Updated guidance from the agency for healthcare research and quality effective healthcare program on including nonrandomized studies of interventions in systematic reviews: a work in progress

Abstract

The updated guidance from AHRQ is welcome and timely, given the increasing interest in using evidence from nonrandomized studies to evaluate health and social care interventions (NRSI) [1]. It is particularly heartening to read endorsement of recent insights both about the opportunities afforded by such studies and the potential pitfalls [2,3]. Of special note are recognition of the “three unique categories of bias” that arise before or at assignment to intervention or comparator status and which threaten the validity of NRSI [3] and the principle of using “study methods rather than study design labels to differentiate among NRSI types” [4].