NOTE NUMBER 345 94424 viewpoint PUBLIC POLICY FOR THE PRIVATE SECTOR JANUARY 2015 Primary Care for the Poor Jorge Coarasa and The Potential of Micro-Health Markets to Improve Care T r a d e a n d C o m p e t i t i v e n e s s G l o b a l P r a ct i c e Jishnu Das M uch o f the pri ma ry cura ti ve ca re pro vi d ed to the po o r by the pri va te This Viewpoint is the result of collaboration s ecto r o ccurs no t a t l a rg e ho s pi ta l s but a t s ma l l , s i ng l e- pers o n cl i ni cs . between the Trade and Whi l e s uch “ mi cro - hea l th” pro vi d ers i ncrea s e a cces s , ques ti o ns pers i s t Competitiveness Global a bo ut qua l i ty. So me ha ve a rg ued tha t the mi cro - hea l th s ecto r need s to Practice and the World Bank Research Group. be better reg u la t ed . Thi s no te ci tes recent s tud i es i n a rg ui ng tha t the Jorge Coarasa is Senior mi cro - hea l th s ecto r need s to be better u n d erst o o d . A mo re evi d ence- Economist, Jishnu Das is Lead Economist, both with ba s ed a ppro a ch ma y ena bl e the Wo rl d Ba nk G ro up to better ta rg et the World Bank Group. i nves tments a nd i nterventi o ns a nd hel p thes e pro vi d ers ful fi l l a n i mpo rta nt ro l e s ervi ng the po o r. Micro-Health Transforms Primary Care for small, single-person clinics that offer low-cost the Poor curative services. The average Indian village In many low-income countries, the private sec- has 4.4 health care providers and the cost of a tor is the predominant source of primary cura- single interaction with a private primary care tive care, which refers to the health system’s first provider in India ranges from 50 cents to $2 interaction with an ill person. Demographic (Das, Muralidharan, et al. 2014). That would and Health Surveys ask household members be anywhere from 10 percent to 40 percent of where they sought care when a child had a fever the minimum daily wage of an unskilled urban or diarrhea. Between 1990 and 2013 (across 224 worker, or between 11 percent and 80 percent THE WORLD BANK GROUP surveys in 77 countries) half the population of the daily wage of a rural unskilled worker.1 At sought care in the private sector, and between the high end, these costs are similar to those in 1998 and 2013, even among the poorest 40 per- the United States as a percent of daily wages.2 cent of the population, two-fifths sought care in These numbers, suggesting both ready avail- the private sector (Grepin 2014). For adult and ability and relative affordability of private care, childhood illnesses combined, private sector call into question the longstanding assumption use in the early 2000s (the latest data) ranged that single public providers at the village level from 25 percent in Sub-Saharan Africa to 63 are the dominant source of health care. In the percent in South Asia (Wagstaff 2013). Private emerging paradigm of micro-health markets, providers typically used by the poor tend to consumers can choose among multiple provid- be micro-health providers operating out of ers, even within small geographical spaces in P rimary C are f or the P oor T h e P o t e n t i a l o f M i c r o - H e a l t h M a r k e t s t o I m p r o v e C a r e countries as diverse as Cambodia, India, and Data from India (Figure 2) includes any pro- Kenya.3 vider to whom patients go for medical advice, In Figure 1, each circle represents registered regardless of the provider’s registration status. health care facilities in Kenya as of 2011. The blue The data come from mapping over 750 pro- circles are private facilities; the black circles are pub- viders in the health markets of 100 villages in lic. More than half the registered health facilities the state of Madhya Pradesh—one of the poor- (4,678 out of 8,776 in the country) are private, rang- est states in a country with very poor human ing from small, one- to two-person clinics to large, development outcomes (a measure combining integrated service providers. The remote arid regions health, education, and income indicators). The 2 of the country in the north and northeast of Kenya qualifications of these providers range from are primarily served by public facilities. But although those with an MBBS (bachelor of medicine, private facilities are more likely to locate in areas with bachelor of surgery—roughly equivalent to a higher population density and higher wealth, they medical doctor in the United States) to degrees are by no means found only in urban areas. Further, from alternative systems of care to no formal the Kenya data show only those facilities registered qualifications at all. Those with no qualifica- by the government and therefore almost certainly tions made up 60 percent of the providers sur- underestimate the number of private micro-health veyed; 10 percent of providers worked in the providers. According to one estimate, about a third public sector; and only 5 percent were private of health care in Kenya, as measured by “health care sector MBBS doctors. Madhya Pradesh is not interactions” or visits, is delivered by informal provid- exceptional. Across India, the average village ers (Sudhinaraset et al. 2013). has more than four primary health care provid- ers—the majority private—delivering care to Figure Formal health facilities in Kenya, by ownership households (Das, Muralidharan, et al. 2014). 