Privatisation in health systems in developing countries: what's in a name?

This is the second issue of RHM on privatisation in health systems. The first, in November 2010, posed the question: “Whose responsibility is the public health in a privatised health system?” Having lived through round one of attempts by the UK government to privatise the National Health Service in England, which the great majority of the country are opposing, I'm now ready to answer this question. The answer is that by removing responsibility for health from national governments, no one is left in charge. There is no other body with the necessary authority and resources to ensure that the right to health is provided for. And if a government reneges on that responsibility, for whatever reason, the people who cannot afford to buy health care will invariably suffer.

One of the strengths of this journal as an international journal is that by focusing on a specific theme, the depth and breadth of the issues as they are manifested in different countries is revealed. With a complex and multifaceted subject such as privatisation, it has been more than worth it to have devoted two journal issues to it. And even then we are still only beginning to get the big picture, since Latin America and the Caribbean, the Middle East and North Africa are greatly under-represented in both journal issues on this topic.

However, as regards Asia and sub-Saharan Africa, some firm conclusions can be drawn from the papers here. Given the rapid move to privatisation in so many countries, and the amount of money that is being channelled by donors into this sector in support of privatisation (non-profit or otherwise), it has been shocking to discover just how little hard evidence there is on the type and quality of health services and standards of care that exist in the private sector in the countries of these two regions. It could almost be said that the private sector is in hiding, as there is so little information about it, at least when it comes to sexual and reproductive health services, or that at best what is published about it has been written by people working in the sector who are blowing their own trumpet, claiming success but with a clear conflict of interest.

The private sector as it exists in the developed world is something completely different from what it is in the low-income countries in these two regions. In fact, the private sector appears to consist almost entirely of small privately owned clinics and nursing homes, often with single proprietors, some of which are being pulled into social franchises, but more of which appear to operate in isolation at a local level without any oversight. They also include social marketing services and faith-based and NGO-run services which are all thought to be doing a good job but this is also not always based on solid evidence either. Sexual and reproductive health care are commonly offered in all these types of private services, or at least antenatal care and sometimes delivery and post-natal care, family planning (at least a few methods), condoms and abortions (safe and unsafe).

However, if the submissions to RHM describing these services are any measure, something urgent needs to be done to improve the evidence base about the private sector across the board.

In selecting papers for this journal issue, I have had to reject a string of papers that were descriptive of the aims and intentions of such services and of some of the so-called public–private partnerships which they are involved in or are promoting. Why? Because the papers read like promotional material. Evidence with any public health value of whether their services were of good quality or were improving health outcomes was absent, making them unpublishable. The absence of these papers in this journal issue is a great loss, and I hope the authors will be able to do what is needed to make their papers publishable, at which point I would welcome them back. Meanwhile, however, the question of how well the private sector is serving its “clientele” remains unanswered and that is a serious problem, which governments, donors, independent researchers, and health advocates need to confront and address.

This is not just my bias. A 2009 paper in the Cochrane Database of Systematic Reviews aimed to examine evidence as to whether social franchising has developed as a possible means of improving provision of health services through engaging the non-state sector in low- and middle-income countries. After a thorough search of the literature (which included randomised controlled trials, non-randomised controlled trials, controlled before and after studies and interrupted time series comparing social franchising models with other models of health service delivery, other social franchising models or absence of health services) two independent reviewers found no studies that were eligible for inclusion in a review. They point to:

“…the need to develop rigorous studies to evaluate the effects of social franchising on access to and quality of health services in low- and middle-income countries. Such studies should be informed by the wider literature to identify models of social franchising that have a sound theoretical basis and empirical research addressing their reach, acceptability, feasibility, maintenance and measurability.” ¹

Similarly, if in order to survive a social marketing organisation has to be concerned with the market position of its own brands of condoms and contraceptives, and with boosting the visibility and therefore sales of its products (as Purdy acknowledges in his paper here, which I actually wanted to publish because it describes how to use social media to educate people about condoms on the Internet) what makes it different to a private, profit-making company that does the same? Just that its prices are lower? Apparently, among all the data on numbers of condoms sold by social marketing organisations, it is not even known whether they are increasing condom use rates, because they mostly don't know who is buying them, let alone whether they are being used consistently and correctly, or with whom. Thus, Knerr comments in her analysis of condom social marketing organisations that: “rigorous and reliable evidence of impact on condom usage and disease prevention are limited, as is evidence of impact on equity of access to condoms for poor populations, women and people living with HIV”.

