Fig. 5

Implementation science study designs that leverage instrumental variable methods

The use of instrumental variable methods in these designs helps to illustrate how the efficacy of the same evidence-based interventions may be impacted by the implementation strategy used to promote its uptake. This is because implementation strategies may function via different mechanisms, target different populations, and also be utilized during different stages of uptake (i.e., early vs. late). A illustrates the study design for the manuscript by Mody et al. assessing the impact of implementing new HIV guidelines in Zambia. B illustrates a hypothetical example using an alternative implementation strategy—clinic-level practice facilitation—for the same EBI—same-day ART initiation. Similarly, C illustrates the study design for the cluster-randomized trial for a mask promotion strategy by Abaluck et al., whereas D depicts a hypothetical example with an alternative implementation strategy—a mask mandate—for the same EBI—individual-level mask use