1 The ready availability and affordability of private micro-health care does not, by itself, mean that micro-health is the right approach: If public care is cheaper and higher quality, the right response may be to limit micro-health markets in favor of an expanded public health sector. The first step, therefore, is to understand how these markets operate, the kind of quality they provide, and the choices that households make. Because the private sector operates in a mostly unsubsidized and lightly regulated fashion in many countries, with prices that are determined in market equilibrium, micro- health markets offer a lens into the fundamen- tal determinants of demand and supply in the primary health care sector. Understanding the Distribution of all providers in Madhya Figure Pradesh, India 2 Private, trained 26% Private, Private (4,678) untrained Public (4,098) Private 59% MBBS 5% Public 10% Source: authors’ analysis of the Kenya Master Facility List (ehealth.or.ke). economics of these markets can provide crucial The problem is twofold. First, there are insights into how poor people access and pay large differences in the knowledge of providers for services and the responses of providers to with formal medical training since the quality these choices. of this training can itself differ substantially across institutions. Second, there is a large From “access” to “access with quality” “know-do” gap in medical practice. Health An expanded understanding of the issue care providers often do a lot less than they requires moving from measuring and think- know they should, the result being that they ing about access to measuring and thinking practice below the frontier of their knowledge. about access with quality. In many places, the The quality of care in practice responds to both problem is not whether people can access pri- knowledge and effort. Although micro-health mary health care, but whether that care is of providers in the informal sector typically have sufficient quality to be of any value to their lower levels of knowledge than, for instance, health status. For example in rural India, only providers in the public sector, their effort lev- 4 in 10 people reporting chest pain to a health els are often higher, bringing them closer to care provider would be correctly diagnosed as parity in the practice quality of their formally heart attack cases; 80 percent of children suf- trained counterparts. fering from viral diarrhea would be incorrectly Consequently, the available evidence does given antibiotics; and less than 25 percent of not confirm a strong correlation between for- the poor who are suffering from hypertension mal training and health care outcomes. It is even know that they have this condition (Lee important at the outset to resist simple compari- et al. 2012). Similarly, in Kenya, one of the only sons. For example, while public facilities offer low-income countries with a national survey on services delivered by licensed professionals, in patient safety, only 1 percent of health facilities many countries these trained doctors and nurses in 2012 achieved minimum compliance with often delegate health care delivery to unlicensed international patient safety standards (Kenya subordinates (Das et al. 2012).4 Recent work Ministry of Health, IFC, WHO 2012). Poor comparing providers with and without training health care quality and patient safety lead to and those in the private versus the public sector poor health outcomes and inflated costs that shows that while training increases diagnostic impose unnecessary financial burdens on fami- accuracy, the increases are relatively small and lies, with economy-wide implications in terms of in many cases, providers without formal medi- lost work and inefficient health care. Quality is cal training deliver care equivalent to what is the central pillar ensuring that health systems available from providers with formal medical are fair, equitable and financially feasible, and training and those in the public sector (Das et al. improving the quality of care is a priority for 2012). In such cases, a blanket requirement that many low- and middle-income countries today. all practitioners have formal degrees would bar many good (unlicensed) providers and maintain More evidence needed on quality of care many poor quality (licensed) providers. While The discussion about micro-health providers on its face the logic of this type of regulation almost inevitably centers on the need for greater seems sound, a closer look at the evidence raises regulation and punitive action based on the questions about whether it serves the intended assumption that inferior quality of care among purpose of improving the quality of care. In micro-health providers must be the major driver extreme cases, the evidence suggests, blanket of poor quality in the health system. Based on enforcement of such regulations could actually qualifications alone, this is a reasonable con- reduce both the quality and availability of care clusion, given the widely varying credentials of for poor populations. micro-health providers already noted. But while A second prevailing assumption in the regulation and punitive action are legitimate health care field has been that poor quality strategies for addressing low quality care, there of care stems from a lack of hardware that pre- is little direct evidence