There is other evidence, however, that equity in access to condoms is a dream when it comes to the poor in poorer countries, including where the HIV epidemic is at its highest. Men in rural northern Kenya, for example, are having to wash and re-use subsidised condoms because they are both too scarce and too expensive to throw away after one use. A March 2011 report said:

“Local TV channels recently showed images of men in Isiolo, in rural northern Kenya, washing condoms and hanging them out to dry; the men said the price of condoms meant they could not afford to use them just once. Other men in the village said when they had no access to condoms, they used polythene bags and even cloth rags when having sex… Condoms are free at government health centres, but in rural Kenya these are few and far between and supplies unreliable. ‘Many rural areas in the country are inaccessible due to the poor road network and this makes distribution of condoms difficult and challenging,’ said Peter Cherutich, head of prevention of the National AIDS and Sexually transmitted infections Control Programme, NASCOP. ‘Because government condoms are mostly made available at health facilities and there are not many in rural areas, this creates another challenge in distribution.’… ‘M usually walks the 5km to his nearest health centre for condoms, but on occasion finds them out of stock.’” ²

So who is buying socially marketed condoms? The same people who would have bought them in a pharmacy at a higher price, because in either case they could afford them?

Effect of MDG5 on attention to sexual and reproductive health

Because the countdown to 2015 for achieving the Millennium Development Goals (MDGs) has begun in earnest,^* and countries are scrambling to make up for the failure of decades to reduce maternal mortality sufficiently, RHM is getting a growing number of submissions on maternal (and infant/child) health, maternity services, and maternal mortality and morbidity. This is obviously a good thing, because this attention has been lacking for a long time and high maternal mortality is a stain on the good name of a very long list of countries.

At the same time, however, in an effort to achieve the maternal and infant part of MDG5, countries are limiting their efforts and not addressing all the parameters of women's sexual and reproductive health (forget men, they are off the radar entirely except for HIV, male circumcision and socially marketed condoms). Three papers (by Desai et al, Rashid et al and Pantha), and a book by Dr Rani Bang reviewed in this journal issue, are a powerful reminder of the fact that a wide range of gynaecological morbidity – prolapse, fistula, sexually transmitted and other reproductive tract infections, urinary tract infections, urinary incontinence, a long list of menstrual problems, lower abdominal pain, low sexual desire, discomfort and pain during intercourse, miscarriage and stillbirth, and complications from unsafe abortion – are not being treated adequately or at all in low-income countries, or only in the so-called informal sector and by pharmacies and drug sellers of various kinds, not necessarily effectively.

Using quantitative data to assess access, equity and service utilisation

Four really important papers have moved this journal into new territory. One by Carla AbouZahr is about how to interpret two new global estimates of maternal mortality for 2008, which were published in 2010, six months apart, and have been causing confusion ever since. Traditionally, the World Health Organization has been responsible for these estimates, which they have been publishing every three years since 1990. In 2010, the Institute for Health Metrics and Evaluation in Seattle, USA, decided that the private sector should have a go and published their own estimates for 2008, without talking to WHO and in a great rush, which was interpreted by many as an attempt to grab the limelight at an impending G8 meeting mid-year, where MDG5 was high on the agenda. AbouZahr's paper examines the differences between the two sets of estimates, and discusses the value and inherent limitations in the modelling exercises needed to reach these estimates, and asks whether modelling exercises are the most relevant to the needs of country-level decision-makers. Existing maternal mortality ratios are based largely on estimates and statistical modelling because of the absence of comprehensive civil registration data on deaths and causes of deaths in most low-income countries. Yet as she points out, “the record of deaths among women of reproductive age derived from civil registration is often the first step in conducting a confidential enquiry into and preventing maternal deaths”. She calls for all countries to start to collect that data in order to be able to use it to improve outcomes.

The other three quantitative papers examine data from nationally representative household surveys on reproductive and sexual health care as a way of measuring equity of access to, utilization of and financing of services. Twenty-five countries are covered by one of these three papers (Limwattananon et al), based on the two most recent Demographic and Health Surveys for each country, and six countries by the second paper (Nguyen et al), which is based on both Demographic and Health Survey data and National Health Accounts and Reproductive Health Sub-Accounts for those countries. The third quantitative paper examines the situation in Thailand in depth (Kongsri et al), based on their two most recent national Reproductive Health Surveys. Being a qualitative sort of person, my first reaction to analyses of survey data such as these was how much quantitative data alone are able to say about countries. But these authors have all written their papers so that even someone as statistically illiterate as I am can understand them, and now that they are done, I can say with no hesitation whatsoever that all three are dynamite.

Nguyen et al, for example, show that in all six countries they studied, reproductive health service utilization levels are highly inadequate, that the continuing large role played by the informal sector raises well-known concerns about safety, efficacy and quality of care, and that there is a clear link between spending and utilisation of services – i.e. higher national expenditure on reproductive health is associated with better utilisation rates, and conversely, under-financing is a key constraint for achieving universal access to reproductive health. They also raise concerns about the extent of donor financing in the absence of corresponding domestic investments, for sustainability.

