Can conditional cash transfers contribute to health systems strengthening?

Underlying the implementation of CCTs and P4P is the question whether these particular health investments are producing desired health outcomes, such as reduced maternal mortality and infant mortality and whether this is good value for the money. All too often health systems pay for what is needed to produce health services and not for their “performance” or outcomes and in many contexts financial allocations to primary care and hospitals are based on inputs, such as number of staff or hospital beds with no link to whether services are delivered.

There is an increasing interest in the potential of CCTs/P4P and other incentive mechanisms to strengthen health systems and in many developing and transition countries. As such they represent system-level interventions targeted specifically at the health financing building block of the health system, yet having ramifications throughout the entire health system, including its governance, human resources, information sub-systems, and service delivery.

In various settings trials and experimentation based on CCTs/P4P is underway with the objective to improve health systems performance. In November 2008, a symposium organised by Swiss Tropical Institute (STI) was dedicated to reviewing some of these experiences and speculating on the way forward. The Symposium took an in-depth look at current experiences and knowledge on supply (P4P) and demand side (CCT) incentive mechanisms to increase the quality and coverage of the population with health services and priority home–based interventions. The one and a half day symposium exposed and examined the measurement and metrics of the implications of incentives on governance, human resources, financing, and donor support of health systems as well as the evidence base and emerging implementation research agenda for informing policy on incentive systems.

In the next sections of this issue of the bulletin Medicus Mundi, some of the experiences with CCTs and P4P schemes are highlighted and analysed. A first paper by Berit Kieselbach and Andreas Kalk points out that result-based payments are not an innovation but have been used for centuries. With regard to health systems they argue that they might provide a window of opportunity for broader management and organisational changes but are also reliant on a comprehensive political framework for example in the area human resource development.

Positive results

Several contributions in this issue are dedicated to supply-side (provider P4P) interventions. Judith Cowley looks at experience from Burundi in the area of skilled birth attendance, hospital referral and traditional birth attendant. Her contribution shows that although health worker incentives can contribute to safer motherhood, other interventions such as increasing the numbers of skilled staff and ensuring adequate basic and post-basic education programmes are equally if not more important so to achieved desired health systems outcomes. Bonaventure Savadogo and colleagues analyse experiences from Rwanda with a P4P scheme and conclude that P4P schemes may show immediate positive results, such improved remunerations for health workers and higher health service attendance, but that these results can not be seen in isolation and have to be related other on-going and overlying health reforms.

With regard to demand-side (patient/population) CCT interventions, the article of Franco Pagnoni presents the potential of community health workers in reducing malaria-related mortality. He summarises critical areas in relating to community workers such as high attrition rates and points towards their potential in improving prompt access for the population to effective malaria treatment. Mohammad Iqbal and colleagues review the role of skilled birth attendants in reducing maternal mortality in Bangladesh and conclude that a performance based payment scheme through vouchers for the poor can improve access to care. Raffael Aye looks in his contribution on the potential of food complements for tuberculosis patients in Tajikistan so to increase treatment adherence thereby stressing that incentives at patient level can substantially decrease household costs related to illness leading potentially to better health outcomes.

What about cost-effectiveness

CCTs schemes are dependent on well-designed information systems allowing judgment on performance. The article of Carla AbouZahr draws the attention to the need for better statistics to track health-system performance thereby also ensure accountability of incentives been awarded to providers and patients.

More generally, little is known about the cost-effectiveness of CCT and P4P schemes, and whether they represent good value for money in relation to other possible investments in health system strengthening. Progress in these areas will require the mobilization of technical and financial support for countries as well as support for the process of improved health-system monitoring. Albeit there is a threat that CCT schemes impose substantial administrative and transaction costs for monitoring performance, it is necessary to improve information systems, particularly about measuring and managing outcomes. In other words, any approach to CCTs will need to find the appropriate balance between administrative costs for tracing the performance based funding and the need to strengthen monitoring systems so to follow up on a set of specific health systems outputs.

CCTs in the health sector can be provided to the demand-side (the public), incentivising increased health service utilization by patients, and/or to the supply-side (the providers), to promote health service provision. The current evidence from lower income countries points to improvements in specific health service utilization for selected services from demand side CCTs and increases in coverage of selected health services resulting from supply-side CCTs when applied to government facilities and non-governmental facilities and improvements.

However, in addition to this recognised potential benefit of CCTs, there are also a number of potential risks, particularly in P4P schemes. These include:
• potential for crowding out or diverting health worker behaviour away from other services;
• potential supplier–induced demand leading to excessive provision of services to those who are not in need;
• potential compromise of quality in order to maximise output;
• potential compromise on equity in order to achieve efficiency goals;
• temptation to artificially inflate performance reports

Potential to improve health systems performance?

During the STI symposium and across the articles of this issue of the Bulletin Medicus Mundi there is a strong message of caution with regard to the value of CCTs and P4P. A common theme was that more evaluation and research was necessary to fully understand the implications, benefits and risks of this intervention. Insufficient evidence is currently available and inadequate research and evaluation is underway. Current evaluations generally suffer from methodological weakness, notably a lack of control group, and inability to control for possible confounders, limiting the conclusiveness of evidence in relation to impact. Evaluations are often funded late and usually start mid-stream. They often lack base-line and counter-factual information. Finally, they often fail to appreciate the effects of the intervention on other building blocks of the health system beyond the financing and service delivery building blocks (e.g. governance, informatics, human resources, etc.)

Overall, it was recognised during the 12th STI Symposium, that health care delivery systems show potential to improve performance. In low- and middle-income countries financing and delivery mechanisms in place have typically failed to protect poor and vulnerable groups and to offer equitable and efficient services to the population. There appears to be increasing recognition of these failures and CCTs/P4P are seen as possible approaches to contract with providers and incentivise patients in ways that reward improved resource allocation. Whereas some argue that cautious and carefully designed, implemented and evaluated change involving the application of CCT and P4P incentives is essential to reducing inefficient health care delivery there is hardly any consolidated evidence available on their potential to improve health systems performance. Thus additional research and documentation is necessary if and how newly introduced incentives do not produce perverse outcomes conflicting policy objectives and induce behaviours consistent with policy goals.

*Kaspar Wyss and Don de Savigny, are both working at the Swiss Tropical Institute (STI) and are interested in health systems performance and monitoring. They co-organised the 12th STI Symposium on 27-28 November 2008 entitled “Health System Strengthening: Role of conditional incentives”. Contact: Kaspar.Wyss@unibas.ch; D.deSavigny@unibas.ch