Interview with James Macinko

James Macinko is a Professor and researcher in public health policy and health systems research and equity, with extensive experience in primary care throughout the world, including the Brazilian Family Health Strategy. He is currently Professor of Health Policy and Management and Community Health Sciences at the UCLA Fielding School of Public Health. He previously held positions as Associate Professor of Public Health and Health Policy at New York University and former director of the NYU MPH program. He was a Robert Wood Johnson Foundation Health and Society Scholar at the University of Pennsylvania from 2006 to 2008 and a Fulbright Scholar in Brazil in 2002. He has worked and still works on several collaborative projects with Brazilian public health researchers and has been involved in both formal and informal consultations with the Department of Basic Care at the Brazilian Ministry of Health (DAB/MoH). This includes his participation in an international seminar on the PMAQ organized by the DAB/MoH and official support to lead the edition of JACM’s supplementary number on PMAQ, published in 2017.

This interview started at the Brazilian Collective Health Association Conference (ABRASCO) held at the Oswald Cruz Foundation (FIOCRUZ) in Rio de Janeiro in late July, 2018, and continued in the following two weeks via e-mails. Although directly based on the Brazilian PMAQ, we cover relevant themes concerning the adoption of PBF/P4P programs worldwide, such as the importance of the program; why it has been developed in the primary care context; risks and results involved in the implementation of the pay for performance program; lessons and recommendations for moving forward; contributions from Social Science researchers. We hope that policymakers, researchers and health workers involved in P4P/PBF programs in other countries will find it relevant and insightful.

Fabiana Saddi (FS): What, in your opinion, is the importance of the PMAQ program?

James Macinko (JM):The Programa de Melhoria de Acesso e Qualidade da Atenção Básica or PMAQ represents, in many ways, the next step in Brazil’s transformation of its national health service from one that sought to increase access to a unique model of community based primary care to that of a “learning health system” capable of producing team- level data to health personnel and managers in a way that should allow them to develop approaches to identify and improve problems and to innovate in order to enhance the overall process and quality of care. This is particularly important because as Family Health Strategy (FHS) expands it is confronting more complex health problems, such as those of chronic disease and comorbidities. It is also confronting an economic and political context where even well-performing public programs, like the FHS, are under increased scrutiny. The evidence base for the effectiveness and impact of the FHS are very strong, yet decision makers (and many members of the Brazilian population) are either unaware of these accomplishments or unwilling to entertain the notion that this program, which is estimated to consume less than 20% of the national health budget (which at 4.5% of GDP is already extremely low in comparison with other middle income and Latin American countries) has already enhanced healthcare access, contributed to reductions in infant and adult mortality, and dramatically reduced inequalities in healthcare access for a majority of the Brazilian population.

FS: Why do you think the PMAQ was developed?

JM:There are some interesting reasons for the development of the program, including Brazil’s highly decentralized healthcare management (its 5,500 or so municipalities are largely responsible for the day-to-day management of primary care) and the fact that financing comes from the federal, state, and local levels. This creates an important principal-agent problem from the point of view of federal authorities who are responsible for working with local health secretariats to set performance targets and to meet national objectives—even if some of these priorities may differ at the local level. In addition, while the FHS has had several periods of rapid expansion, it has stalled at around 60 percent population coverage. Perhaps more importantly, a number of studies have documented important variations in the quality of care provided through the FHS.

FS: What was in place before the PMAQ?

JM:There was already a history of federal incentives to accelerate local adoption of the new primary care model (The Family Health Strategy). For example, while federal transfers include what is known as the PAB (a per capita payment to municipalities for primary care), they also included a number of components of a variable payment scheme for achieving certain additional objectives. One such adjustment was to adopt and then increase coverage of the Family Health Strategy with increasing incentives for municipalities attaining higher coverage.

While these incentives seem to have accelerated the rapid adoption and diffusion of the FHS, other important federal initiatives assured that flexibility at the municipal level did not circumvent the adoption of the program by deviating too far from the basic model tested and originally developed in the state of Ceará. This included development of national guidelines for the construction and equipping of a primary care post/center, a set of clinical guidelines for common conditions, ongoing education and other incentives to assure that different types of health personnel were adapting to the new model of community based care, and the development of a national primary care policy--which, according to many commentators, has been considerably weakened in its 2018 revision-that defined what was meant by primary care, its role in the national health service, and the roles and responsibilities of each sphere of government.

FS: Do you think there any risks to introducing pay for performance in a National Health System like the SUS?

JM:The use of capitated payments (a single yearly payment for each patient covered by the municipality--the current and dominant form of primary care financing in Brazil) can create certain unintended consequences. Capitated payments can lead to healthcare providers doing the minimum possible—or to favor less complex or difficult interventions or patients, since resources are limited and there are few incentives to do otherwise. Early in the expansion of the FHS some of these potential distortions were countered by additional incentives for developing programs for certain high-risk groups (e.g. hypertension and diabetes) and special priority populations such as those for Tuberculosis or Hansen’s disease. Other risks come from the fact that the organizational and capacity of different health teams can vary within municipalities and that municipalities themselves have vastly different levels of management capacity. One risk with such diversity is that better-resourced municipalities may simply be better able to consistently gain incentive payments, which would further increase differences between the highest and lowest performing municipal health systems, rather than narrowing the gap. One way the MOH has tried to avoid this is to stratify municipalities by their overall socioeconomic level Import Imageand to calculate quality scores within each of these socioeconomic strata, but to date, it is unknown how equitably PMAQ resources have been distributed across different settings.

