Deepening our understanding of community response to mass health interventions

Invited Commentary on ‘Refusal of oral polio vaccine in northwestern Pakistan: a qualitative and quantitative study’ by Murakami et al.

Research of the kind presented in this paper (Murakami et al.) — analysis of what influences households to accept or refuse child vaccination — is welcome. The global programme for polio eradication (GPEI) still struggles to understand the complex socioeconomic, political and cultural factors that determine vaccine acceptance or refusal at household and community levels — in particular in areas where that complexity is compounded by wider religio-ideological and geostrategic power struggles. The mixed methods approach applied in this case, is especially encouraging as, too often, social research on mass vaccination programme demand has relied rather heavily on qualitative findings.

In certain ways, the paper foreshadows subsequent action to improve understanding of why households refused vaccination. The 2010–2012 GPEI strategic plan included basic communication indicators to assess the reasons for missed children after each immunization campaign. Today, independent monitors regularly interview parents on reasons for non-vaccination and health workers have to provide reasons for not vaccinating a child on their route. While still imperfect such tracking has served well to identify clusters of refusal households and develop a range of programme responses from training for health care workers to the engagement of local leaders to help resolve local issues.¹

That said that there are two aspects of the paper that we question. The first is whether the study’s design has inadvertently incorporated assumptions about what constitutes the core pathway to refusal, resulting in conclusions which do not add much to the standard answers offered by the predominant demand-side paradigm. The second is a matter of timing which in the context of the polio programme is non-trivial.

The grounded theory on which Murakami’s research is based focuses on refusal as the product of ‘facts’ in the minds of residents which feed ‘community interpretation’, and ‘manifestations of … refusal’. It is a neat theoretical model, but falls short on detail and nuance. Firstly, it under-represents the range and complex interplay of subjective and objective influences that we know from wider research and programmatic analysis shape household and community attitudes. In the grounded theoretical model, three OPV facts lead to five - essentially negative - community interpretations — all related via ‘rumour’ and ‘suspicion’ to Islamic doctrinal concern. This seems a somewhat skewed finding when we consider that, in 2013, almost 40% of household refusals were related not to religion, but to rational objections about absence of non-polio health services, prevailing insecurity, and political opposition to the programme.²

In addition, the model marginalizes the role of wider ideological, political and geostrategic struggles which contextualize the programme in Khyber Pakhtunkhwa (KP), in shaping how households and communities behave. The study appears to proceed from the assumption that a household’s ‘refusal’ to vaccinate is based on information, ideas and attitudes located within the household or community itself — while evidence suggests that refusal is as likely to be the result of the influence on households of local religious or militia leaders who have a bone to pick with Islamabad (over distribution of administrative power, secular versus religious authority and the federal government’s association with foreign powers).²

What we see again and again across the polio programme in its last endemic redoubts is that the attitudes of households and communities to the programme are shaped by multiple factors — without doubt including personal beliefs based in part on a combination of doctrinal commitment and incomplete or inaccurate information, but as well on rational judgement about the disproportionate focus on OPV delivery relative to the failure of other basic vital services, and on the explicit or implicit directive of local religious, traditional, or military powerbrokers, many of whom see compliance of communities with the global programme as a bargaining chip in the wider political struggles which characterize the region.

Turning to our second question on timing, it has been seven years between this paper’s field research and publication and much has changed. At the time the research was in process, Swat was without doubt a major area of concern for the polio programme, with military operations in 2007–2009, major population displacement and subsequent floods. But in the intervening period, there have been no WPV cases in Swat, Buner or Shangla Districts since 2012⁴ and in 2012–2013 they were estimated to have among the lowest rates of polio vaccine refusal in KP.² Furthermore, the front-line between local powerbrokers and insurgents and the Pakistani state and its international allies has moved and while KP remains key to Pakistan’s chances of eradicating polio the epicentre has shifted south, around the densely populated Peshawar valley and into the Waziristans.³ At this stage, in the global eradication end-game, time — and speed — are of the essence. Instructive insights eventually transmogrify into matters of historical record and the currency of research depends on the speed of dissemination.

We need more research of this kind. But we need it quicker; and we need it to start from a broader understanding of what determines household and community orientation to mass health interventions, including — critically — more than a ‘blame-the-victim’ narrative of suspicion, rumour and misconception.