Background
Integration, and integrated health services specifically, are widely being promoted as a way to gain efficiencies, meet clients’ varied health needs and ultimately improve health outcomes. The U.S. Global Health Initiative calls for both “upstream” and “downstream” integration to coordinate and integrate health interventions [
1]. The Global Health Initiative, Paris Declaration, World Bank Sector Wide Approaches, the International Health Partnership Plus
a are models of “upstream” integration and coordination at the national level and higher, where donors, partner governments and other implementers harmonize financing and work together to develop and implement national health plans [
1‐
4]. In the U.S. President’s Emergency Plan for AIDS Relief (PEPFAR) five-year strategy, there is an important priority for “downstream” integration of family planning and HIV services at the health service and individual levels “so that women living with HIV can access necessary care, and so that all women know how to protect themselves from HIV infection” [
5].
Although these are some of the more recent endorsements of integration, in 1978, the Declaration of Alma Ata promoted a comprehensive approach to health starting with primary health care [
6]. While integration is promoted for its potential to provide greater access to more comprehensive care and to create health system efficiencies, there is a lack of consensus about the concept of integration and how to operationalize integration, and the field is lacking empirical evidence for effective models to guide decision making. More information is needed from integrated program designs about specifically what are the changes being made and to what program element (such as governance, funding, service delivery organization, etc.) [
7].
Strong monitoring and evaluation (M&E) systems can provide the information needed to assess progress, generate information for program management and decision making, and produce evidence of impact on health outcomes to inform replication and scale up [
8]. In the last decade, monitoring systems and indicators have proliferated in health, particularly for HIV, while research and evaluation has been neglected or implemented
post-hoc[
9]. Moreover, how building M&E systems from the outset alongside program planning can benefit program planning is not well understood or appreciated. However, new funding for HIV and AIDS control through global health initiatives has created opportunities for increased multi-sectoral participation, political commitment, and transparency in M&E systems [
10], and there is growing momentum to ensure that process and outcome or impact evaluations are planned from the outset to inform decision making [
11,
12].
To respond to the need for evidence of effective integrated services to inform the design and operationalization of integration, in this paper, we prioritize and organize existing M&E principles in to a systematic approach specifically relevant for health service integration initiatives. The primary audience for this approach is those program planners working at the national level. The approach is intended to help decision making related to what health services to prioritize for integration, what systems to strengthen to support integrated health services, and what data to collect to best monitor and evaluate integrated health services. Ultimately, the approach is intended to help establish an evidence base of what works to help inform decisions about these investments at both the national and international levels.
The approach outlined in this paper represents the ideal. In reality, each country will have performed each step at differing levels of depth and breadth. Thus, applications of the approach will reveal gaps. The approach is also iterative, in revisions can take place in earlier steps based on information gained in later steps.
Summary
Integrated interventions are fundamentally client centered and seek to improve the effectiveness and efficiency of providing a continuum of care to improve the health and well-being of those clients. Integration is a core Global Health Initiative principle and is viewed as a means to achieving critical public health goals, including the Millennium Development Goals [
58]. In this paper, we have attempted to apply basic M&E questions for national level M&E systems and M&E best practices embodied in existing frameworks to define a systematic approach at the country level for prioritizing integrated interventions, developing an M&E plan, and establishing an evidence base to inform those who manage these investments. The approach we outline is borne out of and consistent with single-sector M&E frameworks that employ a stepwise approach to M&E to ensure that information is available to inform decisions throughout the cycle of planning, monitoring, data collection, analysis, revision, and evaluation. The steps identified have been discussed in the light of the special considerations and modifications that make them necessary and effective for the M&E of integrated health programming.
The approach is grounded by first defining the health impacts integration is intended to affect. The mutual goal of designing, implementing, and scaling up interventions to improve a particular health outcome or impact is the “glue” that holds together disparate interests, services, and sectors. Second, it calls for a thorough understanding of the key point of contacts for adding new activities and services to maximize public health impact. A third advantage is the role for all levels of policy in creating “essential packages” of integration for all major client presentations. International minimum guidelines draw on the global evidence base and expert opinion. This international guidance informs individual country strategies and guidelines, including appropriate modifications to policies and standard operating procedures guiding practice at the service delivery level. Fourth, this paper describes the role of logic models to outline the plausible causal pathways and define the inputs, roles and responsibilities, indicators, and data sources across the health system. Finally, we recommend improvements to the health information system and in data use to ensure that data are available to inform decisions.
