Vertical dominance: Cost-effectiveness, randomisation, and the bias against horizontal interventions in global health

Interventions in global health are frequently divided into two categories: vertical, which address one disease, and horizontal, which tackle multiple health problems through the building of health infrastructure. When identifying interventions to fund, global health practitioners place great weight on cost-effectiveness, which is determined through cost-effectiveness analyses. These analyses frequently draw on data from randomised controlled trials (RCTs), as they are considered the gold standard for determining causality. I argue that the use of RCT data in cost-effectiveness analyses leads to a bias in favour of vertical interventions and against horizontal interventions. This is because it is significantly easier to randomise vertical interventions compared to horizontal ones, so analyses that draw on RCTs will preferentially report on vertical initiatives. This bias contributes to a trend of underfunding horizontal interventions in global health. I argue that this trend is problematic, as horizontal interventions have the potential to be highly cost-effective. Finally, I argue that global health practitioners should find effective ways of evaluating horizontal interventions to ensure their benefits are captured.