This does not make qualitative work any less valuable as a means of giving greater depth to and confirming quantitative findings. As Collumbien et al say in their paper on youth-friendly services for young people in two rural districts in India, studies may or may not even ask the right questions:

“…Social norms influence the design of research and how findings are interpreted and inform the design of interventions. In our study, for example, questions on seeking abortion services were left out of the questionnaire for unmarried girls, in order to avoid offending the communities where the interventions were being planned. Although it is widely acknowledged that community-based surveys do not produce reliable estimates on abortion in any case, by not even asking the question in an open and non-judgemental way the research itself reinforces these norms. Hence, we become complicit in perpetuating the denial of reality by not producing the evidence. Appropriate interventions are then not identified and young people suffer.”

As the strategic assessment of unsafe abortion in Malawi (Jackson et al) shows so well, qualitative data, i.e. interviews on the ground, observation of services, measurement of quality of care, as well as measurement of financing and cost, trends in utilisation rates, as well as data on health outcomes and epidemiological data describing existing health problems – all reveal the extent of equity of access for everyone who needs care. Papers such as those of Rashid et al, Honda et al, Chelstowska, Jonkers et al, Rispel et al, and Ngo & Hill in this issue add immeasurably to what survey data would say, because they bring the experience of patients and health professionals to the fore, and show the consequences of failings in the health system for the individual patient, who is the raison d'être of public health services.

One of the most interesting linguistic issues in these papers, speaking of surveys, is the description of the top and bottom quintiles of the population in the paper by Limwattananon et al. Although they talk about the rich–poor gap in the paper, they also describe these two quintiles as being “the poorest” and “the least poor” (not the richest) in the country. This is lest we forget that being rich in low-income countries is for the most part still being poor, just not quite so poor as others.

A health systems approach to sexual and reproductive health care

The papers about individual countries published here are about Bangladesh (Rashid et al), India (Desai et al, Bhate-Deosthali et al, Madagascar (Honda et al), Malawi (Jackson et al), Nepal (Pantha et al), Thailand (Kongsri et al) and Viet Nam (Ngo & Hill). Papers from Poland about the privatisation of abortion turning sin into gold (Chelstowska) and the Netherlands about the disadvantages faced by ethnic minority women with complications of pregnancy and delivery (Jonkers et al) and a paper about the growing cross-border travel from developed to developing countries for assisted reproduction services in Asia and the consequences for health systems of catering for rich patients from abroad (Whittaker), add to the richness of the issues raised in this journal issue.

Even when these papers are not primarily about privatisation – such as Desai et al's about the rising prevalence of hysterectomy in young women workers in Gujerat, India since health insurance began to partially cover hospitalisation – they allude to it, recognise it and take account of it. This is a major step forward from RHM's point of view, because privatisation has long been hovering in the background in health systems, often in hiding, and most of us in the sexual and reproductive health field have failed in the past to take enough account of it, if we have done so at all.

These papers address commercialisation of services, serious lack of affordability of services due to out-of-pocket payments (not only for the poorest but also for the lower-middle class in low-income countries), equity gaps due to socioeconomic class, poor standards of care in small private hospitals, and above all, the status and accessibility of health services (public, private, informal) and which sub-groups (almost always of women) are utilising which of them (by socioeconomic class, geographic location, age, education and marital status).

Here is some of what I've learned from these papers:

•	Not enough is being invested by countries in public health services or in the education and training of new generations of health professionals.*
•	Those who are better-off economically, better educated, married, and living in urban areas have far greater access to formal sexual and reproductive health services from both the public and private sectors than those who are poorer, less educated, single and living in rural areas – unless there is universal access to health care, a claim which very, very few countries are able to make.
•	The private sector does not appear to be making up for what the public sector is not providing and cannot afford to provide. This is because, first and foremost, no one is in charge of what that sector does or does not do.
•	Services in the private sector cost more, which limits who can use them. However, the poor are often using them because the public sector is known or believed to offer poorer quality care and less care, e.g. fewer medicines and drugs, fewer forms of treatment, less up-to-date technology.
•	While the private sector costs more than the public sector, the public sector has also become too expensive for many poor people to use, especially when receiving care depends on payment of out-of-pocket expenses, giving “gifts” and travelling long distances to obtain it.
•	Ill-health is a leading cause of indebtedness among the poor. The private sector is not alleviating this well-known problem but probably exacerbating it.
•	There is no effective regulation or monitoring by governments of private sector health services. The stewardship function that neoliberal proponents of the private sector have assigned to governments is not working for precisely the same reasons that public sector health services are failing – lack of capacity and management skills, lack of regulations and the power to enforce them, and lack of sufficient resources.
•	Where services are scarce, higher fees can be charged and under-the-table payments are expected, and corruption can become the accepted basis of access to care and treatment.
•	People from developed countries are increasingly travelling to developing countries to get certain services, especially for infertility treatment, because of their high cost in the private sector in their own countries, as well as due to policy restrictions.
•	And most importantly, the informal sector in developing countries is providing the great bulk of “health care” to the poor, to those living in rural areas, and to young single people, all of whom are being neglected by both the public and private formal sectors.