There are some notable differences and similarities between this approach and approaches to single sector (i.e., vertical, disease oriented) M&E. In both cases the country’s epidemiology informs the strategic planning and M&E goals and targets. The difference compared with single sector M&E is that this approach requires a more systematic assessment of the service delivery system entry points; the set of activities that need to be added; and the national and operational guidelines that need to be tested, revised, or created. Most importantly single sector M&E approaches are often driven by institutions external to the national governments, such as by donors, and those external institutional priorities might not reflect those of each country [
48]. This paper’s approach requires that program planning and M&E systems be determined based on national level priorities. While this may increase the likelihood that the programmatic response is tailored, it does not mean that donor funding will correspond with the country’s priorities. Further, there may be human and financial resource constraints that limit application of the approach.
Because the perspective of this approach is at the country level, it is consistent with calls for greater country ownership [
1,
4]. It is the country and its institutions that coordinate between global and national partners and where international partners bring their agendas in line with national ones, instead of vice versa [
59]. However, it will be a challenge to overcome the vertical, single sector approach to program planning and M&E and adopt practices that respond to country priorities and reflect a more integrated approach.
The approach represents the ideal, and thus it faces certain challenges. Despite a commitment in the public health community to the kind of logical and stepwise approach to M&E that we have recommended in this paper, execution of this kind of approach often falls short even for single service interventions, particularly at the country level where logic models and program impact pathways may be incomplete. Applying the framework requires making serious progress in overcoming parallel health information systems. Strengthening health information systems in general and ensuring that routine data systems are “interoperable” will result in high costs to donors, governments, and partners; the value of which is hard to communicate to donors and politicians [
59]. Successful application of this framework also requires collaboration and cooperation amongst stakeholders of all types and at all levels of public health policy, programming, and practice from the international, national, sub-national, and service delivery level. This framework must be taken to heart by the entire public health community and not just by single actors if it is to be successful in reframing thinking about how integration is implemented, monitored, and evaluated. Evaluation of integrated interventions will also require increased tolerance for the inability to attribute changes to a particular program or donor and for negative or unanticipated outcomes.
More experience is needed to understand better how the M&E approach we outline will be useful in the realities of countries. While there is a role for rigorous studies of integrated models, we argue that more emphasis is needed on documenting the process of designing and implementing integrated models at the national level that take into account the national epidemiological, health systems, social and political factors. The question shifts from “what is the effectiveness of the particular integrated model (compared with non-integrated)?” to “what is the effectiveness of integrated model designed for a specific context?” At the same time, the process by which the model was designed and implemented is thoroughly documented, in order to inform adaptation of the process (rather than a particular model) in another context. Continued rigorous testing of integrated models across various contexts will eventually, through systematic reviews, yield conclusions about the characteristics of the interventions and contexts that are favorable for implementation. But the complexity of integration and sheer number of possible service delivery combinations, health system improvements, and implementation contexts requires a different approach. Public health professionals will continue to build and fill gaps in the evidence base for what is new and what works in integration to promote the health and well-being of global communities, and the M&E approach that we have outlined can help contribute to that ongoing effort.
There are several development trends emerging which increase the likelihood of systematic planning and implementation of integrated intervention strategies, and the potential success of an M&E approach to integration such as the one we propose here. The aforementioned unprecedented and widespread international promotion of integration and country ownership as core principles to improve health outcomes is one of these emerging trends. Another is the attention being paid to health system strengthening in general, and not just for the purposes of integration. Strong health systems are required if moving to a more complex integrated, rather than vertical, approach to health care is to be successfully undertaken. There are also calls for more unified M&E systems, again in general and beyond the needs of specific integration efforts, to allow evaluation efforts benefit from and capitalize on all health information emerging from monitoring and reporting systems [
24,
60].
Competing interests
The authors declare the have no competing interests.
Authors’ contributions
HR and ES made contributed to the manuscript conception and design, drafted and revised the manuscript, and read and approved the final manuscript.