Recognizing research paradigms: Session summary continued

Recognizing research paradigms, methods and impact for people-centred health systems – panel discussion

Third Global Symposium on Health Systems Research, Cape Town, 2 October 2014

Following presentations by James Macinko and Rene Loewenson, session chair Barbara McPake opened a discussion with panellists Bruno Marchal, Clara Mbwili-Muleya, Kumanan Rasanathan using a variety of questions received through Twitter and from the floor.


Question: Whose knowledge matters, who should engage in the process and what is contributed?

From her perspective as a frontline health worker using and implementing health systems research, Mbwili-Muleya noted that as the district level is the most important point of action and need for sustainable change in Zambia’s health system, participatory action research (PAR) should be applied as much as possible at this level. Knowledge is produced from actions in communities, much of which is not anticipated, so involvement of the people it touches is essential.

Rasanathan agreed, noting that health systems are fundamentally about politics and that a PAR  focus on incentives and structures in the district health system would direct transformation for people-centred health systems (P-CHS). People participate because they want change, so transformation should be a key determinant of methodology.

Marchal observed that research methodology and policy-making challenges are socially complex, such that the solutions should accordingly be addressed within a complex systems framework. Only intensive groundwork through PAR in local contexts, followed by de-contextualisation for replication, will achieve this.

Loewensen cautioned that all knowledge and power at local and higher levels matter, so allowing one group’s knowledge to subjugate the experience and influence of another is problematic. There is a rich store of local knowledge to be tapped, but because much of it is seldom implemented, it can take years for basic resources to be supplied. 

Macinko concurred, saying that inclusive fora for dialogue and networking to generate data and implementation must be created.

Question: How do we make data work for people – how should we receive and apply it?

Mbwili-Muleya said that data can inform better ways of working that will positively impact communities. We have health information systems, but health workers cannot analyse the data, and in some cases the data might not be useful for their day-to-day application.  It is important to consult end-users at the outset and involve their understandings in generating data, and to produce user-friendly information for informed decision-making.

Rasananthan observed that we tend to fetishise data in health systems on the premise that “what gets counted gets done”. We mourn the scarcity of local data, but we are not doing enough to gather it routinely, nor do we interrogate whether high-tech journalised research data are helpful for health workers. New technologies should be deployed to address this.

Question: Are the data used for holding authorities to account, and should researchers be engaging in such advocacy more forcefully?

Loewensen’s view was that researchers have the responsibility of facilitating the availability of specific P-CHS data, but in order to build P-CHS on society’s terms, communities should hold governments accountable.

Macinko noted that researchers are often advocates for evidence-informed change, but that this should be stimulated, whilst being cautious of compromising their objectivity. 

Marchal held that as citizens, researchers have a duty to mobilise for change, although it is true that their vested interests as scientists can be disrupted by such a role.

Question: How do institutionalised bodies like UNICEF prioritise and commission knowledge production?

Rasananthan confirmed that UNICEF, as an agency with large capacity and presence, is committed to studying macro-level issues of social justice and commissioning national and local studies.

Mbwili-Muleya emphasised that Africa and the developing world need more research capacity at district level.  PAR has worked at local level to render community enquiry and data, but this knowledge must be moved up to the level of policy-making and programming – and the capacity for this process does not exist.  Frontline workers have a multitude of research questions but no capacity to pose them to bodies like UNICEF. Rasananthan responded that institutions must actively seek and build such capacity, noting that the Symposium series has been making significant progress in this direction over the last four years.

Marchal proposed that this issue also centred on the channels and people involved in priority-setting for research funding agendas, and the need to ensure that such priorities are socially relevant.

Loewenson noted that research should not be segmented. Community members must be incentivised and paid to be involved in health planning and research.

Macinko recommended that capacity-building be an integral part of research programming so that it becomes part of the project ethos.  Funders and researchers want to hear about health workers’ research ideas and problem statements, because they are aware of the gaps in local priorities for knowledge.

Question:  Is there any role for high-level literature outputs in the P-CHS knowledge endeavour?

Rasananthan believed that there is – there is an appetite for such content in widely read publications such as The Lancet, which is known to be a space for contesting and assimilating findings and methods – but these are not the only destinations that should be explored.

Loewenson confirmed that local PAR evidence has been used to generate high-level studies, but the issue is how outputs in journals like The Lancet is being translated into implementation and impact on the ground, so it remains important to investigate the power and control represented by publications like these.

Macinko proposed that the parameters for this exercise be broadened.  PAR and primary health care work is often much more valuable than other forms of research, but it needs to be condensed without loss of richness.  The challenge is to convince the high-level journals and their readers that PAR work strikes at the heart of health matters.

Quoting Macinko’s earlier statement that “we believe we’ve learned a lot”, McPake encapsulated the abiding questions that underpinned the session discussion as follows:

  • What do we really know about what we know?
  • How to we induce people to think harder and more openly about research practice for P-CHS?
  • What works where? (We might envisage a beautiful endpoint in principle, but we need detailed understandings for scale-up and replication.)
  • Who should use the information?
  • By what criteria can we judge common and different issues?
  • Why do we want different definitions of research paradigms and practice?
  • Questions of ethics and the political economy were raised but not answered.



Page updated: 4 October 2014