMBER OF CLAIMS, AND AMOUNTS OF CLAIM.................................................................................................................................................................... 17 FIGURE 7: PERCENT OF CLAIMS BY TYPE OF PROVIDER UNDER THE PHALALA FUND .................................. 23 FIGURE 8: PERCENT OF CLAIMS BY TYPE OF PROVIDER UNDER THE MEDICAL REFERRALS SCHEME....................................................................................................................................................................... 23 FIGURE 9: TOP 10 HEALTH PROVIDERS UNDER THE PHALALA FUND.................................................................. 26 FIGURE 10: TOP 10 HEALTH PROVIDERS UNDER THE MEDICAL REFERRALS SCHEME ................................... 26 FIGURE 11: PRINCIPAL PLAYERS IN A HEALTH CARE SYSTEM.............................................................................. 34 vi EXECUTIVE SUMMARY A. KEY FINDINGS 1. The findings of the general assessment of the Phalala Fund for Swazi citizens and the Medical Referrals Scheme for civil servants are summarized in Table 1. Table 1: Summary of the Findings of the Assessment of the Phalala Fund and Medical Referrals Scheme, 2006 Aspects Phalala Fund Medical Referral Scheme Year established 2001 1995 Government of Swaziland Coverage Swazi citizens civil servants and their dependents Ministry of Health & Social Line Ministries/Ministry of Governance Welfare/Ministry of Finance Finance Clearly defined and Funds appropriated in Budget and Not defined Expenditure Expenditure during the last SE 32.8 million SE 14.4 million period Approximate number of 760 patients 460 civil servants persons who benefited (about SE 41,000 per (about SE 31,000 per during the last period patient) patient) (year) 2,300 claims Approximate number of 3,200 claims (about SE 13,000 per claims (about SE 4,400 per claim) claim) Number of providers of health services who 180 providers 375 providers claimed during the last (about SE 180,000 per (about SE 40,000 per period (doctors, hospitals, provider) provider) laboratories, etc) 22 hospitals (67 percent of 45 hospitals (57 percent of claims) claims) Number of hospitals (about SE 1 million per (about SE 180,000 per hospital; about SE 36,000 hospital; about SE 15,000 per claim) per claim) 142 practitioners (24 304 practitioners (36 percent Number of individual percent of claims) of claims) health practitioners (about SE 55,000 per (about SE 17,000 per (doctors, dentists, practitioner; about SE 6,100 practitioner; about SE 2,700 physiotherapists, etc) per claim) per claim) vii 2. Direct costs. On an annualized basis, both funds will spend approximately Swazi Emalangeni (SE) 60 million (about US$8.6 million) on referrals to South Africa in this financial year for approximately 1,500 patients. During the last period, the two funds already spent SE 51 million for the benefit of about 1,220 patients. Strictly speaking, the funds are noncontributory and are mainly derived from the government budget1. The Phalala Fund did receive some donations2 from the private sector in FY 2002/03. 3. Key weaknesses. The assessment found the following major weaknesses of the two funds. • Only a tiny segment of the Swazi population benefits from the large subsidy. In the period covered under this study, only 760 citizens benefited from the SE 32.8 million expenditures of the Phalala Fund, and only 460 civil servants benefited from the SE 14.4 million expenditures of the Medical Referrals Scheme. (The expenditures for the Phalala Fund and the Medical Referrals Scheme over the FY 2002/03 to 2004/05 were 18 percent, 23 percent and 21 percent of the costs of curative care in the MOHSW – approximately 1 in 5 health SE went to this small group). Thus, the equity impact of the two funds is highly adverse. • The lack of cost-effectiveness guidelines and clear rationing criteria also implies that the funds could be used for patients that are of dubious social value3. For instance, a random selection of a few cases referred under Phalala shows that 4 out of 10 died. • While the extensive provider base in the private sector favors choice among patients, it also precludes fee negotiation and cost control. The health practitioners, hospitals, pharmacies and laboratories are located mainly in the Gauteng province and cover an extremely wide geographic area. There were 180 providers who claimed under the Phalala Fund and 375 under the Medical Referral Scheme. • Fees and other prices are not negotiated beforehand and are completely supplier- determined. Fees are requested at the point of need, when the government is at its weakest negotiating position. Thus, the two funds provide a "blank cheque" for 1 During the social health insurance orientation workshop held in Manzini in November 2005, civil servants’ union representatives argued that indeed they "contribute" to the fund. However, the fact is that the Government periodically raises civil servants’ salaries, and uses part of the salary increase to finance the fund. Under a full-blown social health insurance fund, the contributory nature of the funding is such that both Government and members’ contributions are determined in advance, and their respective contributions are collected regularly. This is not the case with the two funds. 2 Donation of SE 4,586m in FY 2002/03 from Shell Swaziland. Source: Commission of Enquiry in terms of Legal Notice No. 165 of 2004, p. 8. 3 http://www.observer.org.sz/main.asp?id=27216&Section=main&articledate=Friday, September 22, 2006. viii South African doctors and hospitals: whatever amount that they ask is paid for by the Government, since it has no recourse but to pay up. There are no contracts or preferred provider arrangements and negotiated fees in place. • There are no co-payments at all from beneficiaries. This may provide perverse incentives, both from patients and providers, for over-utilization of services. • The local referring doctors are involved only in the medical assessment of the referred patients and are minimally involved in the fee negotiation, or in claims utilization and review. The payments are solely the purview of the Ministry of Finance and Accountant General. This bifurcation of responsibilities engenders supplier-induced cost inflation: No patient tracking ever takes place, providing incentives to the admitting doctor or hospital in South Africa to possibly extend the unnecessary confinement of the patient. No utilization review is ever undertaken, providing incentives to the admitting doctor or hospital to add superfluous referrals, other medical investigations, or medications. • The payment process suffers from management problems. The payor could mistakenly pay a bill twice or thrice due to multiple "reminder" billings. Bills covering the Phalala Fund may, in some instances, be sent to the Medical Referrals Scheme for payment and vice versa. The transaction costs for both funds are high given the large number of providers involved, and the low number of claims for many of them. All claims processing are done manually with no links between the clinical review and the administrative/financial processes. • Referrals and care that should be covered out of the Motor Vehicle Accident (MVA) Insurance Fund or Workmen’s Compensation Fund are being covered by the Phalala Fund or the Medical Referrals Scheme. • The funds are financially nontransparent in that they do not produce any periodic or annual reports including summary tables of their operations, much less an audited financial statement. • Overall, there is a lack of comprehensive management information systems that would support the administrative, financial, logistic and clinical activities of the two funds. While these two funds are multimillion-dollar payors, their information system is extremely rudimentary. 4. Indirect and opportunity costs. Various indirect costs are being covered by the Central Transport Authority, the MOHSW, the Accountant General’s office, and other line Ministries to administer the funds, provide the accompanying nurses and drivers, as well as the travel and subsistence costs. The estimated subsistence and travel costs for the two funds amounted to approximately SE 3.1 million. ix 5. The burden of costs that accrue to the patients and their families especially within the Phalala Fund have not been quantified and include the following costs: (a) the costs in getting to Mbabane Government Hospital and back to their homes, and (b) the costs of the travel documents and the process of obtaining them. B. OPTIONS FOR IMPROVEMENT 6. Local resource conservation. A guiding principle in restructuring the Phalala Fund and the Medical Referrals Scheme is that the funds should flow within Swaziland as far as possible, rather than flowing out of the country. A medium term option is that of creating a managed care system in which the funder, the providers, and the patients operate within one system with shared risk amongst all players. This would entail substantial linkages with the private sector in Swaziland and/or South Africa, with the bulk of the health care provision for beneficiaries to be rendered within Swaziland. 7. Needed short-term actions. The following are proposed: • From being funded straight out of the budget, the Medical Referrals Scheme should be converted into a contributory Social Health Insurance (SHI) fund with seed capital from the government, and with regular contributions from employees (the civil servants) and the government as employer. The key areas of concern may be that of coverage (willingness of civil servants to contribute), and co- payments for health services and pharmaceuticals that may be required in certain instances. This new fund should have its own governance structure (similar to the Pension Fund) and operate under autonomous legislation with possible line reporting to the Ministry of Public Service and Information, if not the MOHSW (as is the case with SHIs in many countries). • The Phalala Fund should also become autonomous with reporting via the MOHSW. An Advisory Board (with community/private sector representatives) can assist in its governance and functions. The Phalala Fund needs special consideration as a health funding tool with the use of health insurance principles (including re-insurance for special medical conditions and health services). It should also adopt patient case management and utilization review. The feasibility of would-be beneficiaries sharing in some of the costs should also be explored, including one that allows for periodic contributions. • The provision of specialist medical care within Swaziland should be vigorously explored through: (a) partnership with the Swaziland private health sector; (b) building up both the public and private facilities - perhaps through amenity wards - at MGH with support from the Swaziland and/or South African and/or other country private sectors; and (c) developing a new private hospital facility with the eight major clinical specialties with the support of laboratory, radiology and pharmacy facilities as well as visiting specialties in some disciplines (e.g., oncology). In all instances, funding would be needed for the initial capital costs as well as the first few years of operations. x • Membership, logistic support, and claims administration including utilization review can be done by one entity (utilizing the current staff in the Phalala office and the Accountant General’s office). This can be in-house or contracted to a third-party health insurance fund administrator. This may be extended to case management and clinical review with the involvement of a senior nurse and doctor. 8. Needed immediate actions. • Defining the referral and coverage criteria for both funds; • Setting up preferred provider arrangements with a limited number of service providers in South Africa as well as the private sector in Swaziland; • Improvements in logistic and transport management; • Negotiating with a provider of a "step down" facility in South Africa, either in Pretoria or Johannesburg (depending on which providers will be used) as well as having a professional nurse and social worker based within the Swaziland High Commission in South Africa to support the patients and to visit the various health providers for quality assurance on a regular basis; • Enhancing the administration of the Phalala Fund and Medical Referrals Scheme with appropriate physical infrastructure, computerized management information systems, human resource development (clinical, social, and administrative), and training; • Recovery of monies from the Motor Vehicle Accident and Workmen’s Compensation Funds; and • Information to relevant stakeholders and the civil servants directly and/or through their representatives of the efficient use of the Phalala Fund and the Medical Referrals Scheme. 