Use of Instrumental Variable in Prescription Drug Research with Observational Data: A Systematic Review

Objective: Instrumental variable (IV) analysis may be useful approach for addressing unmeasured confounding bias in observational studies of prescription drugs by substituting actual exposure with IVs that are: 1) strongly related to treatment assignment; and yet 2) uncorrelated with factors predicting outcome (key assumptions). Study Design and Methods: We conducted a systematic review assessing the use of IV methods, summarizing major types of IVs and evaluating evidence for assumptions. We searched MEDLINE, OVID, PsychoInfo, Econlit and economic databases from 1961 to 2009. Results: We identified 26 studies. Most (n=16) were published after 2007. We identified five major types of IVs: regional variation (n=8), facility prescribing patterns (n=5), physician preference (n=8), patient history/financial status (n=3) and calendar time (n=4). Evidence supporting the validity of IV was inconsistent. All studies addressed the first assumption; however, there was no standard for reporting evidence that IV sufficiently predicted treatment assignment. For the second assumption, 23 studies provided explicit argument that IV was uncorrelated with factors predicting outcome, and 18 supported argument with empirical evidence. Conclusions: Use of IV methods is increasing in prescription drug research. However, we did not find evidence of a dominant IV. Future research should develop standards for identifying IVs and reporting performance on key assumptions.