Application of the building blocks framework in field studies
All three studies used a mixed method approach and used the Building Blocks as a conceptual framework. The measles study was primarily based on qualitative data, with most information collected through interviews and observations. The polio study used document review, interviews, and participant observation for qualitative data, and health utilisation data as well as national coverage data for quantitative analysis. The new vaccine study used semi-structured interviews, questionnaires and routine health service utilisation data. See Table
1 for more details.
Table 1
Study characteristics
Measles eradication[ 10, 17] | Semi-structured interviews with key informants (national, regional, district and facility levels) | Bangladesh, Brazil, Cameroon, Ethiopia, Tajikistan, Viet Nam |
Reviews of secondary documents |
New vaccine introduction
| Semi-structured interviews with key informants (national, regional and district levels) | Cameroon Ethiopia Guatemala Kenya Rwanda, Mali |
PCV
| Questionnaire with health facility staff |
Rotavirus
| Routine health service use data |
HPV
|
|
| Semi-structured interviews with key informants (national, district and community levels) | Nepal, India, Pakistan, Nigeria, Ethiopia, Rwanda, Angola |
Participant observation in polio campaigns, surveillance, and routine health post activities |
Reviews of documents |
Routine health service use data |
DPT3 coverage data from DHS and IHME |
| Attended births and antenatal care coverage data from DHS | |
For all three studies, modified versions of the Building Blocks framework were used to develop study instruments, categorise data, shape the analysis and guide policy recommendations. For example, in the new vaccine study the framework was modified to be specific to vaccination services, with sub-categories identified in each building block (e.g. within ‘human resources’ were availability of staff, training, remuneration, satisfaction and supervision). The questionnaires were structured according to the building blocks (see Table
2 for examples).
Table 2
Examples of questions structured according to the building blocks
Service delivery
| Do measles campaigns affect your capacity to reach remote areas for routine outreach services? | Has the number of outreach activities changed because you started offering the new vaccine? | Are routine immunization activities affected during polio campaign days? |
Health workforce
| Do measles campaigns take staff away from routine activities? | Did the training focus solely on the new vaccine or did it cover issues relevant for other vaccines or health services too? | Are health workers’ motivation levels the same as before polio campaigns began? |
Health information system
| Was there any change in the processes for identifying high risk groups and their vaccination coverage rates? | Have immunisation documents been reprinted to include the new vaccine? If yes, has this changed the time required and data completeness? | Has the surveillance system changed as a result of polio? |
Medical products, vaccines and technologies
| Have measles campaigns lead to additional infrastructure, such as waste management equipment? | Has the cold chain capacity related to the new vaccine had any impact for products other than vaccines, such as ARVs? | Have there been any changes in the cold chain infrastructure over the last 10–15 years? Are any of these changes a result of polio? |
Financing and sustainability
| Have donor funds been earmarked to measles campaigns? | Has funding requirements for the new vaccine affected the level of funding for other routine health related activities? | Is funding for polio separate from other health programs? |
Leadership/governance
| Do you think measles campaigns tend to strengthen or weaken policy processes? | Did the planning for the new vaccine have any effect on planning activities of other health services? | In the past, have government officials given a high level of attention to routine immunization activities? Has this changed as a result of polio? |
In the three studies, thematic content analysis was used to explore the interview data using the modified Building Blocks framework to map and chart data. As described in Table
2, this analysis was triangulated with data obtained through staff observation, document review, and staff facility surveys.
Discussion
Assessing the impact of a programmatic intervention on health systems can be a daunting task because of the absence of any controls or counterfactuals, and the complex, dynamic nature of health systems. On reflection, we found that many weaknesses of the Building Blocks framework resulted from its very elegance and simplicity. A framework is, indeed, needed to support both researchers and policy makers in explaining the reality of health systems, and to help simplify what is complex in order to facilitate policy making. Therefore, despite its many limitations, the WHO Building Blocks framework is valuable as it provides a common language and reference for researchers and policy makers. This is arguably of more use than the creation and use of different frameworks for each study conducted or policy issue raised. Nonetheless, the fact that it is a common framework means there are also common blind spots.
The simplicity allows specific effects of an intervention on individual building blocks to be described adequately. However, this comes at a cost. First, it does not allow to capture the dynamic interactions between the elements of a health system, a key feature that makes health systems complex. Second, the Building Blocks model does not provide any rationale of what makes health systems tick. As a result, the overall effect on the health system of a specific intervention might still be poorly understood and possibly even misinterpreted.