9. Conclusions. This report concluded with a list of short-term actions needed to achieve the objectives of a reform program for the two funds. It also identified the key actors, a possible timeframe for the actions from these actors, and the respective budgetary resource requirements for each action. Over the medium- term, however, what is required is a projectized set of technical assistance support for converting one or both funds into a full-blown social health insurance program. This could very well be done in tandem with the intended upgrading of the Mbabane Government Hospital, which requires a sustainable revenue base that could be provided by the social health insurance program. xi I. INTRODUCTION A. BACKGROUND 1. Swaziland is a lower middle-income country with the highest HIV/AIDS prevalence in the world. An area that has received some scrutiny in recent years is the financing by the government of Swaziland of the medical referrals of patients and civil servants from Swaziland to South Africa. In Swaziland, the two programs dealing with medical referrals involve multi- million dollar transactions and benefit an extremely small segment of the population. The Phalala Fund, which caters to ordinary citizens, is funded through a "trading account" of the Ministry of Finance (MOF) but implemented through the Ministry of Health and Social Welfare while the Medical Referrals Scheme is implemented through all Ministries. Between April 1, 2005 and March 31, 2006, the fund spent SE 70.6 million (US$11.8 million) for about 1,500 patients, or an average of SE 47,000 (US$6,700) per patient. Translated into per-capita terms, the expenditures of this fund alone are about US$11.80, which is roughly the total national health expenditures per-capita of Malawi, Mozambique, or Tanzania. 2. This assessment is the first attempt to analyze "Referrals Abroad", a common but little- understood line item in many African countries’ budgets. Even with World Trade Organization’s inclusion of trade in health services, very little information exists on this phenomenon. Public expenditure reviews conveniently omit them, despite their size, because of non-existent data. The assessment aims not only to evaluate the patterns of utilization of the two referral funds, but seeks to generate options on how these two funds can be converted from being passive buyers and payers of services, to active purchasers of care. This will require a major institutional change. The findings of the study will be used as input into the proposed upgrading of the Mbabane Government Hospital as part of the options for sustainability of the funds, the hospital, and Swaziland’s overall health system. B. OBJECTIVES AND ORGANIZATION OF THE STUDY 3. This study was undertaken at the request of the MOHSW. Based on discussions with the Ministry’s Planning Unit, it was agreed that the study should cover two tasks: (1) an assessment of the Phalala Fund and the Medical Referrals Scheme for civil servants, and (2) an assessment of the options to improve the governance and management of the two referral funds 4. Task 1. This task aims to evaluate the experience of patient referrals abroad from the Medical Referrals Scheme and the Phalala Fund by: (a) Organizing the available administrative and service statistics data on number of patients served, diagnosis of illness, amounts paid for by the MOF, attending doctor and hospital used by the patients, and laboratory tests performed and other inputs provided. Summary statistics will also be analyzed, e.g., volume of patients per doctor or hospital, and cost per doctor or hospital. (b) Analyzing the process of and institutional structure for patient assessment/screening, referral, treatment, billing, discharge, transport, and payment; identifying weaknesses in this process; and proposing ways for improvement. 5. Task 2. This task aims to explore options to improve the governance and management of the two funds, including but not limited to: (a) Conversion of the Medical Referrals Scheme into a full-blown mandatory and contributory medical aid scheme. This should permit for more 1 professional management of the scheme, including use of health insurance principles and practices in patient gate-keeping, utilization review, claims processing, management information systems, reserve requirements, etc. (b) Active fee negotiation, perhaps on an annual basis, before actual referrals are made. (c) Limiting the number of “accredited� providers to those willing to provide services at a negotiated fee, through "preferred provider" arrangements, to reduce medical inflation. (d) Contracting out the billing and claims review to a professional Third-Party Administrator (TPA). (e) Harmonizing the roles of the Referral Board (as gatekeepers) and MOF/Accountant General (as payor) of the scheme so that patient tracking and utilization review can occur. (f) Computerization of the information system. 6. The report is organized as follows. Chapter II provides an assessment of the administrative framework, management, and processes of the two funds. Chapter III provides an analysis of the medical referral process and statistical data. Chapter IV presents options for improving the operations of the two funds. 2 II. ASSESSMENT OF THE ADMINISTRATIVE FRAMEWORK, MANAGEMENT AND PROCESS OF THE TWO SCHEMES A. LEGAL FRAMEWORK 1. Phalala Fund. The Phalala Fund was set up in 2001 and is administered under the Administration of the Specialist Medical Aid Fund Regulations, 2001, under section 12 (4) of the Finance Management and Audit Act 1967 (Act No. 18 of 1967)4. The regulation sets out the objective of the fund and its administration, the process and criteria for referrals as well as the reporting structures. The fund was established under Legal Notice No. 149 of 20015. The objective of the fund is to assist deserving Swazi citizens to access specialist medical care within the Kingdom of Swaziland or, in special circumstances, outside the Kingdom. The sources of funds are from the government and donations2. Governance of the fund is through the Permanent Secretary of the MOHSW, with quarterly and annual reports provided to the Permanent Secretary, Ministry of Finance. The Minister of Finance shall present the annual report to Parliament within six months of the end of the financial year. The fund shall be audited by the Auditor General. 2. Civil Servants Referral Scheme. The Government Medical Referrals Scheme was established in 1995 under the Establishment Circular no. 17 of 1995 – Guidelines for the Operation of the Medical Referrals Scheme for Government Employees6. The Joint Negotiations Team agreed on the guidelines for the scheme, which set out the definitions and eligibility criteria for dependents, the aims, criteria and process for referrals, as well as certain transitional arrangements. The scheme was to be set up in terms of the Finance and Audit Act and to date, does not appear to be set up as a pooled scheme. B. ADMINISTRATION OF THE PHALALA FUND 3. The Phalala Fund is administered through the MOHSW and has dedicated offices with administration and finance personnel based at the MGH. The finance personnel moved to MGH about three months ago from the MOHSW building. The Specialist Care Medical Aid Committee appointed by the Minister for Health and Social Welfare (after consultation with the Cabinet), administers the fund. The fund operates as a trading account under the Accountant General with the Accountant General approving all payments. 4. The Phalala Fund pays for all hospital fees, doctor and patient transport, doctor and patient accommodation, prescribed surgical or clinical treatment, and equipment and medical supplies. 5. The Phalala Committee comprises five persons: the Director of Health Services (Chair), the SMO (or representative) (Secretary), relevant specialists (co-opted), representative of the nursing cadre, representative of the special health unit, and the medical practitioner in private 4 Final Report. Commission of Enquiry in terms of Legal Notice No. 165 of 2004.Appendix I, p. 36. 5 This differs from p. 36 in the Commission of Enquiry Report where it refers to Legal Notice No. 195. 6 Final Report. Commission of Enquiry in terms of Legal Notice No. 165 of 2004. Appendix II, p. 41. 3 practice. According to the rules, the committee members shall receive sitting and traveling allowances7. 6. The Phalala Committee had to develop the criteria for referrals and circulate them to all practitioners within Swaziland. Broad guidelines are given as follows: The government hospitals will receive all referrals. The committee should assess the referral. The Committee has to be convinced that all local medical capacity has been exhausted, before it makes a referral. The nature of the patient’s condition should be life threatening or creates undue hardship; the fund could not be used for cosmetic purposes. The patient is not a beneficiary of any other scheme which should bear the costs. Of note are the clauses that by interpretation may allow for patients to be seen in the private sector in Swaziland if the medical interventions can be carried out by them. In addition, there is support for specialist skills to be brought into Swaziland. The committee should approve the institutions outside Swaziland to which patients are referred. (This clause in all probability allows for waiver of the relevant procurement procedures). Provision is made for the Chair or the SMO to authorize referrals for emergencies covering emergency care within and outside Swaziland. 7. Expenditures of Phalala in 2001/02 period were SE 495,4768 (for a four-month period). The bulk of the expenditures were on hospitals and doctors. The specialist doctors from South Africa visited MGH in FY 2003/04 and this may account for the increased expenditure on doctors in that year (Table 2). The referrals to South Africa in that year decreased considerably resulting in lower Hospital and Laboratory costs. However, the costs of hospitalization and laboratory services may have been picked up by MGH in its expenditure. The total expenditure of SE 36.7m in FY 2003/04 differed slightly from the total expenditure in FY 2004/05 when the visiting specialist service was stopped. The visiting specialists saw approximately 6,000 patients at MGH (Table 3) which was far higher than the numbers of patients referred to South Africa on an annual basis. It is not known how many patients were seen on an outpatient basis or as inpatients at MGH. An assessment of the Medical Referrals Scheme in the overlapping period (2002/03 to 2003/04) showed that many of the civil servants still visited South Africa for their health care needs despite the presence of the visiting specialists at MGH. Some 408 patients were referred to South Africa during this period under the Phalala Fund. 7 This has not occurred leading to the committee not functioning optimally; however, the Senior Medical Officer (Dr. Ezeogu) and his specialist members meet every two weeks to assess the patients that need referral to South Africa. 8 Final Report. Commission of Enquiry in terms of Legal Notice No. 165 of 2004, p. 8. 4 Table 2: Expenditure on the Phalala Fund, FY2002/03 to FY2004/05 Payments for FY 2002/03 FY 2003/04 FY 2004/05 (SE ‘000) (SE ‘000) (SE ‘000) Hospitals 11,256 8,019 15,847 Laboratory 3,142 2,642 5,203 Doctors 8,072 25,052 8,182 Drugs 431 392 724 Accommodation 245 263 0 Burial benefit 64 34 33 Transport 207 65 0 Specialists (refreshments) 167 67 0 Local Clinics 32 25 0 Miscellaneous 13 51 66 Wages 6 83 182 Ambulances 6 12 0 Postage 0 2 0 Adjustment 4,586 0 6,952 Total Expenditure 28,227 36,707 37,189 Source: Treasury Annual Reports 2002/03, 2003/04, 2004/05, Ministry of Finance Table 3: Number of Patients Attending Visiting Specialists and Patients Referred to South Africa, August 2001 to March 2003 Specialties Number of Patients Attending Number of Patients Visiting Specialists Referred to South Africa Cardiology 605 131 Cardiothoracic - 18 Gynecology - 3 Internal Medicine 701 20 Maxillofacial 150 32 Neurosurgery 502 55 Oncology - 35 Opthalmology 820 - Orthopaedics 720 69 Paediatrics 480 18 Radiology 1,020 - Surgery 420 - Urology 740 27 Total 4,200 4089 Source: Phalala Fund. Report from August 2001 - March 2003. Dr. P.K. Dlamini, Minister of MOHSW, p.15. 9 Incorrect total in Phalala Fund Report. August 2001 – March 2003. Dr P.K. Dlamini, p. 15 5 C. ADMINISTRATION OF THE MEDICAL REFERRALS SCHEME FOR CIVIL SERVANTS 8. The scheme came about in 1995 through a negotiated process between government as the employer and the civil servants’ representatives. The scheme covers the civil-servant-member and the following dependents: the legally registered spouse and children (biological and legally adopted) under the age of 21 or up to 25 if in full time education or disabled and fully dependent on the civil servant. 9. The principles guiding the scheme6 are: (a) Coverage of all government employees who are Swazis and their dependents. (This excludes expatriates). Of note is the fact that the "…fund will scrutinize all requests". (b) The Medical Referrals Board will consider all cases and MGH will be the initial referral hospital. (c) Apart from the need for the referral, note is taken that the patient should benefit from the treatment. The criteria cover emergency and non-emergency patients (similar to the Phalala Fund). 10. Provision is made for the setting up of a committee comprising members of the Joint Negotiations Teams10. This committee will review all cases handled by the Medical Referrals Board. This will be done without contravention of medical practice or ethics. The committee will also review the operations of the fund. The Medical Referrals Board comprises the Senior Medical Officer of MGH (chair); four specialists and heads of the following departments in MGH - internal medicine/pediatrics, general surgery, orthopedics, and obstetrics/gynecology; and co-opted members (from other specialties as well as from the visiting external specialists). The Board reports to the Director of Health Services. 