FS: What have been the results so far?

JM: Perhaps one of the most important achievements has been to increase funding to the FHS, both through incentive payments and those intended to enhance basic health facility infrastructure, strengthen supply chains, stimulate the use of EHRs, and begin the process of creating a culture where data are used for routine decision-making. The program has also generated a lot of data that have been used for many purposes and developed a core team of researchers, government officials and health professionals with greater stakes and engagement in improving access and quality of primary care.

FS: Are there any lessons learned or recommendations for moving forward with the PMAQ?

JM: First, I think that efforts should be made to make the current method of data collection more transparent and independent. Currently, some of the data for evaluating healthcare teams relies on these same teams reporting a great number of indicators, both online and during the external evaluations conducted by independent universities. This can lead to intense pressure at the health center level for everyone to provide the best face to quality and access problems because that should lead to greater funding. One way to remedy this would be to rely more on review of Electronic Health Records and to base assessment on at least some of these clinic-level indicators. This would require additional investment in data extraction and analysis but could not only provide more objective data but also contribute to the continued development and use of these fairly new health information systems increasingly adopted in most municipalities. Data collection and target setting also needs to begin to focus on important national priorities, such as quality of care for older individuals—an area that receives very little attention in the current scheme.

Second, more needs to be done to understand whether incentives are really driving quality improvements that matter to improving population health. This is a basic problem inherent in many target-based pay for performance programs: health care providers will perform the activities related to incentives, but may not be willing or (due to time constraints) able to perform other vital functions. Perhaps even more importantly, there is the possibility of perverse incentives whereby municipalities may have incentives to invest in the highest performing teams to assure they continue to receive full incentive payments, rather than to do the possibly more time consuming and resource intensive work of bringing the bottom-performing teams up to speed.

To date, data from the external evaluations have shown a number of deficiencies in terms of basic supplies and equipment and documented the existence of many performance improvement initiatives. But there are still no studies that demonstrate whether the incentives themselves are associated with better quality of care, whether such quality improvements have been maintained over subsequent PMAQ cycles, or whether any of the observed quality improvements have actually resulted in better health outcomes.

Third, greater understanding is needed of the way the incentive payments are actually being used. This is another area not captured in current evaluation schemes. Federal guidelines leave it to the municipalities to decide how best to use incentive payments. Incentives can be used as salary bonuses to the highest performing teams, as investments in infrastructure or equipment, as a means to develop new teams or provide support to existing ones, or they can be put to other uses that the municipality deems important. While it may be the case that different strategies are needed in different municipal contexts, there is very little known about which strategies really stimulate quality improvements and which may create tensions among health personnel or between health workers and municipal managers.

Finally, more effort is needed to communicate the important work of the FHS to users and policymakers. At the outset of the PMAQ, an extremely ambitious online portal was created to allow users to access data on the quality assessments of their own health centers. It is currently not known whether such data are accessed, understood, or used by the population to make decisions about seeking healthcare. Such efforts could potentially go further, including greater attention to community outreach efforts to promote the importance and benefits of a strong primary care foundation. Interestingly, some of Brazil’s private health plans have begun to move in this direction by introducing primary care doctors (many of whom might have gone into the public sector) who take the role of one’s personal physician. I’m optimistic that continued efforts will drive increased demand for this robust model of primary care and the evolution of programs such as PMAQ will stimulate the changes needed to assure the FHS will be able to meet emerging needs.

FS: This is a Brazilian Social Science journal. A good number of Brazilian Social Scientists --from Anthropology, Sociology, Political Sciences will be reading this interview. How do you think the Brazilian Social Science could contribute to increase our knowledge about PMAQ?

JM: There is a great need for social scientists to bring their strong theoretical frameworks and diverse analytic tools to this project. In terms of Economics, there are many unanswered questions about the impact of financial incentives on firm behavior under conditions of uncertainty. From Political Science, it seems that there could be important insights gained from application of tools such as Game Theory to understand the complexity of the team-municipal-federal relationships and how each level may try to gain advantage. To date, there is also little discussion of how local politics may play out at the municipal and health team levels. Anthropology has an important role in helping us to understand the meaning that different actors in the system ascribe to the diverse aspects of the PMAQ process that only in-depth qualitative research can reveal.

From Sociology there are important organizational and institutional analyses, and a clear need for understanding how underlying structural inequalities in municipal capacity may affect team performance, ability to perform quality improvement activities, and resulting reactions to the evaluative process. In short, PMAQ represents a golden opportunity for a multidisciplinary social science analytic program to understand the particulars of the program, but also to help inform important social and economic theories at the individual, organizational, and governmental levels.