Who is the private sector in middle- and low-income countries?

Based on the papers in this and the previous issue of RHM, the private sector in low- and middle-income countries does not mostly consist of large, shiny new corporate premises, staffed with highly skilled professionals. Many, perhaps most, are single-proprietor small hospitals and nursing homes, mostly with less than 30 beds, with extremely varying quality of staff and premises, delivering uncertain health outcomes. If the ones in Maharashtra, as described in the paper by Bhate-Deosthali et al, are typical, they often have very limited facilities, few qualified staff, poor standards of care and even worse record-keeping. The cover photo on this journal issue (from a photo agency) is shocking, but it was described by Bhate-Deosthali as “typical” when I asked her what she thought of it while I was writing this editorial. Indeed, some though not all of the photos she sent me of the small hospitals she and her co-authors had visited in Maharashtra were just as bad. Of the 261 small hospitals she and her co-authors studied in Maharashtra, 146 were providing “maternity services” and many of the owners were obstetrician/gynaecologists, yet 137 did not have a qualified midwife, and though most claimed they provided emergency care, including caesarean section, only three had a blood bank and eight had an ambulance. And these are the sorts of facilities that are soon to be pulled into a public–private partnership with the state government. Will the state government really be able to reduce maternal deaths without spending a lot to upgrade and staff them properly, just as they would have to do with public sector clinics?

However, the most important discovery to arise from the papers in this journal issue for me is the fact that what is called the private sector by academic and NGO researchers includes what they euphemistically label as the “private informal sector”, that is, the one that consists of traditional birth attendants who cannot (on their own) reduce maternal deaths, various kinds of healers, drug sellers who are not trained pharmacists, and village doctors who are not doctors. If data about this untrained, often uneducated sector in other studies, are subsumed in data on the private sector in other studies, which was initially the case with some of the papers published here – until I asked the authors to disaggregate them – this is the most elaborate form of hiding of all. Thus, for example, women giving birth at home were initially counted as having had an informal private sector delivery. If that is being interpreted to claim that the private sector in low- and middle-income countries is feasible and working, even in rural areas and for the poorer populations, then something is very wrong. Once the data were disaggregated, it became crystal clear that there is no such thing as “the private sector” outside urban areas in low- and middle-income countries, but rather that most deliveries and many other forms of sexual and reproductive health care, including contraception for young single people and drugs and other means to treat sexually transmitted diseases and other reproductive tract infections, as well as unsafe abortions and treatments for a range of chronic gynaecological morbidity – are still being obtained from unskilled, untrained providers by the majority of poor and rural women.

As Limwattananon et al conclude in their paper on inequities in where women delivered their babies in 25 low-income countries between 1995 and 2006:

“Fragile health systems that result from inadequate public investment in health system infrastructure in rural areas, and lack of skilled and trained health professionals, are among the major impeding factors in reducing maternal mortality. Honouring commitments by governments and international development partners is an important entry point for health systems strengthening. Strengthening private, for-profit providers is not a policy choice for poor, rural communities.”

I was criticised for putting my anti-privatisation bias in the call for papers for the two RHM journal issues on this theme. I have seen nothing in the papers I accepted for publication in these two issues, nor in the papers that were submitted but not accepted, nor in the papers summarised in the Round Up section on public health systems, financing and privatisation, to make me change my mind on this subject.

Moreover, as Kongsri et al show for Thailand, where universal coverage has been national policy before the MDGs were even dreamed of:

“…High coverage and low inequity in access to services are the result of extensive investment in the health system by successive governments, in particular in primary health care at district and sub-district levels.”

I conclude that only universal access to good quality health services and attention to the social determinants of health will deliver health for all in any and every aspect of health, and that politically committed governments are the only ones who can and have delivered it.

Notes

* Ironically, the more comprehensive Programme of Action of the International Conference on Population and Development 1994 was also meant to reach fruition in 2015, and certainly will not, yet this is getting almost no press. The MDG5 indicator “universal access to reproductive health” is spelled out in the ICPD programme, but is rarely remembered as the source of what is supposed to be achieved.

* For example, in the 10 years since heads of state of African Union countries met in Abuja, Nigeria, and pledged to set a target of allocating “at least 15%” of their annual budget to improve the health sector, only Rwanda and South Africa have reached that objective, according to WHO.³