11. The Medical Referrals Fund shall be established from which payments for referrals can be made11. The Accountant General’s office issues the payment guarantees upon approval from the Medical Referrals Board. Mention is made of payment of ".. named medical institutions in South Africa"12. Transport expenses will be covered by the fund. (In reality, the ministries seem to be carrying this cost where a referred civil servant seeks care in South Africa and claims back from the line ministry, the costs of own transport). Subsistence allowances in terms of General Orders A845 as amended will be paid by the fund. (This seems to be covered by the line ministries as well). The transitional arrangements seem to have continued with the Accountant General’s office continuing with the payments of the bills. (An accounting officer administers the payments in liaison with the other accounting officers looking after the line ministries). 12. The administrative processes for both the Phalala Fund and the Medical Referrals Scheme (Figure 1) are quite complex and extensive and are directed mainly towards the claims processing aspects of the funds rather than clinical review and case management activities. 10 This committee was never set up. 11 There was not a clearly defined Fund covering the Medical Referrals Scheme as with the Phalala Fund which is clearly designated and appropriated and can be viewed in the Treasury Annual Reports. 12 There is no approved list of institutions (hospitals) in South Africa or any list covering other Health Practitioners or providers. 6 Figure 1: Administrative Processes for the Phalala Fund and the Medical Referrals Scheme Technical committee Patient / (Evaluates the patient; meets Doctor seeks Civil Servant every 2 weeks; can review financial referred emergencies) assessment from provider Letter to Director Letter to relevant (Health & Welfare) Ministry Director generates Treasury generates Guarantee Letter Guarantee Letter Phalala office makes Phalala office makes bookings with facility, bookings with facility, arranges transport & civil servant arranges immigration visa transport & immigration Invoice sent to Invoices sent to Accounts Division in Accountant General’s Phalala office office Cross checking and Cross checking and sign offs from sign offs from accounting perspective accounting perspective Sent for payment via Sent for payment to relevant ministry accounts Accountant General; officer in Treasury; Reconciliation cheque Reconciliation cheque / / invoice invoice Phalala office calls provider or provider Accounts officer calls calls Phalala office for provider or provider cheque calls office for cheque 7 D. MANAGEMENT AND STAFFING OF THE TWO SCHEMES 13. There are two "typists" who support the Medical Referrals Board’s activities, make and manage the bookings and arrange the logistics for referrals to South Africa from MGH and the return of the patients to MGH. In addition, there are three accounting officers (one permanent and two temporary) that process the payments for the Phalala Fund and liaise with the providers in South Africa. There is one messenger13. The team relies on nurses from the nursing services of MGH to accompany the patients and drivers from the Central Transport Authority (CTA). Subsistence allowances (US$120 per day) are paid to the driver and nurse. In addition, funds are provided for the fuel and toll fees (SE 700 per journey). The Medical Referrals Scheme has a dedicated accounting officer in the Accountant General’s office. 14. The Phalala Fund officers have two offices at MGH with the SMO using his office as an additional space. The Board uses the SMO’s office for its deliberations. There are two stand- alone computers in the Phalala office for secretarial functions and one stand alone computer in the Phalala Accounts office (an old 386)14. The Accounts Officer in the Accountant General’s office has access to an on-line computer where she can track the payment process. 15. Patient files are kept in filing cabinets in the Phalala office. There is generally very little, if any, feed- back from the referring institutions, laboratories, radiological examinations and doctors based in South Africa. 16. For the purposes of this evaluation, all staff involved with both funds were interviewed. Data for the Phalala Fund was extracted manually from a book in the SMO’s office, a diary in the Phalala booking office, a Government of Swaziland register and a single entry Analysis book for the Medical Referrals Scheme in the Accountant General’s office. Mr. Dumisani Shongwe assisted together with Computer Services to extract expenditure data from the old 386 computer in the Phalala accounting office. 17. The SMO is the key person for both funds. He outlined the process as detailed in Figure 1. At times, the committee is placed under tremendous political pressure, with comments such as "It is our King’s fund" to send patients to South Africa over and above the medical criteria. Approximately 15 patients are screened at each meeting and nearly all are referred. The SMO has many varied functions including management activities in MGH as well as providing an extensive clinical service in his discipline. 18. He had submitted a memo15 about the Phalala Fund to the MOHSW highlighting the challenges as being the lack of specialists in 16 specialties, and the lack of equipment and appropriate infrastructure in some specialties. He has suggested that over the next three years, the gradual employment of various specialists be made as well as improving the equipment and physical infrastructure at MGH. His proposed budget is for SE 10 million in year 1, SE 16 million in 13 Did not find a messenger at the Phalala Office. 14 For a fund that pays out in the millions of rands, it has severely under-invested in computers and appropriate information system. 15 Undated memo from Dr Austin Ezeogu to MOHSW. 8 year 2, and SE 18 million in Year 3. He suggests that there will be substantial reductions in the referrals to South Africa if this is carried out. 19. Limitations of the study • Given the split between the clinical referral process and the payment procedure, it was difficult to link patient clinical outcomes to reimbursements. Patient outcomes relating to deaths and discharges as well as average length of stay, type of procedure, and improved or worsening condition could not be assessed. Files of the top ten patients with the highest expenditures in the Phalala Fund were assessed and none of them had any clinical records of note. • It was not possible to differentiate a civil servant’s claim from a dependent’s claim owing to the use of the employee number as the defining variable. • The manual systems of recording in the multiple data sets in the Phalala and Accountant General’s offices serve mainly an accounting and reimbursement function rather than a clinical review, case management, and service utilization function. Thus, it was not possible to assess whether the amount of paid out claims has any relation to the severity of the condition being treated. • It was not possible to link the guarantee amounts to the final payments as the guarantee letters are in individual patient files and not on a database for final reconciliations with invoices and payments. On a number of occasions, a second or third letter is generated from the SMO to the Director of Health or to the Accountant General seeking additional guarantees. • There are additional indirect costs attributed to the line ministries for subsistence and travel as well as to the CTA for transport. Other indirect costs are bank fees. The assessment of the Phalala Fund showed that 372 cheques were issued for a total amount of SE 24 million16. (There were no cheque records for approximately SE 8.8 million in the Phalala Fund). Twenty three cheques were cancelled costing approximately SE 100,000 and possibly re-issued in the Medical Referrals Scheme17. • Opportunity costs arise for the nurses, ambulance drivers and the patient’s families while accompanying the patients. This was not assessed in this study. • A further detailed analysis of the cause of the illness or injury that led to the referral of patients to South Africa may show that other funds such as the Motor Vehicle Accident Fund and the Workmen’s Compensation Fund should be carrying those costs. 16 Assessment of computer records in the Phalala office at MGH. 17 Assessment of the sample Analysis book for the Medical Referrals Scheme from the Accountant General’s office. 9 E. PROCESS OF REFERRALS AND PAYMENTS 20. In terms of the administrative process, the typists function both in a secretarial role as well as social worker and logistics officer (making bookings and transport arrangements as well as scheduling the accompanying nurses and drivers). As outlined in Figure 1, a patient who is referred to South Africa is initially assessed by the medical (technical) team chaired by the SMO who generates the letter of referral to the Director of Health in the MOHSW. She then generates a guarantee letter to the provider which is sent to the provider by the Phalala office in order for them to send a booking confirmation letter back so that a medical visa can be procured for the patient from the South African High Commission in Mbabane. The Medical Referrals Scheme has a guarantee letter generated by the Accountant General’s office. 21. The four accounts officers (three at the Phalala Fund and one for the Medical Referrals Scheme) spend an enormous amount of time moving documents among various offices (MOHSW, Finance, and other line ministries), processing payments, as well as answering telephone and faxed queries from providers who in many instances are rude and disparaging. 22. Patients have to get to MGH to be assessed for referral. This causes great stress (moral, physical and financial) on the patients and their families in arranging transport to MGH from other places in Swaziland. In many instances, they are accommodated within MGH while awaiting referral to and from South Africa leading to additional stress on that institution. The patients also have to procure passports at a cost of SE 80 or SE 100 depending on whether it is an emergency travel document or not. Passport photos are additional costs. Fortunately the medical visa for South Africa is free but the documentation needed is quite demanding. 23. Elective patients under the Phalala Fund are taken to South Africa by 16-seater vehicles or ambulances twice a week (Tuesdays and Thursdays). In some instances, the vehicles are inappropriate for the type of patients that they transfer. The vehicles invariably leave Mbabane late in the day and arrive in the night at health facilities in South Africa. The facilities are far apart which compounds the problem. Other problems of note that could not be evaluated include: (a) Patients being discharged at hospitals in South Africa and waiting up to five days to return (especially with a weekend discharge). (b) Patients missing outpatient treatment appointments especially with late arrival in South Africa necessitating a re-visit out of schedule. (This would be unacceptable especially with patients receiving chemotherapy and other follow-up treatment which need standard intervention regimens/protocols and time periods). (c) The support of the patient’s family and the inability of the patient to contact family members in Swaziland when hospitalized in South Africa. 24. The Phalala office and the High Commissioner’s office in South Africa do provide support in the event of deaths of patients with the patient’s body being brought back to Mbabane through a private contractor funded by the Phalala Fund. The families of the patients have to take the body to the patient’s village or home from Mbabane. In the event of a death of a civil servant, the line ministry assists the civil servant’s family to get the body back to Swaziland. 10 III. ANALYSIS OF THE MEDICAL REFERRAL PROCESS AND STATISTICAL DATA A. MANAGEMENT OF THE MEDICAL REFERRAL PROCESS 1. The process is geared towards financial management. The accounting officers in the Phalala office and Accountant General’s office diligently review invoices and facilitate the payment process. Much of this is a manual process with numerous forms to be filled in as well as obtaining the necessary authorizations and approvals. However, there are minimal linkages between the clinical assessment/diagnosis and the claims administration. 2. Despite the diligence given by staff, the accounting process may still be flawed. There are manual recording systems and missing data with respect to invoices and cheque numbers as well as discrepancies in the amounts between invoices and cheques. At times, payments are made on the basis of ‘Various Patients’. This amounted to SE 1.74 million in the Phalala Fund assessment. The manual recording systems did not capture cheque numbers for a total amount of SE 8.8m under the Phalala Fund16. The process to generate a cheque payment is laborious and time consuming, notwithstanding duplicate payments, no cheque reconciliations, lost and cancelled cheques17 and time consumed by the accounting staff to track payments. The staff also face verbal abuse over the telephone from health providers that have not been paid. 3. Governance of the Phalala Fund is weakened by its split between the MOHSW and Ministry of Finance. It operates as a "trading account" in the Ministry of Finance and has an annual cap. The Medical Referrals Board functions with Dr. Austin Ezeogu as the chair while the other structures are non-functional. The Director of Health should prepare quarterly reports and an annual report for submission via the MOHSW to the Ministry of Finance and thereafter to the Parliament (within six months of the closure of the financial year). To date, there has been one report reviewing the Phalala Fund18 and none for the Medical Referrals Scheme. 