Our response to Peter Smith's paper is framed by our understanding of the concept of universal health coverage (UHC). While supporting the 2010 World Health Report's definition of UHC, i.e. access for all to needed, effective services while providing financial protection, we would go further to say that everyone within a country should be entitled to effective access and financial protection on the same terms (e.g. comparable access to comparable services according to need and not socio-economic status). We recognise that in all countries, there will be some differentials due to the rich buying more, or what they perceive to be ‘better’ health care, but these differentials should be at the margin. Smith's recommended approach would lead to these differentials being considerable rather than marginal.

We differ from Smith not only in terms of the values or principles that we believe should underlie the interpretation of the UHC concept and what this implies for health systems; our country experience perspective is that of low- and middle-income countries (LMICs) while his perspective is clearly that of high-income countries (HICs). He is prescribing an approach for LMICs, as these are (with very few exceptions such as the United States) the countries that are trying to move to UHC, but is doing this from a HIC perspective. Admittedly, HICs face considerable challenges in sustaining their universal health systems; in a sense, ‘the horse has already bolted’ because of the way in which these systems have developed, with a curative, high-tech and often a specialist and hospi-centric focus. But we believe there is still time to ‘close the stable door’ in LMICs and to build our universal health systems on a different foundation and to manage population expectations better. Instead of exploring options from the perspective of HICs, which we believe have allowed the development of unnecessarily high-cost systems, we should instead look at the examples of LMICs that have achieved ‘good health at low cost’ (Halstead et al., Reference Halstead, Walsh and Warren1985; Balabanova et al., Reference Balabanova, McKee and Mills2011). Certainly the health systems of HICs are held up (often by private for-profit organisations in LMICs) as the only ones that are ‘world class’ and by implication, the ones that LMICs should aspire to. These perceptions need to be challenged and the examples of quality, effective, low-cost health systems held up as alternative aspirational models.

Smith's paper concludes rather forcefully that “if policymakers wish to progress towards universal health coverage it is imperative that they implement user charges” (our emphasis), although he does add that these charges should be ‘appropriately’ implemented. It appears that he agrees with the position that he mentions at the start of his paper that ‘imposition of user charges is an inevitable consequence of seeking to match demand and supply’. Effectively, he is assuming that user charges are the only way in which to balance demand and supply, as he does not present a well-argued case that alternative approaches to achieving this balance are ineffective.

Our entire understanding of how UHC can be achieved in LMICs in a sustainable way is founded on using a range of mechanisms to balance demand and supply that focus on the service coverage, as opposed to cost coverage, dimension of UHC. What are these mechanisms for promoting sustainable progress to UHC in LMICs and why do we believe they can be effective?

A particularly positive approach that is being adopted by a range of LMICs is to build their UHC systems ‘from the ground up’, i.e. giving particularly strong emphasis to primary health care (PHC) services. Many more people can be reached by PHC than hospital-based services, and improved access can be achieved more equitably given the many benefit incidence studies that identify PHC services as generally being pro-poor, and certainly far less pro-rich than hospital-based services (O'Donnell et al., Reference O'Donnell, van Doorslaer, Rannan-Eliya, Somanathan, Adhikari, Harbianto, Garg, Hanvoravongchai, Huq, Karan, Leung, Ng, Pande, Tin, Tisayaticom, Trisnantoro, Zhang and Zhao2007; Mills et al., Reference Mills, Ataguba, Akazili, Borghi, Garshong, Makawia, Mtei, Harris, Macha, Meheus and McIntyre2012). There are numerous examples of countries that have successfully built a strong PHC platform that includes services delivered within communities. In Cuba, community-based polyclinics and a network of neighbourhood-based family physicians and nurses form the cornerstone of the primary health care system, which is able to address most of the health care needs of the population (Dresang et al., Reference Dresang, Brebrick, Murray, Shallue and Sullivan-Vedder2005). Similarly, Brazil in 1988 introduced the Family Health Program (PSF), with an interdisciplinary health team, often including community health workers, responsible for a geographical catchment area (Macinko et al., Reference Macinko, Almeida, dos Santon Oliveira and Klingelhoefer de Sá2004). Although challenges in terms of access remain, the Brazilian Sistema Único de Saúde (Unified Health System, SUS) provides universal, free health care. Another often cited example is the Ethiopian community health extension worker programme, where community workers provide a wide range of preventive, promotive and basic curative services (Haines et al., Reference Haines, Sanders, Lehmann, Rowe, Lawn, Jan, Walker and Bhutta2007). Access to basic health care services and a strong emphasis on preventive and promotive interventions are the backbone of these health systems and have achieved (and in the case of Cuba, sustained) significant improvements in health outcomes.

The importance of preventive and promotive interventions cannot be underestimated. It is often argued that a key reason for spiralling expenditure in high-income countries is the ageing population and high burden of non-communicable diseases (NCDs). While not a straightforward task, it is feasible for LMICs to reduce the potential future burden of NCDs (e.g. obesity, diabetes, hypertension) through maintaining a strong focus on preventive and promotive interventions.