that would justify basing vents doctors and other trained professionals such punitive action solely on suppressing prac- from achieving their technical knowledge fron- titioners who lack formal training. tier. Here again, recent research has demon- strated this position to be largely incorrect. A 2. Scaling up interventions to improve qual- number of studies have found that, in primary ity requires understanding and addressing care settings, improvements in infrastructure market failures. and supply of materials including drugs may As with the design of policies and regulations to be necessary but are not sufficient to improve improve quality, market-level investments will have to quality (Das and Hammer 2014). be grounded in a robust understanding of the causes underlying market failures. This idea is nascent for Recommendations primary health care, but the approach has already been tested in recent research. For example in 4 1. Effort, rather than hardware or training, Uganda, after noticing that pharmacists often sold may count the most. counterfeit drugs at prices identical to those of real There is a need to go beyond regulation and infra- drugs, researchers worked with a non-governmental structure improvements in the quest for quality. organization (NGO) to introduce a new, high-quality One consistent finding from new research across option into the market. The NGO pharmacist sold many settings, ranging from India to Rwanda only real drugs and had a strong reputation to back to Uganda to the United States, is that large that claim. One year later, when the researchers went improvements in quality can result from chang- back to look at the other existing pharmacists, they ing the level of effort that providers exert in their found that the number of counterfeit drugs had interactions with patients. Effort, in turn, is sensi- fallen dramatically. Introduction of the high-quality tive to interventions that change the financial, option forced other pharmacists in this market to non-financial, and intrinsic incentives that pro- change their products and reduced the chances that viders face. For example, a doctor working in his a low-income household seeking medicine would private clinic in Madhya Pradesh is 23 percentage be sold a fake (Björkman-Nyqvist, Svensson, and points more likely to give the right treatment for Yanagizawa-Drott 2013). unstable angina than the same doctor working in his public clinic (Das, Holla, et al. 2014). The 3. Changing the way impacts are measured clear financial incentive for greater effort rep- will lead to smarter investments. resented by working in a fee-for-service model A new approach will be required to understand and rather than for a fixed salary is only one way to measure the impact of public and private invest- achieve better outcomes. Similar results can be ments in primary health care markets. Traditional obtained by improving community or adminis- approaches like measuring how many users attended trative accountability, or even by playing on the a private clinic or a new public health center may intrinsic motivation of health care providers. For no longer be adequate. The problem is that an instance, recent research shows that effort levels increase in the number of users of a new facility can be improved through better peer monitoring does not necessarily mean that quality improved in or increasing providers’ pride in their own work the market where that facility operates. An accurate (Brock, Lange, and Leonard 2014).5 measure of how such a shift affected quality would These observational and experimental studies require measuring the quality of the clinics that provide important “proof-of-concept.” Moving people no longer use. This was illustrated clearly from the conceptual to the scale-up stage requires in an experimental study of user fee reductions a different investment approach. It is increasingly in Ghana. The study showed higher utilization of evident that improving the quality of primary health care services following a reduction of fees, health care requires interventions at the market- but no change in health outcomes (Ansah et al. level. These are interventions that do not “pick 2009). The reduction increased demand for public winners” in the sense of mobilizing public and clinics, but the private clinics that people no longer private funding to individual clinics or hospitals. patronized were probably of equally high quality, so Instead, they improve the overall functioning of increased traffic at the public clinics did not improve markets, allowing different clinics to improve under the overall quality in the market. new policy environments according to their own capabilities and constraints. In other words, a new Conclusion approach requires moving from “fixing the pipes” Most global health funding focuses on public to “fixing the institutions that fix the pipes.” providers. Investment in the private health sec- tor is directed mostly towards large integrated 6. Sixty-three percent of IFC investments in health in providers that serve a predominantly urban, FY12–13 were in integrated service providers. relatively wealthy clientele.6 To the extent that the “other” private sector, composed of micro- References health providers, enters policy discussions, the Ansah, E., S. Narh-Bana, S. Asiamah, V. Dzordzordzi, emphasis is primarily on regulation. This note K. Biantey, K. Dickson, J. Gyapong, K. Koram, B. argues that although regulation is an important Greenwood, A. Mills, and C. Whitty. 