4. The status of the Medical Referrals Scheme in accounting terms is not known. It is administered within the Ministry of Finance and its actual expenditures are difficult to track owing to the involvement of all line ministries. 5. Both funds are poorly defined in terms of contributions and benefits. The funds have no defined contribution, no clearly defined benefits, and especially in the case of the Medical Referrals Fund, no expenditure cap - it is open ended. The Phalala Fund however, has an expenditure cap. 6. Both funds have no fee negotiation with providers, and no utilization review or case management process. The review of a sample of claims shows that the amounts paid approximate the guarantee letters and at times exceed the guarantee amount, necessitating a second or third guarantee letter being generated. There was no documented evidence that the final paid amount is lower than the quoted amount in the guarantee letter. Thus, whatever the providers ask for, it is guaranteed through the letter. And whatever is guaranteed, is paid for. 18 Phalala. Specialists’ Care Medical Fund. Report from August 2001 – March 2003. (His Excellency) Dr P K Dlamini. 11 7. The patient transport system for the Phalala Fund is inadequate and at times inappropriate for the type of medical conditions. (a) There is an extensive bureaucratic process of sourcing the vehicles, the drivers and the accompanying nurses including their subsistence allowances. (b) The transport leaves late and arrives late at South African hospitals - at times in the early hours of the morning. This is partly due to the large number of hospitals in the Gauteng area that the patients are referred to. There have been instances where the patients sleep in the vehicle prior to admission. (c) The transport facility is only available on Tuesdays and Thursdays from Mbabane with the return trip on Wednesdays and Fridays except for emergencies. Thus some patients who are discharged on a Saturday wait till the next Wednesday to return and thus incur additional costs of hospitalization mainly related to accommodation needs. (d) Some patients receiving chemotherapy or other procedures on an outpatient basis may find the long journey tiring. (e) The patients do not get any subsistence allowances. (f) The civil servants either use their transport or the line ministry provides the transport. The civil servants are reimbursed for the use of their vehicle according to set government rates. 8. The management information, administrative, financial and clinical systems are wholly inadequate given the levels of expenditures. Figure 2 and Table 4 show the level of funds for the two schemes (SE 51 million). Despite this magnitude of financial flows, the system is almost completely manual. These many manual procedures and registers that are filled in which adds to the bureaucracy but does not give appropriate information nor linkages between the clinical assessment and the claims. Figure 2: Expenditure on the Phalala Fund and the Medical Referrals Scheme, FY02/03 to FY04/05 14 FY 04/05 37 10 Medical Referral FY 03/04 37 Phalala 5 FY 02/03 28 Expenditure (SE m) 0 10 20 30 40 Source: Personal interview with Ministry of Finance and review of Treasury Annual Reports (2002/03, 2003/04 and 2004/05). 12 Table 4: Comparative Expenditure by the Phalala Fund and the Medical Referrals Scheme Against Curative Medicine Expenditures by the MOHSW for the General Population, FY02/03 to FY04/05 Items FY 2002/03 FY 2003/04 FY 2004/05 (SE million) (SE million) (SE million) Curative Medicine19 (Vote 45, Health and Social Welfare) 180 206 246 Phalala Fund 28 37 37 Medical Referrals Scheme 5 10 14 Total 33 47 51 % of Curative Services 18% 23% 21% Source: Treasury Annual Reports (FY 2002/03, FY 2003/04, FY 2004/05). 9. The social support infrastructure is inadequate. While the medical and health needs of the patients are being met, their social and pastoral care is non-existent. (a) Patients access the referral system from rural villages; are sent to large hospitals without family support in a foreign city; and are subjected to multiple clinical investigations. All these create additional stress on the patients, who are already stressed to begin with because they are seriously sick. Also, patients have to fund themselves to get to MGH or from MGH back to their homes. (b) The application for travel documents has to be done by the patient or family members with various fees being charged. (Fortunately, the application for a medical visa to South Africa is free). (c) The Phalala fund personnel and the MGH try to assist the patients with some social support activities viz., contacting families, arranging temporary accommodation in MGH etc. 10. Both funds benefit a small part of the population. In the case of the Phalala Fund, when the annual expenditure cap is reached, no more patients are accepted nor referred20. (In November 2006, the government temporarily stopped further referrals under the Phalala Fund). This creates moral and ethical dilemmas especially for patients needing repeat visits and therapeutic interventions according to set protocols. 19 Expenditure on ‘Curative Medicine’ was taken rather than the total Health expenditure in the MOHSW (Vote 45) records as this reflects the equivalent services paid for by the Phalala Fund and the Medical Referrals Scheme. 20 http://www.observer.org.sz/main.asp?id=27003&Section=main&articledate=Monday, September 18, 2006 13 B. REVIEW OF PATIENT AND PROVIDER DATA 11. Major difficulties were faced in reviewing the data with respect to the Phalala Fund and the Medical Referrals Scheme. The sources of the data were in different facilities, ministries and with various individuals. The periods of review did not overlap – Phalala Fund (2002 to 2006 with a concentration in the 2005 to 2006 years) and Medical Referrals Scheme (2002 to 2003). In addition, the specific data sources covering Diagnosis and Referral Specialties also covered different periods. The data covered mainly financial information with little clinical data. A great deal of time and effort was spent in entering the data onto spreadsheets and "cleaning" the data to enable some meaningful analysis to be done. Nevertheless, the analysis provides stark and quite worrying findings. 12. The number of "unknown" or uncategorized patients is large, raising issues about the ability of the referral financing schemes to track their patients. For the purpose of this study, claims processing data over the period 2002 to 2006 with concentration on claims processing in years 2005 and 2006 for the Phalala Fund, and years 2002 to 2003 for the Medical Referrals Scheme were assessed .Some 33 claims that were submitted for payment under Phalala were for "unknown" or "various" patients (see Table 5); these entailed a total payment of about SE 2.083 million, or roughly 6.4 percent of the total cost of all claims in Phalala. Similarly, 37 claims that were submitted for the Medical Referrals Scheme were for "unknown" patients, entailing a payment of SE 105,269 representing about 0.7 percent of the cost of all claims in that scheme. It cannot be determined whether these claims in Phalala and in the civil servants' scheme were legitimate or not, but were paid for by the two funds anyhow. Table 5: Overview of the Phalala Fund and the Medical Referrals Scheme, 2000s Phalala Fund Medical Referrals Scheme Average Average Items Total Amount Total No. of Amount No. of Amount of Claims Amount of Claims of Claim Claims of Claim (SE) Claims (SE) (SE) (SE) Known 30,707,452 2,471 12,427 14,300,518 3,196 4,475 Patients Unknown 341,804 28 12,207 105,269 37 2,845 Patients Various 1,740,873 5 348,175 0 0 0 patients Total 32,790,13021 2,504 13,095 14,405,787 3,233 4,456 Period 2002 to 2006 w/ concentration of 2002 to 2003 covered claims processing in the last 2 years Patients 756 known patients 455 known civil servants covered Source: This study 21 Rounding error. 14 13. The average cost per patient is high. Table 6 shows summary information covering both funds. About SE 40,618 (US$5,803)22 was spent on the average Phalala Fund patient and SE 31,430 (US$4,490) on the average civil servant under the Medical Referrals Scheme, with an average of three and seven claims, respectively. Of note, is the "unknown" and "various" category of patients amounting to SE 2.1 million covering 33 claims for the Phalala Fund and SE 105,000 covering 37 claims for the Medical Referrals Scheme. The latter may also be interpreted as the civil servants’ being referred for less severe conditions. If these "unknown" and "various" claims are included, the total expenditures for the period under review for Phalala would be SE 32.8 million for 2,504 claims; for the Medical Referrals Scheme, total expenditures would be SE 14.4 million for 3,233 claims. Table 6: Average Amounts of Claim and Average Number of Claims Per Patient Under the Phalala Fund and Medical Referrals Scheme Average Average No. of Known Expenditures No. of Amount per No. of Patients (SE) Claims Fund Patient (SE) Claims Phalala Fund 756 30,707,452 40,618 2,471 3 Medical Referrals Scheme 455 14,300,518 31,430 3,196 7 Phalala Fund Unkown 2,082,677 33 Medical Referrals Scheme Unkown 105,269 37 Source: This study. 14. Some of the top five disease/injury conditions that are referred represent preventable conditions. For instance, cancer of the cervix (Figure 3) is highly preventable, and the possibility of early diagnosis of cervical cancer through preventive cervical screening programs will reduce costs as well as reduce mortality, morbidity and the psychological effects of the disease. This is a public policy issue and needs prioritization. In addition, there are excellent private sector specialists in gynecology in Swaziland and a first step could be engagement with them to provide advanced gynecology care to such patients. 15. Three of the top 5 referral specialties reflect general specialties23. Figure 4 shows that a third (32 percent) of the top five referral specialties are general specialties, i.e., orthopedics, ophthalmology, and Eye, Nose, and Throat, raising issues about the referral funds’ intended focus on highly specialized care. Orthopedic surgery is a general specialty and warrants an active presence at MGH. Fields such as ophthalmology can be enhanced using telemedicine and cutting down the numbers of referrals, and therefore, the outflow of scarce Swazi resources to a foreign country. 22 Exchange rate of US$1=SE7. 23 This data has to be treated with caution as it reflects the assessment of notes from Dr Ezeogu’s book. There were many patients with no identifiable referral specialty. 15 Figure 3: Top 5 Disease Conditions and Number of Patients Referred Under the Phalala Fund to South Africa, January to November 2006 Hydrocephalus, 5 Eye, 17 Head Injury, 9 Cardiac disease, 9 Ca Cervix, 12 Source: Based on assessment of "Diagnosis of Patients" in Ms. Mumcy Khumalo’s diary in the Phalala office. Figure 4: Percent of Top 5 Referral Specialties Under the Phalala Fund, May to October 2005 Orthopaedic, 6 Ears, Nose & Throat, 16 Urology, 8 Cardiothoracic, 8 Opthalmology, 10 Source: Based on assessment of notes in Dr. Austin Ezeogu’s book. 16. A few very-high-cost patients take up a significant proportion of the total expenditures of the funds. The expenditure amounts for the top four patients in the Phalala Fund were in excess of SE 500,000 per patient, with one patient in excess of SE 1 million (Figure 5). The cost of the top 10 patients’ claims amounted to SE 5.3 million and represented about 16 percent of the total claims expenditure for the Phalala Fund. For the Medical Referrals Scheme, three civil servants’ costs exceeded SE 500,000 with one civil servant’s costs amounting to SE 0.7 million (Figure 6). The cost of the top 10 civil servants’ claims amounted to SE 4.2 million and represented about 30 percent of the total claims expenditure for the Medical Referrals Scheme. 16 Figure 5: Top 10 Referred Patients in Phalala Fund, Their Number of Claims, and Amounts of Claim 1200000 30 Cost of Claims (SE) 1000000 25 Number of Claims 800000 20 600000 15 400000 10 200000 5 0 0 1 2 3 4 5 6 7 8 9 10 Patient Source: This study Figure 6: Top 10 Referred Civil Servants, Their Number of Claims, and Amounts of Claim 800000 100 700000 90 Cost of Claims (SE) Number of Claims 80 600000 70 500000 60 400000 50 300000 40 30 200000 20 100000 10 0 0 1 2 3 4 5 6 7 8 9 10 Civil Servant Source: This study 17. There is no association between the number of claims per referral and the associated costs. Phalala Fund patients had a maximum of 25 claims while one civil servant had 91 claims (Figures 5 and 6, as cited earlier). In general, the civil servants had more claims, reflecting more consultations for probably less severe conditions. It may also reflect referrals of civil servants at an early stage of disease or injury and pressure from the civil servants to be referred. 