Injuries/Violence
Population health in armed conflict: How to conceptualize and measure a prominent threat to global health Maya Luetke* Maya Luetke Signe Svallfors Elizabeth Heger Boyle
The humanitarian impact of armed conflict remains a significant international issue, with an estimated 2 billion people residing in fragile or conflict-affected settings. Despite growing research on armed conflict and its impact on human populations, few studies have explicitly considered the methods necessary to assess such relationships (i.e., how to use disaggregated and granular conflict data, measure and operationalize conflict, and other potential considerations). In this study, we identify key methodological considerations for conducting research on the effect of armed conflict on population health, including how data structures, study design, and methodological decisions might impact conclusions. First, we discuss particular characteristics of existing armed conflict datasets, such as identifying and dealing with the spatial and temporal imprecision of conflict events within such datasets. We demonstrate how different data and measurement choices may result in spatial misclassification, bias, and spurious conclusions. Lastly, we provide an empirical example using armed conflict data from the Uppsala Conflict Data Program (UCDP) and combining it with health data from the Demographic and Health Survey (DHS) data in Nigeria. We quantify the impact of armed conflict on current contraceptive use and demonstrate how different measurement and data choices impact the point estimate of the relationship between our exposure and outcome. We find that measurement choices do indeed impact the magnitude and significance of the point estimates of the association between armed conflict and current contraceptive use among women of reproductive age (15-49) in our empirical example (Figure 1). Our findings reinforce the importance of deliberate methodological decision-making and transparency in describing such decisions (including the rationale) so that future research can build reliable and comparable evidence on the impacts of armed conflict on population health. In this project, we recommend concrete strategies – and an empirical worked example – for employing such data for public health research.
