Public-private partnerships for health require thoughtful evaluation.

There are now about 50 public–private partnerships operating internationally to provide the means of combating diseases associated with poverty. Their aim is to develop, or improve access to, health products such as drugs, vaccines, contraceptives, microbicides, diagnostics and bednets. Since getting under way during the last few years, they have been variously criticized but usually with no distinctionmade between their different ways of working (1–3). The designation ‘‘public–private partnership’’ is claimed by a wide variety of arrangements. They range from small singleproduct collaborations with industry to large entities hosted inUN agencies or private not-for-profit organizations. They also include legally independent ‘‘public interest’’ (but actually private sector) entities such as the Global Fund to Fight AIDS, Tuberculosis andMalaria. Although the private sector includes not-for-profit or civil society entities as well as commercial ones, most commentators define ‘‘PPP’’ as involving some for-profit organizations. These can be companies whose core business expertise is outside the health sector, such as communications, or inside it, such as a pharmaceutical research, raising different questions. The word ‘‘partnership’’ has been used loosely to include communication, consultation, coordination, and collaboration. ‘‘Public–private partnerships’’ has also been used to refer to the privatization of health services delivery, although here it is governments alone which set the rules, and this causes further confusion. In fact, a strict definition of ‘‘PPP’’ would probably require a significant degree of joint decision-making. For most commentators ‘‘PPPs’’ imply innovative interaction, and exclude traditional transactions such as grant-making or procurement of goods and services. But ideas of ‘‘innovation’’ differ too. Simply calling an interaction a ‘‘partnership’’ does not ensure that it actually involves joint decisionmaking, or innovation. Reality needs to be distinguished from perceptions by means of impartial observation and documentation. To evaluate criticism it is important to understand the actual policy-making and implementation processes in different so-called partnerships. Arrangements orchestrated from within UN agencies (such as the Global Alliance for Vaccines and Immunization (GAVI), Roll Back Malaria (RBM), STOP TB, and others), operate by their host’s policies and procedures. The Boards and working groups of such PPPs act as advisory bodies to the heads of their host agencies and their other partners. Critics have complained of inappropriate influence by business. But, for example, to include one person representing the pharmaceutical industry in the 16-member Board of GAVI adds necessary expertise while being highly unlikely to overturn the entire policy-making systems ofWHO,UNICEF, theWorld Bank and the other members, especially as the host of the GAVI secretariat (UNICEF) retains a veto on its actions and each partner has the right to ignore the board’s decisions. Legally independent partnership entities such as the International AIDS Vaccine Initiative (IAVI) or Medicines for Malaria Venture (MMV) are actually private sector not-for-profit organizations, since their assets are not state-controlled. Hence, these types of PPP face the same decision-making and governance issues as the nongovernmental organizations (NGOs) which often severely criticize them. All organizations claiming to beworking in the public interest need to deal effectively with four issues: representation of intended beneficiaries, funders and other stakeholders; conflicts of interest, which include biases arising from any person’s organizational affiliation or strongly held convictions; accountability; and transparency. There is no perfect model, and decisions are best left to each partnership to determine what works best for them and their ‘‘clients’’, with the obvious caveat that they must comply with the relevant laws, ethical conventions, and international and national policy frameworks. Genuine deference to local decision-making would acknowledge that the elected governments of poorer countries are better placed than NGOs based in richer ones to decide what is in the interests of their populations. Poorer countries have to deal withmany external partners and partnerships, all of them eager to help, each in its own particular way. At present the number of partnerships is below the number of development aid agencies and well below that of international NGOs working in health. Nonetheless, ‘‘partnerships’’ themselves can reduce the transaction cost overload by working together more closely, both internationally and within countries. The policy frameworks shaping PPP (and NGO) inputs on health problems must come ultimately from national governments. Further studies of partnerships on the ground will help to optimize their integration. Partnerships, funders, governments, UN agencies, and companies could all adopt a more strategic approach to optimizing the aggregate health benefits of public–private collaboration, but the problems that are unique to those PPPs that include for-profit partners must be better defined. Critics have not differentiated these from problems shared by any targeted effort, whether public, private or joint, or from problems common to all efforts in poorer countries that are externally funded. The rationale for public–private collaboration in health work is not simply to capture money from profit-making enterprises on the one hand or facilitate the intrusion of business into the public policy setting on the other. True partnership is really about combining different skills, expertise and other resources— ideally in a framework of defined responsibilities, roles, accountability and transparency — to achieve a common goal that is unattainable by independent action. The health benefits of these social experiments must be maximized and potential risks minimized. The current crop of PPPs can in time yield a body of experience on which to construct evidencebased ‘‘best practices’’. Agreement on these may seem far-off but we can help build consensus now by rejecting ideological suspicions, and pragmatically analysing the many existing and emerging efforts. It is populations afflicted by the health problems associated with poverty which stand to gain the most from such an approach. n