18. The top 10 referred civil servants who had the highest claims expenditure under the Medical Referrals Scheme came from 11 ministries24. The Ministries that the top 10 referred civil 24 This does not mean that these ministries have greater numbers of ill civil servants. Data are needed about the numbers employed in each ministry over the study review period to give a proper denominator for statistical calculations and ministry specific referral rates and claims profile which may be comparable across ministries. 17 servants came from are reflected in Table 7. It must be noted that there may have been transcription errors during the capturing of the data in the Accountant General’s office or that the civil servants worked for different ministries during the period under review, hence the 11 ministries. It must be noted that the claims expenditure may cover the civil servants and/or their dependents. Table 7: Top 10 Civil Servants’ Amount of Claims Under the Medical Referrals Scheme, by Ministry Cost of Claims Civil Servant Ministry No. of Claims (SE) Information 125 Public Service 670,710 62 Agriculture & Co-operatives 226 Enterprise & Employment 659,790 71 3 Defence 540,511 13 4 Economic Planning & Development 439,315 76 5 Police 429,002 91 6 Foreign Affairs & Trade 321,622 66 Health & Social Welfare 727 Public Works & Transport 310,287 18 8 Police 293,494 41 9 Public Works & Transport 289,492 35 10 Education 284,002 20 Total 4,238,225 493 Source: This study. 19. The case fatality rate of referred patients is high. Analysis of the health outcomes of the referred patients indicates high death rates, many of them after incurring very high costs. For instance, the top 10 most expensive cases28 in the Phalala Fund were assessed as to medical outcome, and the results are shown in Table 8. These top 10 patients cost Phalala SE 5.3 million, but four of the 10 patients died, indicating a case fatality rate of 40 percent. The criteria for referral need re-visiting given that one of the patients was 80 years old, clearly with a medical prognosis not warranting the referral. Though largely illustrative, these results 25 Possible duplicate payments; payment periods overlap with the two Ministries. 26 Payment periods overlap with the two Ministries. 27 Possible duplicate payments or transcription errors. Payment periods overlap with the two Ministries. 28 Note that the selection of these cases was not random; it focused instead on the most expensive cases, which also reflects severity of illness. Clearly, the case fatality rate of 40 percent is extremely high for such an extremely expensive program. It can be surmised, though not proved at this stage, that less expensive cases (ergo, less severe cases) could have better outcomes and lower case fatality rates. Of those presumed alive in this table, it cannot be ascertained how many eventually died soon after the discharge from the hospital. Due to time and data constraints, proper survival analysis could not be undertaken for this study. 18 indicate the urgent need to have a clearer rationing criteria on who and what condition should be eligible to use these funds. While there are clear ethical issues involved in approving or disapproving patients for referral, the principle of opportunity cost should be invoked, since scarce government resources could very well save so many more lives than only a few who incur these costs, and who often end up not in better condition anyhow. It is recommended that a more thorough analysis using disability-adjusted life years be undertaken. Table 8: Age of Patients, Diagnosis, Costs and Outcome of Medical Interventions in the Top 10 Patients Referred Under the Phalala Fund Age Diagnosis Costs (SE) Outcome 23 Major abdominal trauma; road traffic accident 1,026,343 Dead Presumed 42 Rheumatic Heart Disease29 670,858 alive Presumed 43 Cerebro-vascular accident 612,942 alive 80 Cancer of the esophagus 393,373 Dead 37 Road traffic accident 512,376 Dead Presumed 56 Anorectal lesion (cancer of the anus) 477,512 alive Presumed 38 Skin cancer 372,981 alive 11 Head injury/tracheostomy 351,330 Dead Presumed 9 Head injury; road traffic accident 332,916 alive Presumed 18 Chronic osteomyelitis (femur) 324,659 alive Source: This study. 20. Some conditions are being funded under the Phalala Fund, when they could very well be funded by other insurance programs, like the Motor Vehicle Accident and Workmen’s Compensation Funds. From the table above, three patients were referred under the Phalala Fund as opposed to claims being made under the Motor Vehicle Fund. These patients cost the Phalala Fund approximately SE 1.87 million and needs to be recovered from the Motor Vehicle or Road Accident Fund. Another patient that was injured at work was referred under the Phalala Fund and those costs need to be recovered from the Workmen’s Compensation Fund. There may be other instances where the costs are borne by the Phalala Fund as opposed to other funds. 29 Could not access the clinical notes; the diagnosis may be inaccurate. 19 21. Analysis of referral and claim data by each Ministry using the Medical Referrals Scheme shows varying patterns. During the period 2002 to 2003, some 455 known civil servants were referred and cost the scheme SE 14.4 million (Table 9). Using this table, the following observations can be noted: • The Ministry of Education was the top ministry with a 129 referred civil servants (26 percent of referrals) who accounted for 837 claims (26 percent of claims) for which it spent SE 3.5 million (24 percent of costs). This probably reflects the fact that this Ministry is the largest employer in government. • The Ministry of Defense had the highest average amount spent per claim at SE 10,963. This probably reflects the severity of injuries incurred by Defense personnel, although this may not be the case in times of war. • The Ministry of Foreign Affairs and Trade had the highest average cost (SE 161,754) and average number of claims (36) per civil servant30. 22. The large number of South African providers participating in the two funds precludes the possibility of fee negotiation and cost control. The Phalala Fund uses approximately 180 private providers (health practitioners, hospitals, laboratory and radiology) covering a wide geographical area in South Africa, especially in the Gauteng province (Table 10). Even more so, the Medical Referrals Scheme uses about 375 providers, also within a wide geographical area in the Gauteng province (Table 11). The large number of providers clearly favors the patients, who can have as wide a choice of hospitals, laboratories, and doctors as possible. However, from the perspective of funds management, this excessive number of providers engenders serious difficulties. First, it makes claims processing and quality of care monitoring far more challenging than when the funds only deal with much-fewer providers. Indeed, misrouted claims have been reported. There are also anecdotal accounts of the patient-carrying vehicle going around Gauteng looking for the right hospital. But more seriously, a multiplicity of providers makes it nearly impossible for the funds to undertake pro-active fee negotiation. Indeed, the cost of negotiating with 180 providers (under Phalala) and 375 providers (under the Medical Referrals Scheme) would be extremely prohibitive. However, if the two funds were more selective of providers and limit their number to a manageable size, they could negotiate better fees with the fewer "preferred providers" by promising them a bigger volume of patients (i.e., volume discounts operating on the fund side, economies of scale operating from the provider side). 30 Note that there were only 2 civil servants in referred in this Ministry. 20 Table 9: Claims Profile of Ministries Under the Medical Referrals Scheme, by Ministry24 Avg. No. of No. of Avg. Cost Claims Avg. Referred per Civil per % of Cost of % of Civil Servant Civil Total Cost of Total Claim No. of Total Ministry Servants (SE) Servant Claims (SE) Costs (SE) Claims Claims Education 129 27,284 6 3,519,619 24.4 4,205 837 25.9 Police 82 27,570 7 2,260,745 15.7 3,884 582 18.0 Health & Social Welfare 84 19,485 5 1,636,703 11.4 3,637 450 13.9 Public Works & Transport 29 50,010 7 1,450,279 10.1 7,040 206 6.4 Enterprise & Employment 12 85,999 12 1,031,988 7.2 6,973 148 4.6 Information 12 69,159 9 829,912 5.8 7,756 107 3.3 Defense 17 47,075 4 800,267 5.6 10,963 73 2.3 Agriculture & Co-operatives 34 21,491 6 730,704 5.1 3,513 208 6.4 Economic Planning & Development 7 77,740 14 544,180 3.8 5,728 95 2.9 Justice & Constitutional Affairs 19 17,355 5 329,738 2.3 3,435 96 3.0 Foreign Affairs & Trade 2 161,754 36 323,507 2.2 4,556 71 2.2 Treasury & Stores 8 23,213 7 185,700 1.3 3,376 55 1.7 Correctional Services 10 15,555 4 155,549 1.1 4,093 38 1.2 Natural Resources & Energy 10 12,356 4 123,564 0.9 2,874 43 1.3 Deputy Prime Minister 10 10,055 5 100,549 0.7 2,139 47 1.5 Income Tax 4 21,906 11 87,624 0.6 1,991 44 1.4 Finance 3 18,412 8 55,237 0.4 2,302 24 0.7 Parliament31 4 11,181 4 44,723 0.3 2,631 17 0.5 Tourism, Communications & Environment 8 4,881 3 39,049 0.3 1,502 26 0.8 Public Service 3 11,301 4 33,902 0.2 2,825 12 0.4 Geological Survey & Mines 1 29,649 6 29,649 0.2 4,942 6 0.2 Home Affairs 2 14,367 9 28,734 0.2 1,690 17 0.5 Fire & Emergency Services 3 7,268 3 21,804 0.2 2,180 10 0.3 Unknown 3 6,342 4 19,026 0.1 1,730 11 0.3 Housing & Urban Development 3 5,115 2 15,345 0.1 3,069 5 0.2 Customs & Excise 1 7,350 4 7,350 0.1 1,838 4 0.1 Audit 1 340 1 340 0.0 340 1 0.0 Total 50132 28,754 6 14,405,787 100.0 4,456 3233 100.0 Source: This study. 31 Not a Ministry. 32 Includes civil servants captured in multiple ministries as well as "unknown" civil servants whose claims were paid by that ministry. 21 Table 10: Claims Data of Health Providers Under the Phalala Fund, by Type of Provider or Service Avg. Avg. Avg. Avg. No. Amount Amount No. of Amount per of Patients No. of No. of Provider per Amount (SE) per Providers Provider per Patients Claims Patient Claim (SE) Provider (SE) (SE) Health 142 54,942 4 518 15,061 7,801,738 1,273 6,129 Practitioner Hospital 22 1,003,103 15 320 68,963 22,068,272 612 36,059 Laboratory 5 197,378 20 99 9,969 986,891 230 4,291 Radiology 4 194,953 8 33 23,631 779,811 80 9,748 Swaziland 2 230,182 52 103 4,470 460,365 139 3,312 Private Ambulance 2 6,980 4 7 1,994 13,960 11 1,269 Dialysis 2 56,582 4 7 16,166 113,163 21 5,389 Blood 1 529,543 70 70 7,565 529,543 133 3,982 Unknown 2 18,193 36,385 5 7,277 Total 180 182,167 4 759 43,202 32,790,130 2,504 13,095 Source: This study. Table 11: Claims Data of Health Providers Under the Medical Referrals Scheme, by Type of Provider or Service Avg. Avg. No. Avg. Avg. Amount of Civil No. of Amount Amount No. of No. of Provider per Servants Civil per Civil Amount (SE) per Providers Claims Provider per Servants Servant Claim (SE) Provider (SE) (SE) Health 304 17,212 1 367 14,257 5,232,369 2,002 2,705 Practitioner Hospital 45 183,056 6 274 30,064 8,237,524 567 14,504 Ambulance 8 10,443 2 14 5,968 83,546 23 3,632 Laboratory 7 48,145 13 93 3,624 337,012 470 717 Pharmacy 7 11,709 4 31 2,644 81,962 59 1,389 Dialysis 2 96,943 3 6 32,314 193,885 58 3,343 Blood 1 232,598 35 35 6,646 232,598 52 4,473 Unknown 1 6,891 2 2 3,446 6,891 2 3,446 Total 375 38,415 1 460 31,317 14,405,787 3,233 4,456 Source: This study. 23. As expected of a referral financing program, claims from the two funds are heavily oriented towards hospitals. In the Phalala Fund, hospital claims represented 67 percent of the claims (Figure 7) and cost SE 22 million (Table 10, cited earlier). The health practitioners, who provided the care, cost SE 7.8 million (24 percent). Laboratory costs were 3 percent. In the Medical Referrals Scheme for civil servants, hospital claims represented 57 percent of the claims (Figure 8) and cost SE 8.2 million (Table 11, cited earlier). The health practitioners who provided the care cost SE 5.2 million (36 percent). Laboratory costs were 2 percent. 22 Approximately SE 300,000 was spent on dialysis for both funds, a service which is currently provided at MGH. Figure 7: Percent of Claims by Type of Provider Under the Phalala Fund Other Providers 9% Health Practitioners 24% Hospitals 67% Source: This study. Figure 8: Percent of Claims by Type of Provider Under the Medical Referrals Scheme Other Providers, 6% Health Practitioners, 36% Hospitals, 57% Source: This study. 24. The two referral funds do not use the public health sector in South Africa, and rarely use the local private sector in Swaziland, even if it is possible to do so. The Phalala Fund incurred just above 1 percent (SE 460,000) on the private health sector in Swaziland. The overall average cost per claim was SE 13,000 with hospital claims averaging SE 36,000. A health practitioner claim at SE 6,100 per claim is quite high and may reflect over servicing and/or high level specialist interventions. 23 25. The extremely low utilization of local health services may indicate over-referrals. Of note is the SE 113,000 spent on dialysis with the average dialysis claim costing SE 5,400. The fact is that there are dialysis services provided at Mbabane Government Hospital at present. Of course, these funded episodes may have been acute interventions for patients while in South Africa for other medical interventions. Under the Phalala Fund, Ears, Nose and Throat (ENT) was the top referral specialty followed by ophthalmology in the top five specialties that patients were referred to in South Africa. However, it must be noted that there is an ophthalmology clinic at Good Shepherd Hospital, and the MGH eye wing itself could be utilized by the Good Shepherd doctor, if a public/private partnership model were adopted. 