Leveraging contact network structure in the design of cluster randomized trials

Abstract

Background: In settings where proof-of-principle trials have succeeded but the effectiveness of different forms of implementation remains uncertain, trials that not only generate information about intervention effects but also provide public health benefit would be useful. Cluster randomized trials (CRT) capture both direct and indirect intervention effects; the latter depends heavily on contact networks within and across clusters. We propose a novel class of connectivity-informed trial designs that leverages information about such networks in order to improve public health impact and preserve ability to detect intervention effects. Methods: We consider CRTs in which the order of enrollment is based on the total number of ties between individuals across clusters (based either on the total number of inter-cluster connections or on connections only to untreated clusters). We include options analogous both to traditional Parallel and Stepped Wedge designs. We also allow for control clusters to be "held-back" from re-randomization for some period. We investigate the performance epidemic control and power to detect vaccine effect performance of these designs by simulating vaccination trials during an SEIR-type epidemic using a network-structured agent-based model. Results: In our simulations, connectivity-informed designs have lower peak infectiousness than comparable traditional designs and reduce cumulative incidence by 20%, but with little impact on time to end of epidemic and reduced power to detect differences in incidence across clusters. However even a brief "holdback" period restores most of the power lost compared to traditional approaches. Conclusion: Incorporating information about cluster connectivity in design of CRTs can increase their public health impact, especially in acute outbreak settings, with modest cost in power to detect an effective intervention.