The Building Blocks framework considers health systems to be complicated, suggesting that one can describe the system by detailing all of the building blocks within it (“the sum of the building blocks is the whole system”). However, health systems are complex; like a living organism, they are dynamic, with interacting components - at various geographical levels - that lead to adaptation and to the emergence of new dynamics. These interactions can be both predictable and unpredictable. They generate feedback loops that will continue shaping the systems and its different components. The WHO Building Blocks framework does not capture this complexity and, consequently, is not well suited to research on the interaction of programmes with health systems.
To avoid falling into the trap of conceptualising health systems as a black box [
30], a health systems framework needs to make its assumptions concerning the role and relative importance of the components explicit, and especially how they interact. Indeed, the main weakness of the Building Blocks framework comes from the assumption that scrupulous description of specific effects on all the individual building blocks helps to understand the system as a whole. In reality, it simply presents a checklist of six functions. This is reinforced by most of the guidance on measuring health system strengthening that uses the Building Blocks framework, which just provide generic indicators for sub-dimensions [
31]. The framework, indeed, neglects a ‘whole system’ perspective.
We also found that demand-side issues were missing from the framework; others have also noted this [
18] and other missing elements, such as behaviour change [
32]. However, our criticism is not simply of the framework, but also of how it is used, particularly the focus on the six building blocks to the neglect of the process and outcome aspects of the framework.
The overuse of the Building Block framework poses a risk of considerable “group think” and may contribute to a lack of critical appraisal of health systems and a persistent view that health systems can be fixed as if they were complicated instead of complex (i.e. by only addressing individual components in silos, rather than considering the system as a whole).
There are a number of options for progressing the use of health systems frameworks in research. One option is to ensure that the existing WHO framework is used to its maximum potential. We believe that if the Building Blocks framework is to be used in research, it should be used in a flexible manner, including process and outcome components rather than focusing solely on the six key functions. Another way forward might be to modify the framework before using it for a new research study. Formative research could develop an explicit hypothesis of how the intervention under study may affect the health system, leading to a modified model, which then can be rigorously piloted. Issues of demand, power, process, decision-making and accountability should be explicitly considered in such a modified model. This would have the benefit of better focusing the study on issues pertinent to the topic under study whilst retaining the benefits of a common framework. A more radical option would be to identify an alternative framework that is better suited to research exploring the impact of programmes on health systems. One possible framework is that developed in the World Health Report 2000, which was used to assess the relationship between key functions and objectives of a health system [
33] in an attempt to evaluate change to the systems as a whole rather than to the sum of its parts. Alternatively, there may be scope for revising the Building Blocks framework, for example as suggested by Savigny & Adam [
31]. They acknowledge that it lacks interactions between components, arguing pointedly that it does not constitute a health system. They attempt to introduce dynamic thinking into the framework, but do not provide a clear way forward on how to operationalize this in a research project.
Summary
As with frameworks in general, the WHO Building Blocks framework is valuable because it creates a common language and shared understanding. Its simplicity and universality are its strengths. No framework is ideal, but a framework is only as good as the understanding it can generate. For applied research, the Building Blocks framework falls short of what is needed to holistically evaluate the impact of specific interventions on health systems. We believe it would be improved by making four amendments: integrating the missing “demand” component; incorporating an overarching, holistic health systems viewpoint; explicitly including considerations of decision-making and power; and including scope for interactions between components.
Nevertheless, the WHO Building Blocks framework should not be used automatically nor without careful consideration. We believe that if this framework is used, it should be in a more flexible manner than we and other researchers have done to date. It should be adapted according to the specific research question through formative research and piloting. It should be focused on systemic research hypotheses rather than exploring each and every possible effect on sub-components of the building blocks.
Continued critical reflection of and debate around its role and potential for development is necessary if the field of health systems research is to keep progressing.
Ethics
The manuscript is submitted in the “debate” section and therefore is not a research paper. However, the three studies that are discussed in the papers have all received ethical approval from their implementing institutions (the London School of Hygiene and Tropical Medicine (2009 and 2010) and Middlebury College (2011). Ethical approval was gained from all the countries where fieldwork was conducted, unless the country did not require an ethical approval.
Research projects discussed in the debate paper
Project: New Vaccine, from licensing to adoption: ethical approval from the LSHTM number 5739.
Project: Impact of measles eradication on health systems: ethical approval from the LSHTM number 5596.
Project: Impact of polio campaigns on health systems: Middlebury College, ethical approval number 11196.
Competing interests
We declare that we have no conflicts of interest in the authorship or publication of this contribution (financial or non financial).
Authors’ contributions
All authors have contributed substantially to the conception, design, writing/drafting and critical revising of the publication. All authors have given final approval of the version to be published and have agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. SMJ has led the drafting of the paper and UKG, SC, HB, and BM have all substantially contributed to various drafts of the manuscript and reviewed the final version of the paper.