We are not suggesting that specialist and hospital-based services should be ignored. However, particularly within low-income countries, expanding the range and quantity of these services will take far longer than at the PHC level, but should nevertheless be pursued over time in line with available resources. Pressure should be maintained for each government to meet the commitments it has made through signing the International Covenant on Economic, Social and Cultural Rights (ICESCR) to “take steps, … to the maximum of its available resources, with a view to achieving progressively the full realization of the rights recognized in the present Covenant” (Article 2.1), which include the right to health.

As services expand, starting from a solid foundation of high quality, accessible PHC services, as well as first level referral hospitals, and progressing more slowly at higher levels of care, we believe the most critical issue for ensuring affordability and sustainability is to entrench mechanisms for active or strategic purchasing of health services. There has been far too much emphasis in the UHC debates on the revenue collection function of health care financing systems, some consideration of the pooling function, but undoubtedly insufficient consideration of the purchasing function.

Active purchasing requires an institution that employs staff with specific skills and good information systems to take on the role of purchaser. While space does not permit full discussion of the alternative forms a purchaser (or multi-purchaser system) could take, we would motivate for a single purchaser or payer, which has decision-making autonomy, strong governance structures and effective mechanisms for holding it to account. The major rationale for this preference is the economies of scale that a single purchaser can achieve and the purchasing power that it can exert. For us, the importance of active purchasing lies in the strategies used by the purchaser to influence the behaviour of both providers and users. The approach suggested by Smith seeks only to influence the behaviour (and demands) of the population and does not require active purchasing at all. Indeed, we would argue that resorting to implementing user charges is an acknowledgement of failure to exercise strategic or active purchasing adequately.

What are the ways of influencing user behaviour other than introducing user charges? Probably the most important mechanism is to insist on PHC providers always being the first point of contact, except in critical emergencies, and strictly adhering to the referral route. This is a key reason why it is important to invest in the PHC level first; if the population receives quality services at PHC facilities, they will not feel the need to seek care directly from specialists or hospitals. Many of the most common conditions can be effectively diagnosed and treated by PHC providers, and at far lower cost than by specialists and hospitals. Given that specialists are a very scarce resource within LMICs, PHC gatekeeping and appropriate referral ensures that these scarce skills are devoted to treating those who are in greatest need of specialist care. Also, there is more job satisfaction for specialists when treating patients who really need their expertise than those with uncomplicated, common conditions (as happens in many HICs).

What are the mechanisms that a purchaser can use to influence the behaviour of providers? Clearly, a key mechanism is provider payment methods. There is growing experience of which provider payment methods best promote efficient provision of quality services. A single purchaser system can be effective in exerting considerable influence over the type of provider payment system applied, payment rates and annual increases in these rates. Increasingly, global payment caps are being used to promote affordability and sustainability.

Another of the core mechanisms for influencing provider behaviour is developing and ensuring adherence to an essential drug and devices list or formulary, accompanied by standard treatment guidelines. While this should be evidence-based, it does not require undertaking health technology assessment of the “contents of the entire ‘health basket’”, which Smith accepts is ‘infeasible’. In practice, there is already considerable consensus on what should or shouldn't be included in a drug and devices formulary for the most common conditions in LMICs. A key challenge will be to find strategies for addressing the intense public pressure that will inevitably grow over time to adopt the latest and most expensive technologies, even where there is limited or no additional effectiveness (often spurred on by aggressive marketing by pharmaceutical and other medical technology corporations and the doctors whose complicity they leverage).

Formularies are a form of service rationing (e.g. it dictates that cardio-thoracic surgeons may use ‘x’ type of stent but not a far more expensive ‘y’ type of stent with limited, if any, proven greater clinical efficacy). Smith regards user charges as ‘the most obvious rationing device to put in place in the face of budgetary constraints’, possibly because he notes that rationing of services ‘have attracted widespread hostility’. Undoubtedly rationing is controversial, but generally it is more so where it is implicit (e.g. simply allowing waiting lists to develop) and where there is a lack of transparency about how rationing decisions are made. While we do not have an evidence-base to substantiate our view, we believe that rationing by means other than user charges is feasible if the need for rationing is made explicit, there is transparency and mechanisms to ensure procedural justice in the rationing decision-making process and a strong sense of social solidarity is built around the UHC system. This is especially achievable where there is a progressive constitution, which guarantees the right to health, but explicitly gives priority to the rights of the population or society over individual rights (see Box 1). This must be supported by intensive public engagement and education about the values underlying the UHC system, particularly in relation to the implications of the principle of social solidarity, why there is a need for rationing and why the needs of the population must take precedence over the needs of individuals in certain rationing decisions.

In conclusion, LMICs have the opportunity to move towards UHC systems in different ways to the path taken in HICs, and to avoid a situation of spiralling health care expenditure. Key strategies in this regard include: building a solid foundation of quality PHC services that are not restricted to curative care but include preventive, promotive and rehabilitative services; prioritising active purchasing of services; being explicit and transparent about rationing; and building support for the principle of social solidarity. There will also need to be sustained efforts to manage public expectations and to ensure that preventive and promotive services are not neglected to pre-empt the potential tsunami of NCDs. User charges are not the only way to address the challenge of balancing supply and demand for health services.