Missing Outcome Data in Epidemiologic Studies

Abstract

Missing data are pandemic and a central problem for epidemiology. Missing data reduce precision and can cause notable bias. There remain too few simple published examples detailing types of missing data and illustrating their possible impact on results. Here we take an example randomized trial that was not subject to missing data and induce missing data to illustrate 4 scenarios in which outcomes are 1) missing completely at random, 2) missing at random with positivity, 3) missing at random without positivity, and 4) missing not at random. We demonstrate that accounting for missing data is generally a better strategy than ignoring missing data, which unfortunately remains a standard approach in epidemiology.

Keywords: bias; error; generalized computation; imputation; missing data; precision.

Figure 1

Causal diagrams for possible missing-data scenarios in the IPOP Trial. W denotes the covariate short cervix, A denotes treatment with 17α-hydroxyprogesterone caproate, Y denotes preterm birth outcome, and M denotes a missing value for the outcome. Boxes denote restriction to observed data. IPOP, Improving Pregnancy Outcomes With Progesterone.