2009. Effect policy lever, there is a need to broaden the set of of Removing Direct Payment for Health Care 5 interventions to supportive measures and institu- on Utilization and Health Outcomes in Ghana- tional responses that can influence consumer and ian Children: A Randomized Controlled Trial. provider behavior. Accomplishing this requires a PLOS Medicine, Speaking of Medicine, Commu- better understanding of the way private providers nity Blog, http://www.plosmedicine.org/article/ operate and the drivers of consumer choice in a info%3Adoi%2F10.1371%2Fjournal.pmed.1000007. market context. A comprehensive view of health Björkman-Nyqvist, M., J. Svensson, and D. Yanagizawa- systems would include better understanding all Drott. 2013. The Market for (Fake) Antimalarial types of providers, their interactions, and how Medicine: Evidence from Uganda. Harvard their behavior mediates the impact of potential University (accessed October 20, 2014), http://www interventions. This task requires a major effort, .hks.harvard.edu/fs/dyanagi/Research/Fake particularly in countries with many widely dis- Drugs.pdf. persed villages. But it can be done; indeed, it Brock, M., A. Lange, and K. Leonard. 2014. Giving and has been done. Teams working in India and Promising Gifts: Experimental Evidence on Reci- Cambodia have completed major regional and procity from the Field. European Bank for Recon- national surveys mapping out the availability struction and Development, Working Paper No. 165. and quality of public and private providers in Brock, M., A. Lange, and K. Leonard. 2013. Generos- representative village samples. The commit- ity Norms and Intrinsic Motivation in Health Care ment of resources should be measured against Provision: Evidence from the Laboratory and Field. the potential benefits for improvements in the European Bank for Reconstruction and Develop- overall quality of health care in the very facilities ment, Working Paper No. 147. that a majority of the poor use. Das, J., and J. Hammer. 2014. The Quality of Primary Care in Low-Income Countries: Facts and Econom- ics. The Quality of Primary Care in Low-Income Countries: Facts and Economics. Annual Review of Notes Economics, 6: 525–553. 1. Current Minimum Wage Rate data, Labour Depart- Das, J., A. Holla, M. Kremer, M. Aakash, and K. Muralid- ment website, National Capital Territory (NCT) Delhi, haran. 2014. Quality and Accountability in Health: India. http://www.delhi.gov.in/wps/wcm/connect/ Audit Evidence from Primary Care Providers. World doit_labour/Labour/Home/Minimum+Wages/. Bank. 2. Doctor Visit Costs, Debt.org website, http://www Das, J., A. Holla, V. Das, M. Monahan, D. Tabak, and .debt.org/medical/doctor-visit-costs/. B. Chan. 2012. The Quality of Medical Care in 3. Based on experiences of World Bank’s Human Clinics: Evidence from a Standardized Patient Study Development Unit, East Asia and Pacific Region, in in a Low-Income Setting. Health Affairs, 31(12): Cambodia’s Rural Health Markets. 2274–2784. 4. When standardized patients were sent to public Das, J., K. Muralidharan, A. Holla, and M. Kremer. primary clinics, in 64 percent of interactions they were 2014. Preliminary results from the Medical Advice, attended to by a subordinate who was not the doctor in Quality and Availability in Rural India (MAQARI) charge. These ranged from compounders, or pharma- study. cists, to janitors. Grepin, K. 2014. Trends in the Use of the Private 5. Refers to separate articles by same authors in 2014 Sector: 1990–2013, Analysis of Demographic and and 2013 as noted in References. Health Surveys. Mimeo. P rimary C are f or the P oor T h e P o t e n t i a l o f M i c r o - H e a l t h M a r k e t s t o I m p r o v e C a r e Kenya Ministry of Health, International Finance Corpo- ration (IFC), World Health Organization (WHO). 2012. Kenya National Patient Safety Survey. Lee, J., P. Arokiasamy, A. Chandra, P. Hu, J. Liu, and K. Feeney. 2012. Markers and Drivers: Cardiovascular Health of Middle-Aged and Older Indians. Aging in viewpoint Asia: Findings from New and Emerging Data Initiatives. Sudhinaraset, M., M. Ingram, H.K. Lofthouse, and is an open forum to D. Montagu. 2013. What Is the Role of Informal encourage dissemination of Healthcare Providers in Developing Countries? A public policy innovations Systematic Review. PLOS ONE. for private sector–led and Wagstaff, A. 2013. What exactly is the public-private mix market-based solutions for in health care? World Bank, Let’s Talk Development development. The views Blog, http://blogs.worldbank.org/development published are those of the talk/what-exactly-public-private-mix-health-care. authors and should not be attributed to the World Bank or any other affiliated organizations. Nor do any of the conclusions represent official policy of the World Bank or of its Executive Directors or the countries they represent. To order additional copies contact Jenny Datoo, managing editor, Room F 5P-504, The World Bank, 1818 H Street, NW, Washington, DC 20433. Telephone: 001 202 473 6649 Email: jdatoo@worldbank.org Produced by Carol Siegel Printed on recycled paper This Note is available online: http://www.worldbank.org/fpd/publicpolicyjournal