26. The average cost of claims and hospital-to-practitioner claims cost ratios of the two funds indicate that the typical civil-servant patient has probably a less severe illness or of more ambulatory nature than the typical Phalala patient. Tables 12 and 13 show, respectively, the claims profiles of the Phalala Fund and the Medical Referrals Scheme. The average cost per claim for the Medical Referrals Scheme was SE 4,511 at 2002/03 prices. Hospital claims average SE 14,528 and health practitioner claims average SE 2,705 (Table 13). Even with medical inflation at 15 percent, this shows that the referred civil-servant-patients were probably referred for less severe conditions than the Phalala Fund patients. In addition, the hospital-to- health practitioner claims cost ratio of 2.8 : 1 in the Phalala Fund versus 1.6 : 1 in the Medical Referrals Scheme allude to more ambulatory or outpatient consultations being done rather than inpatient or hospital care. Finally, the average laboratory costs for the Phalala Fund were SE 4,300 while average laboratory costs to the Medical Referrals scheme were SE 720 (Table 14). Of note, the hospital-to-health practitioner ratios for average costs per claim were equivalent at 5.4 : 1 for both funds. All costs were at least 50 percent below that of the Phalala Fund except for ambulance services where the costs were trebled. Finally, the civil servants were typically referred to general specialty hospitals while the Phalala patients were typically referred to high - level hospitals, such as the Heart Hospital. Table 12: Amount, Number, and Average Cost Per Claim Under the Phalala Fund, by Type of Provider No. of Avg. Cost per Provider Amount (SE) % Claims Claim (SE) Hospital 22,068,272 67.3 612 36,059 Health Practitioner 7,801,738 23.8 1,273 6,129 Laboratory 986,891 3.0 230 4,291 Radiology 779,811 2.4 80 9,748 Blood 529,543 1.6 133 3,982 Swaziland Private 460,365 1.4 139 3,312 Dialysis 113,163 0.3 21 5,389 Unknown 36,385 0.1 5 7,277 Ambulance 13,960 0.0 11 1,269 Total 32,790,130 100.0 2,504 13,095 24 Table 13: Amount, Number, and Average Cost Per Claim Under the Medical Referrals Scheme, by Type of Provider Number of Average Cost per Provider Amount (SE) % Claims Claim (SE) Hospital 8,237,524 57.2 567 14,528 Health Practitioner 5,232,369 36.3 2,002 2,705 Laboratory 337,012 2.3 470 717 Blood Transfusion 232,598 1.6 52 4,473 Dialysis 193,885 1.3 58 3,343 Ambulance 83,546 0.6 23 3,632 Pharmacy 81,962 0.6 59 1,389 Unknown 6,891 0.0 2 3,281 Total 14,405,787 100.0 3,233 4,511 Source: This study. Table 14: Laboratory Claims and Costs Under Both Funds Phalala Fund Medical Referrals Scheme Average Average Cost per Number Cost per Number of Claim Amount of Claim Lab Amount (SE) Claims (SE) Lab (SE) Claims (SE) 1 734,565 89 8,254 3 143,347 297 483 2 247,916 129 1,922 1 115,152 107 1,076 3 3,230 8 404 2 66,937 51 1,312 4 602 3 201 6 4,874 9 542 5 579 1 579 7 3,960 5 792 8 2,743 1 2,743 Total 986,891 230 4,291 Total 337,012 470 717 Source: This study. 27. Claims data show a few providers with significant number of claims, and a large number of providers with extremely small number of claims. For the Phalala Fund, seven hospitals, two medical practitioners, and one laboratory group were in the top 10 providers, based on their total claims expenditure profile (Figure 9). About 25 percent of the total costs were from Hospital A33 alone and amounted to approximately SE 8 million. These top 10 providers accounted for 72 percent of the claims expenditure which amounted to SE 23.6 million. For the Medical Referrals Scheme, there were six hospitals and four doctors in the top 10 provider 33 The health providers (hospitals, practitioners and laboratories) are re coded and labeled differently to protect anonymity for the purposes of this assessment. 25 category. About 23 percent of the total costs were from Hospital C alone and amounted to SE 3.3 million (Figure 10). Hospital A, which was the top hospital in the Phalala Fund (SE 8 million) was tenth in the Medical Referrals Scheme (SE 250,000). Dr ABC was number 5 in the Phalala Fund (claiming SE 1.8 million) and number 5 in the Medical Referrals Scheme (claiming SE 418,000). These top 10 providers accounted for 50 percent of the claims expenditure which amounted to SE 7.3 million. The rest of the providers to the two funds, however, have small numbers of claims and claims expenditures relative to these top 10 providers. Figure 9: Top 10 Health Providers Under the Phalala Fund 9000000 400 No. of Claims Cost of Claims (SE) 300 6000000 200 3000000 100 0 0 A F D lG os A os B os E os C D D os BC ra D bo C H ital H tal H ital H tal ry H tal l ta ta rA La r B to pi pi pi pi pi p p os os H H Health Provider Source: This study. Figure 10: Top 10 Health Providers Under the Medical Referrals Scheme 4000000 250 Cost of Claims (SE) 200 No. of Claims 3000000 150 2000000 100 1000000 50 0 0 pi I H al H D lJ os K lA os G os C D BC D D D E os l H pita H EF D C H ital ta H ital ta rA rC rB t pi pi p os r p os H Health Provider Source: This study. 26 28. The claims profile of a few providers (especially individual practitioners) with large amounts of reimbursements from the two funds should warrant further investigation. With respect to the Phalala Fund, Hospital G had the highest average cost per claim (SE 174,000) (Statistical Annex Figure 1 and Table 15). This would be expected as the facility is a Heart hospital. The other hospitals and doctors have great variance in their average costs and will depend on the type of facility, the diagnosis and the interventions. Dr ABC’s cost claim average for the Medical Referrals Scheme was half that of the Phalala Fund at SE 15,000 and SE 30,000, respectively. Dr ABC is a neurosurgeon and this may account for his high cost average. In the Medical Referrals Scheme, Dr. BCD has an average claim cost profile of SE 24,000 which is relatively high and may need further investigation. Table 15: Top 10 Health Providers Under the Phalala Fund, by Their Number of Claims and Average Cost Per Claim Health Provider Average Amount Per Claim No. of Claims Hospital G 173,971 6 Hospital B 80,396 55 Hospital C34 63,815 36 Hospital D 36,070 50 Hospital F 31,849 35 Dr ABC35 29,031 61 Hospital A34 26,248 307 Hospital E 18,979 73 Laboratory A 8,254 89 Dr BCD35 7,595 125 Source: This study. 29. There was very little overlap between the referral patterns of the two funds, except for the top referral doctors. This may partly be due to the different reporting periods for the two funds in this assessment. Six medical doctors overlapped with both funds (see Statistical Annex Figure 2) with the top two being the same doctors in both funds (Dr. ABC and Dr. BCD) (Table 15, cited earlier, and Table 16). There are two more doctors in the top 10 providers to the Medical Referrals Scheme. 30. The large number of providers with extremely low number of claims poses large transactions costs. Based on the log scale (See Statistical Annex Figure 3), 15 hospitals in the Phalala Fund had less than 10 claims with 7 hospitals having only 1 claim each. Similarly, 34 hospitals in the Medical Referrals Scheme had less than 10 claims with 14 hospitals having 1 claim each. The transactions costs associated with many providers who have few claims may not justify the referral unless these hospitals are the only providers of such highly specialized services. 34 Same hospital in both funds. 35 Same medical practitioner in both funds. 27 Table 16: Top 10 Health Providers Under the Medical Referrals Scheme, by Their Number of Claims and Average Cost Per Claim Health Provider Average Amount per Claim No. of Claims Hospital H 35,221 31 Hospital C34 24,436 135 Dr BCD 23,797 17 Hospital I 23,189 22 Dr ABC35 15,478 27 Hospital K 9,460 28 Hospital J 7,935 59 Hospital A34 4,740 52 Dr EFG 4,427 60 Dr CDE 1,332 218 Source: This study 31. There is a large variation in the average claim cost profile between the hospitals in the Phalala Fund and the Medical Referrals Scheme. There were 22 hospitals in the Phalala Fund with a mean claim value of SE 1.0 million per hospital, as opposed to 45 hospitals in the Medical Referrals Scheme with a mean claim value of SE 183,400 per hospital (Table 17). The higher mean claim value for Phalala probably reflects the sicker patients being referred under that scheme. Thirteen hospitals overlapped both funds (see Statistical Annex) and the possibility exists to consolidate the “preferred provider� arrangements with some of those hospitals as a starting point. Table 17: Amount, Number, and Average Cost Per Claim of Hospitals in Both Funds, Summary for All Hospitals Phalala Fund Medical Referrals Scheme (n=22 hospitals) (n=45 hospitals) Summary Statistics Average Average No. of Amount No. of Amount (SE) Amount per Amount per Claims (SE) Claims Claim (SE) Claim (SE) Mean 1,003,103 28 64,388 183,400 13 16,476 Maximum 8,058,224 307 413,741 3,300,072 136 83,577 Minimum 1,181 1 236 1,056 1 592 Median 107,072 4 29,882 43,444 3 10,594 Source: This study. 28 32. There is a large variation in the average cost per patient in the Phalala Fund and the Medical Referrals Scheme. The average amount spent per patient under the Phalala Fund is SE 43,202 as opposed to SE 31,317 under the Medical Referrals Scheme (Tables 18 and 19), and the difference is driven by the twice-higher hospitalization costs of Phalala. This probably reflects higher morbidity or more serious patients being referred under the Phalala Fund as opposed to the Medical Referrals Scheme. Alternatively, it could also reflect much better amenities enjoyed by Phalala patients, compared to civil servants36. The average cost per patient by doctors alone (i.e., individual health practitioners) is almost the same in the two funds (SE 14,000 to 15,000). There was a 1:1 ratio of patients to health practitioner in the Medical Referrals Scheme compared to a 4:1 ratio in Phalala. The average cost per hospital was SE 1.0 million in the Phalala Fund as opposed to SE 183,056 in the Medical Referrals Scheme. There was a 15:1 ratio of patients to hospitals in the Phalala Fund as opposed to a 6:1 ratio of civil servants to hospitals. Table 18: Provider and Patient Referral Claims Profiles for the Phalala Fund Avg. Avg. Avg. No. Avg. Amount No. of Amount per of Patients No. of No. of Provider Amount per Amount (SE) per Providers Provider per Patients Claims Patient (SE) Claim (SE) Provider (SE) Health Practitioner 142 54,942 4 518 15,061 7,801,738 1,273 6,129 Hospital 22 1,003,103 15 320 68,963 22,068,272 612 36,059 Laboratory 5 197,378 20 99 9,969 986,891 230 4,291 Radiology 4 194,953 8 33 23,631 779,811 80 9,748 Swaziland Private 2 230,182 52 103 4,470 460,365 139 3,312 Ambulance 2 6,980 4 7 1,994 13,960 11 1,269 Dialysis 2 56,582 4 7 16,166 113,163 21 5,389 Blood 1 529,543 70 70 7,565 529,543 133 3,982 Unknown 2 18,193 36,385 5 7,277 All 180 182,167 4 759 43,202 32,790,130 2,504 13,095 36 Without additional corroborating information, it is difficult to analyze and make definitive conclusions about the differences in costs between Phalala and Medical Referrals Scheme patients. The conventional wisdom is that higher costs are a reflection of higher severity. However, if amenities are taken into account, such costs of amenities dilute the higher costs of severity. 29 Table 19: Provider and Patient Referral Claims Profiles for the Medical Referrals Scheme Avg. No. Avg. Avg. Avg. of Civil No. of Amount per Amount No. of Amount per No. of Provider Servants Civil Civil Amount (SE) per Providers Provider Claims per Servants Servant Claim (SE) Provider (SE) (SE) Health Practitioner 304 17,212 1 367 14,257 5,232,369 2,002 2,705 Hospital 45 183,056 6 274 30,064 8,237,524 567 14,504 Ambulance 8 10,443 2 14 5,968 83,546 23 3,632 Laboratory 7 48,145 13 93 3,624 337,012 470 717 Pharmacy 7 11,709 4 31 2,644 81,962 59 1,389 Dialysis 2 96,943 3 6 32,314 193,885 58 3,343 Blood 1 232,598 35 35 6,646 232,598 52 4,473 Unknown 1 6,891 2 2 3,446 6,891 2 3,446 All 375 38,415 1 460 31,317 14,405,787 3,233 4,456 Source: This study. 33. The number of claims for some patients is excessive and opens the possibility for fraud. (See Statistical Annex Figure 5). In Phalala, the mean number of claims was three, but one patient had 27 claims, the maximum number found in this study (Table 20). Some 94 percent of the patients had less than 10 claims per patient while 46 percent of the patients had one claim. For Phalala, the mean cost of a claim for was SE 40,995, with the maximum at SE 1.0 million. However, one claim had a value of SE 25. For the Medical Referrals Scheme, the mean number of claims was seven with a maximum of 91 claims for one civil servant. Some 80 percent of the civil servants had less than 10 claims per civil servant while 31 percent had one claim. The mean cost of a claim for each civil servant was SE 31,430 with a minimum claim of SE 76. Keeping track of all these numerous claims in both funds is onerous, given the manual system being used, the under-staffing, and the lack of professional claims processing system. Numerous claims, some with extremely low values, also leaves open the feasibility for misrouting, double-invoicing, double-payment, and other inadvertent mistakes, if not intended fraud. Finally, the claims process may be at great administrative and social cost to the MOHSW and the other line ministries to send one patient to one practitioner in South Africa for one consultation (resulting in one claim) in many instances. Table 20: Amount, Number, and Average Cost Per Claim of Referred Patients from Both Funds Summary Phalala Fund Medical Referrals Scheme Statistics (n=757 known patients) (n=455 known civil servants) Amount No. of Average Amount No. of Average (SE) Claims Amount per Claims Amount per Claim (SE) Claim Mean 40,995 3 10,741 31,430 7 4,405 Maximum 1,026,343 27 205,269 670,710 91 51,710 Minimum 25 1 25 76 1 76 Median 9,656 2 4,416 7,784 3 2,243 Source: This study. 