Absolute and Relative Bias in Eight Common Observational Study Designs: Evidence from a Meta-analysis

TL;DR

This meta-analysis compares bias and variance across eight common observational study designs, revealing that incorporating more design elements reduces bias and variance, but confounding remains a concern.

Contribution

It systematically evaluates how different design elements affect bias and variance in observational studies through a comprehensive meta-analysis.

Findings

Bias decreases with more design elements

Lowest bias observed in designs with all three elements

Bias within 0.10 SD in 59-83% (Bayesian) and 86-100% (non-Bayesian)

Abstract

Observational studies are needed when experiments are not possible. Within study comparisons (WSC) compare observational and experimental estimates that test the same hypothesis using the same treatment group, outcome, and estimand. Meta-analyzing 39 of them, we compare mean bias and its variance for the eight observational designs that result from combining whether there is a pretest measure of the outcome or not, whether the comparison group is local to the treatment group or not, and whether there is a relatively rich set of other covariates or not. Of these eight designs, one combines all three design elements, another has none, and the remainder include any one or two. We found that both the mean and variance of bias decline as design elements are added, with the lowest mean and smallest variance in a design with all three elements. The probability of bias falling within 0.10 standard deviations of the experimental estimate varied from 59 to 83 percent in Bayesian analyses and from 86 to 100 percent in non-Bayesian ones -- the ranges depending on the level of data aggregation. But confounding remains possible due to each of the eight observational study design cells including a different set of WSC studies.