In a newly accepted paper in the prestigious Annals of Statistics, Dr. Donna Spiegelman and her colleagues Drs. Daniel Nevo (Tel Aviv University) and Judith Lok (Boston University) have developed an innovative adaptive study design, the Learn as You Go (LAGO) design, which allows implementation scientists to optimize multiple component intervention packages for cost, power, and overall benefit in pre-determined stages, learning as the trial progresses.
More specifically, LAGO is a multi-stage, multi-center design that simultaneously identifies the optimal intervention as well as estimates its impact. Studies with such a design are carried out in stages after each of which the results are analyzed, the intervention package is reassessed, and a revised version is put to use in the next stage. LAGO thus provides rigorous methods to adapt to local, changing contexts, which is the usual setting in public health, and aims to reduce the chance of implementation failure for interventions whose efficacy has been proven at a smaller scale.
The authors illustrate their methods in the BetterBirth Study, which aimed to improve maternal and neonatal outcomes among 157,689 births in Uttar Pradesh, India, through a multi-component intervention package.
Citation: Nevo D, Lok J, Spiegelman D. Analysis of "Learn-As-You-Go" (LAGO) Studies. Annals of Statistics. 2020. Accepted. arXiv:1808.06310