30 C. ANALYSIS OF THE INDIRECT COSTS OF REFERRAL 34. Assumptions. This study made an assessment of the indirect costs of the two referral funds, using subsistence and travel costs to South Africa. An average of SE 50037 per day for subsistence and SE 1.0038 per km. for travel was used. The following were also assumed: (a) 25 percent of the number of claims for both funds was used as the base figure to reflect consultations or visits to South Africa. (b) Many claims would be for 1 visit to South Africa as that visit would entail medical assessment, laboratory investigations, hospitalization and medication. (c) In the case of the Phalala Fund, a nurse and driver would accompany the patients. (d) In the case of the Medical Referral Scheme, the subsistence would be for the civil servant. (e) The Phalala Fund patients usually travel by a 16-seater vehicle. (f) Each visit entailed 1 day. (g) The distance used was 360 km (Mbabane to Johannesburg); this is likely to be exceeded on all occasions by both Phalala Fund (multiple health facilities being visited in South Africa) on any one visit and civil servants visiting facilities not only in Johannesburg. Fuel and toll fees at SE 700 per return journey are included. 35. Results. Based on the above assumptions, approximately SE 1.5 million and SE 1.6million was spent on subsistence and travel for the Phalala Fund and the Medical Referrals Scheme respectively (Table 21). Table 21: Subsistence and Travel Costs for Both Funds Phalala Fund Medical Referrals Scheme No. of Claims 2,504 No. of Claims 3,233 25% 626 25% 808 Nurse + Driver (SE 500 per day) 1,000 Civil Servant (SE 500 per day) 500 Distance (km) (one way) 360 Distance (km) (one way) 360 Fuel & Toll Fees (SE 700 per day) 700 Fuel & Toll Fees (SE 700 per day) 700 Subsistence 626,000 Subsistence 404,125 Travel 450,720 Travel 581,940 Fuel & Toll Fees 438,200 Fuel & Toll Fees 565,775 Total 1,514,920 Total 1,551,840 Source: This study. 37 Current rates are US$120 per day; used a conservative figure of SE 500 because of time periods. 38 This is a conservative figure and is probably a minimum amount used only for cars. 31 IV. OPTIONS FOR IMPROVEMENT OF THE TWO SCHEMES A. THE ASSESSMENT FINDINGS AS BASIS FOR CHANGE 1. Summary. At present, both funds are a net drain on the government budget with little, if any, impact evaluation of the expenditures to date on the real outcomes of treatment (i.e., in terms of disability adjusted life-years of patients who were referred39). This study has shown that approximately SE 60 million in direct health costs from the Phalala Fund and Medical Referrals Scheme will be spent in South Africa with a negligible amount to be spent in Swaziland. The expenditures of the two funds are almost exclusively for the private health sector in South Africa. The funds are non-contributory with a full subsidy from government either through the MOHSW or the line ministries. The administrative support is geared towards logistic and payment activities (mainly claims processing); there is a need to extend this to social support, utilization review, case management, and claims profiling activities, which are just as important. The criteria for referral and coverage need re-assessment; an ex-post examination of a few cases revealed that some of them should not have been referred either because the medical prognosis is bad, or because the cases could be treated domestically. The provider base is too extensive and widely dispersed geographically. Under existing practices, no pro-active provider negotiation exists on patient fees and volumes as embodied in contracts, perhaps because the roles of the payor (Ministry of Finance) and medical personnel doing the referring are completely bifurcated. In addition to the large direct costs, the indirect costs are significant - subsistence and travel costs amounted to SE 3.1 million, at minimum - and these are borne by the MOHSW, other line ministries, and in many instances by the patients and their families. 2. The clinical assessments of patient and provider data show that: • The expenditure at SE 41,000 per Phalala Fund patient and SE 31,000 per civil servant under the Medical Referrals Scheme seems excessive and benefits a small part of the general population (757 known citizens and 455 known civil servants). • Very little "work–up" or "back-referral" occurs with Phalala patients or civil servants which necessitated many visits to South Africa. This may be due to the lack of facilities in Swaziland. • A large number of patients in both funds are sent for 1 consultation to South Africa – either they are too ill and cannot have further treatment in South Africa or alternatively, they are sent back for treatment in Swaziland, or their condition was not that severe in the first instance. (Such referrals will result in high transaction costs). • Mainly "acute" or "late-stage" patients are sent under the Phalala Fund with little rationing in place. The Medical Referrals Scheme sends many civil servants for consultations only as opposed to hospitalization. While no review of the clinical files of the civil servants has taken place, in many instances, the civil servants seem to be referred for less serious conditions to South Africa. 39 This is important and feasible to do, but would require a much-longer time frame of analysis. It would also require ethical approval of reviewing the medical records of patients and their outcomes, and interviewing a sample of the surviving patients and their families, if they have subsequently died. 32 • Some patients in the Phalala Fund ought to have been cared for by funds from the Motor Vehicle Accident Fund or Workmen’s Compensation Fund. 3. Referral financing or hospitalization insurance? Because this assessment only dealt with the referral schemes per se, a key issue that was not analyzed is whether it is more cost-effective to fund the whole range of hospitalization (including cases that do not require referrals abroad), or whether the government should focus only on referrals abroad. The problem with the latter (i.e., existing arrangement) is that the schemes tend to "wait" until the condition of the patient has deteriorated so much before s/he can enjoy the benefit of being referred. Moreover, at this stage, the situation of the patient has gotten so grave that the costs will surely become prohibitive. Converting the schemes into full-blown health insurance schemes will allow coverage of non-referral cases, thus the patient can be cared for at an earlier stage of the disease progression. If preventive and promotive services are included in the package, as in most managed-care (health maintenance organizations) schemes, this will also widen the benefits of those covered under the schemes. B. PRINCIPLES TO GOVERN CHOICE OF OPTIONS FOR IMPROVEMENT 4. "Shared risks" model. In an ideal system, the patients40, the providers41 and the funders42 are sharing the risk and ensuring an optimum health status of the individual with relevant and appropriate health care interventions (Figure 11). In non-ideal systems, patients may seek unnecessary or inappropriate care; there may be "over-servicing" or inappropriate interventions by providers (a phenomenon of perverse incentives known as supplier-induced demand); funders may restrict or ration using "blunt" instruments, e.g., restriction of access due to exhausted funding, delay in authorizations, or simply refusing to pay the provider. (The shared risk model is applicable in most public or government funded health systems where the funder and provider is government; a private not-for-profit example is the Kaiser Health System in California). 40 The Patients are the persons referred under the Phalala Fund and Medical Referrals Scheme. 41 The Providers are the hospitals, health practitioners, laboratory, radiology, pharmacy, etc. that assess, intervene, or care for the patient. 42 The Funder is the Phalala Fund and Medical Referrals Scheme. 33 Figure 11: Principal Players in a Health Care System Patient Heath System Funder Provider 5. Social health insurance. In a small country like Swaziland, a medium-term approach should be towards setting up a social health insurance (SHI) fund which will operate on health insurance principles – key principles being risk-pooling, social solidarity, and sustainability of the financing of health care. (Ideally all persons should be covered under a National Health Insurance Fund. However, the large numbers of unemployed, migrant, informal, and self- subsistent persons make it difficult to get contributions from as well as enrolments into the fund). Thus, SHI becomes a more feasible goal with all employees in the formal sector (having a tax number) being covered. The government is a key player in the formal economy in Swaziland and can kick-start the SHI by converting the Medical Referrals Scheme to a Government Employee Medical Scheme (GEMS). With time, it can invite other parastatals, state enterprises as well as some of the larger corporations to join. Ideally, all sectors including the small and medium enterprises should be encouraged to become part of the SHI. 6. "Buy-in" from stakeholders. The other key stakeholders are the trade unions and their "buy-in" to the concept is vital to the success of setting up a GEMS. (A workshop has taken place in November, 2006 that sensitized them to the concepts of health insurance; a further workshop is planned for early 2007). 7. Sustainability of financing. The proposed health financing mechanism can lead to greater sustainability of health care provision through defined contributions of both the government (as employer) and civil servants. Service provision can be done by the government health system, the private (for-profit and not-for-profit) sectors, or through public-private partnerships. Swaziland has a private health sector, albeit small, both in the for-profit and not-for- profit sectors. There are also elements of health provision at MGH that can benefit from economies of scale, e.g., the Intensive Care Unit, laboratory, pharmacy and radiology facilities. Telehealth 34 support to the public and private health sectors in Swaziland can assist with diagnosis as well as medical care and support. 8. Alternative models of service delivery. Alternative health care financing and provider models that can be developed in Swaziland include: • Public financing and public provision of health care with the re-tooling and expansion of MGH facilities, technology and personnel; • Public financing and private provision through the use of the existing and/or expanded private health sector in Swaziland; • Public financing and public/private partnerships through the use of joint facilities and personnel from both the public and private health sectors43. 9. "Localizing" the spending. As far as possible, the expenditures flowing from both funds should circulate within Swaziland. This would lead to further expansion of the health sector in Swaziland with job opportunities for other health practitioners and support staff. Indirectly, the Swazi tax revenues will be enhanced; opportunity costs (accompanying nurses and drivers plus vehicles) would be reduced; and patient and family support would be enhanced given the proximity to a "home" health service base in Mbabane or Manzini. 10. Principles for continuing referrals. Where necessary, some referrals to South Africa can still continue under the following conditions: • The two funds must put in place, jointly or separately, a system with a limited pool of providers contracted to provide care to referred Swazi patients under a preferred provider arrangement with negotiated fees, discounted on the prospect of a large volume of patients44. • Except for emergencies, a local unit based at MGH should manage patient work-up and back referrals. • The criteria for referrals should be revised based on need, diagnosis, age, prognosis of medical condition, and availability of health services within Swaziland, whether in the government or in the private sector. Moreover, given the political nature of referral decisions, the main responsibility of referring should not be lodged in a single individual. "Second opinion" arrangements should be further utilized. 11. HIV/AIDS. Given the high prevalence of HIV/AIDS in Swaziland, special mechanisms should be put in place within the SHI to support those civil servants and their dependents that need to access health services for HIV/AIDS-related conditions. 43 Examples of these public/private partnerships were identified in the earlier report on "General Assessment of the Mbabane Government Hospital". 44 In this case, the two funds individually or jointly act as monopsonist(s). 35 C. NEEDED ACTIONS Short-Term 12. Governance. Ensuring effective governance of both funds by ensuring that the legal, financial and administrative provisions are met. Both funds should become semi-autonomous with a medium-term view of becoming fully autonomous with reporting lines via the MOHSW and/or Ministry of Public Service and Information (MOPSI). (Action: Planning Office, MOHSW, MOPSI, MOF) 13. Administration • Creating one structure to oversee all administrative functions of both funds – bookings, logistics, transport, claims processing, and payment. (Action: Planning Office, Phalala Office, Accountant General) • Greater efficiency in administration, logistics and transport through capital investments in physical space, computerized management information systems, filing and claims processing systems, furniture and fittings. (Action: Planning Office, Dr. Ezeogu, Phalala Office) • Separating out the Phalala Fund and the Medical Referrals Scheme as separate trading accounts within the MOHSW and MOPSI respectively with the Ministry of Finance having an oversight function. (Action: Planning Office, MOHSW, MOPSI, MOF) • Dedicated human resource personnel with a doubling of the current administrative staff and the addition of a social worker. (Action: Planning Office, Phalala Office) • Recovery of funds from the Motor Vehicle Accident Insurance Fund and the Workmen’s Compensation Fund. (Action: Planning Office, Consultant, Phalala Office) 14. Clinical management • Dedicated senior medical officer with one senior professional nurse to manage the process – utilization review, case management, conduct of audits, and negotiations with providers. (This is different from the Medical Referral Board which has other functions). (Action: Planning Office, Dr. Ezeogu, Consultant) • A support office to be based at the Swazi High Commission in South Africa comprising a senior professional nurse, a social worker, and a driver plus infrastructure to provide administrative, logistic and case management support for referred patients. (Action: Planning Office, High Commission) • Development of a halfway house or step-down facility arrangement in South Africa and Swaziland which will enable timely discharges, back referrals, and elective patient work-ups at 36 lower cost. This can be done in partnership or by contract with a local South African provider and at MGH. (Action: Planning Office, Consultant) 15. Preferred provider arrangements. Professional consultation and hospital fees as well as other costs of health care interventions should be negotiated with a limited pool of providers (ambulance, hospitals, health practitioners, radiology, laboratory, health technology and pharmacy) in a localized area in South Africa as well as within Swaziland. These arrangements should be embodied into various contracts and pre-approved negotiated fees. (Action: Planning Office, Consultant, Dr Ezeogu, Legal Team) 16. Building local capacity • Conduct of a gap analysis of health service provision (public and private sectors) within Swaziland to ensure adequate coverage for referral of patients and civil servants within Swaziland. (Action: Planning Office, Consultant) • Strengthening of the eight general specialties at MGH (Medicine, Surgery, Orthopedics, Obstetrics and Gynecology, Pediatrics, Anesthetics, Family Medicine, and Psychiatry), the support specialties of Clinical Pathology (Laboratory), Radiology, Intensive Care and Pharmacy through human resource (SE 10 million) and capital investments (SE 5 million). The human resources include high-level professional nurses as well as specialist-doctors. This can be enhanced by allowing private practitioners to use MGH as well as MGH doctors to work in the private sector. (Action: Planning Office, Consultant, Dr. Ezeogu, Swaziland Medical Association, MGH doctors, and senior management) • Negotiate back-referral support drugs with pharmaceutical manufacturers and ensure support infrastructure at MGH or within private pharmaceutical sector in Swaziland. (Action: Planning Office, Consultant, Dr. Ezeogu, MGH Pharmacy) • Provide training and support (information technology and software packages; claims processing, utilization review, case management, and preferred provider negotiations, arrangements and contracts; and clinical support/telehealth). (Action: Planning Office, Consultant) 17. Funding for improvement. Funding (SE 18 million) for the short-term initiatives can be made by a special allocation under the Phalala Fund or special additional appropriation by government. Thus, the Phalala Fund will still have approximately SE 17 million for recurrent costs. In addition, the financial resources from the Medical Referral Scheme can bolster the overall capacity of the administrative and clinical service interventions through local referral of civil servants and revenue from their claims on a full-cost recovery basis. 37 18. Guiding principle to use local capacity. As the local capacity is built up within Swaziland, all patients (Phalala Fund and Medical Referrals Scheme) will have to access local (Swaziland- based) services prior to referral to South Africa. 19. Timeframe and budget. The above recommendations and their budgetary implications are summarized in Table 22. Table 22: Timeframe and Budget for Short-Term Interventions Short-Term Objectives By Whom Timeframe Budget (SE) Strengthen governance of Planning office (PO), Jan – March 200,000 funds MOHSW, MOPSI, MOF, 2007 Administration PO, MOPSI, MOF Jan – June 2007 1,000,000 (this 1. Create one includes salaries for structure for both additional staff for funds 1 year) 2. Ensure efficiency gains – capital investments 3. Create separate trading accounts 4. More human resources + social worker 5. Recover funds from Motor Vehicle Accident and Workmen’s Compensation Funds Clinical Management PO, Dr. Ezeogu, High Jan – June 2007 1,500,000 (this 1. Dedicated Senior Commission includes salaries for Medical Officer 1 year plus travel and Senior Nurse and subsistence for 2. Support office at provider High Commission agreements but in South Africa excludes provision (Nurse, social of care at step down worker & driver) facility) 3. Develop step down facility in South Africa & MGH Develop preferred PO, Dr. Ezeogu, MOF, Jan – June 2007 300,000 provider arrangements in Legal team South Africa & 38 Swaziland Build local capacity PO, Consultant, Dr. Jan – March 15,000,000 1. Gap analysis of Ezeogu, Senior MGH 2007 (including capital local provision of management, Senior investments (SE 5 services (public & Nurse managers, million) and private sectors) Swaziland Medical salaries (SE 10 2. Strengthen Association, Swaziland million) for 1 year) specialties at private health sector MGH (capital, human & technological resources) 3. Strengthen back referral support drugs 4. Provide management information systems 5. Develop health provision and financing models 6. Provide training and support to clinical and administration staff including Telehealth Total Costs SE 18,000,000 20. Technical assistance consultant or project. The above actions are comprehensive and technically complex. They require specialist knowledge in several areas. Thus, there is a need for dedicated consultant(s) to undertake the tasks. The cost of this technical assistance support was not calculated as part of the above budget. It is assumed that this technical assistance (TA) cost would be funded by the government itself through its own budget, or funded through donor funds, or some other source. 39 ANNEX 1: List of Persons Interviewed Name Designation Ministry Minister Majozi Sithole Minister Finance Minister Njabulo Mabuza Minister Health and Social Welfare Ms. N. Dlamini Permanent Secretary Health and Social Welfare Ms. Muntu Mntungwa Under Secretary Health and Social Welfare Victor Ndlangangamandla Under Secretary Public Service and Information Titus Khumalo A/AS Public Service and Information Felton Mhlongo Principal Human Resource Public Service and Officer Information Nellie Dlamini Director Public Service and Information Cyril Kunene Permanent Secretary Public Service and Information Thabsile Mlangeni Principal Finance Officer Finance Sanet Mzungu Senior Finance Officer Finance Nina Dlamini Principal Finance Officer Finance Meshack Shongwe Acting Permanent Secretary Finance Khabona Mabuza Deputy Accountant General Accountant General’s office Ncamsile Masango Accounts Officer Accountant General’s office Mumcy Khumalo Typist Phalala Fund Emily Nzimande Typist Phalala Fund Lindiwe Motsa Accounts Officer Phalala Fund Bongiwe Vilakati Accounts Officer Phalala Fund Mcebo Dlamini Accounts Officer Phalala Fund Africa Hadebe Acting Auditor General Auditor General’s office Fanisile Mabila Acting Deputy Auditor General Auditor General’s office Andreas Dlamini Deputy Auditor General Auditor General’s office Themba Kunene Principal Auditor Auditor General’s office Phestecia Nxumalo Principal Auditor Auditor General’s office S M Kunene Principal Secretary Defence Dr Austin Ezeogu Senior Medical Officer Mbabane Government Hospital, Phalala Fund & Medical Referrals Scheme Sifundza Wandile Executive member Swaziland National Association of Teachers (SNAT) Bheki Mamba Executive member SNAT Thabsile Dlamini Executive member SNAT Quintin Dlamini Executive member Swaziland National Association of Civil Servants (SNACS) 40 Statistical Annexes ANNEX 2: Average Cost per Claim of Top Ten Health Providers A. Phalala Fund 200000 350 Avg. Cost of Claim (SE) 300 150000 No. of Claims 250 200 100000 150 50000 100 50 0 0 A os C os D os C D lG os B os A bo l E lF C B ry l l l l ta ta ta ta ta La ita ta rA rB to pi pi pi pi pi pi p ra D D os os H H H H H H H Health Provider B. Medical Referrals Scheme for Civil Servants 40000 250 Avg. Cost of Claim (SE) 200 30000 No. of Claims 150 20000 100 10000 50 0 0 os H lC lJ E lI D os C FG lK lA D C B ta ta l ta ta ta ta rC rB rA rE pi pi pi pi pi pi os os D D D D os os H H H H H H Health Provider 41 No., Avg. Cost and Cost No., Avg. Cost and Cost ANNEX 3: of Claims (Log Scale) of Claims (Log Scale) A. Phalala Fund H P H 1 10 100 1000 10000 100000 1000000 P 1 10 100 1000 10000 100000 1000000 10000000 H rA P H rA H rB P H rB H Pr P P E H rC H rA P P A H rD H rB P P B r H E H rC P P C H rF r P H DD H rG P B. Medical Referral Scheme for Civil Servants P H rK r Cost H H 42 Cost H Pr P P D r H rI P H EE P H rJ H rF P P F r r H K H GG P P H rL rH P H H H rM P P Avg. Cost Avg. Cost H rI H rN P I P r H JJ H rO H Pr P Top 20 Health Practitioners Top 20 Health Practitioners P H H rP r P No. No. H KK H rQ H rP P P LL r H rR P H MM P r H S rN P N rT Total Number of Claims and Total Cost of Claims of Top Twenty Health Practitioners ANNEX 4: Log Scale of Hospitals’ Claims Profile A. Phalala Fund Cost Avg. Cost No. No., Avg. Cost and Cost 10000000 of Claims (Log Scale) 1000000 100000 10000 1000 100 10 1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 Hospitals (n=22) B. Medical Referrals Scheme for Civil Servants Cost Avg. Cost No. 10000000 No., Avg. Cost and Cost 1000000 of Claims (Log Scale) 100000 10000 1000 100 10 1 Hospitals (n = 45) 43 ANNEX 5: Claims Profile of Health Practitioners A. Phalala Fund Cost Avg.Cost No. 10000000 No., Avg. Cost and Cost of Claims (Log Scale) 1000000 100000 10000 1000 100 10 1 Health Practitioner (n = 142) B. Medical Referrals Scheme for Civil Servants Cost Avg. Cost No. 1000000 No., Avg. Cost and Cost of Claims (Log Scale) 100000 10000 1000 100 10 1 Health Practitioner (n = 304) 44 ANNEX 6: Number of Referred Patients and the Cost of their Claims A. Phalala Fund Cost Avg. Cost No. 10000000 No., Avg. Cost and Costs of Claims (Log Scale) 1000000 100000 10000 1000 100 10 1 Patients including unknowns (n = 759) B. Medical Referrals Scheme for Civil Servants Cost Avg. Cost No. 1000000 No., Avg. Cost and Costs of Claims (Log Scale) 100000 10000 1000 100 10 1 Civil Servants including unknowns (n = 460) 45 ANNEX 7: Amount, Number, and Average Cost Per Claim of Hospitals, by Individual Hospital A. Phalala Fund (n = 22 Hospitals)33 Number of Average Cost per Claim Hospital Amount (SE) Claims (SE) 1 8,058,224 308 26,163 2 4,421,767 55 80,396 3 2,297,336 36 63,815 4 1,803,522 50 36,070 5 1,385,480 73 18,979 6 1,114,698 35 31,849 7 1,043,826 6 173,971 8 626,139 4 156,535 9 413,741 1 413,741 10 304,536 5 60,907 11 124,417 1 124,417 12 89,727 17 5,278 13 84,920 1 84,920 14 83,745 3 27,915 15 60,888 3 20,296 16 56,910 1 56,910 17 28,917 2 14,459 18 26,216 1 26,216 19 26,093 3 8,698 20 14,518 1 14,518 21 1,472 1 1,472 22 1,181 5 236 Total 22,068,272 612 36,059 Source: This study. 46 B. Medical Referrals Scheme (n = 45 Hospitals) Number of Average Cost per Claim Hospital45 Amount (SE) Claims (SE) 3 3,298,799 135 24,265 5 1,091,847 31 35,221 10 510,149 22 23,189 468,154 59 7,935 9 264,866 28 9,460 1 246,504 52 4,740 230,646 3 76,882 8 212,388 9 23,599 208,603 9 23,178 2 191,024 7 27,289 4 155,410 8 19,426 154,456 11 14,041 109,735 25 4,389 98,526 11 8,957 95,345 9 10,594 85,432 29 2,946 83,577 1 83,577 81,347 7 11,621 6 69,590 5 13,918 65,734 1 65,734 50,613 3 16,871 47,630 8 5,954 43,444 1 43,444 40,477 1 40,477 38,552 3 12,851 34,179 3 11,393 33,030 6 5,505 31,365 53 592 31,110 2 15,555 28,611 2 14,306 7 25,915 2 12,958 14 24,080 4 6,020 45 Numbers in the Hospital column refer to the same hospital overlapping with the provision of services to Phalala Fund patients. 47 19,110 3 6,370 15,358 2 7,679 11,949 1 11,949 10,346 1 10,346 16 7,005 1 7,005 11 6,312 1 6,312 3,066 1 3,066 2,866 1 2,866 2,796 2 1,398 2,530 1 2,530 2,446 1 2,446 1,547 1 1,547 1,056 1 1,056 8,237,524 567 14,528 Source: This study. 48 ANNEX 8: Map of Swaziland 49 50