Dealing with missing data in observational health care outcome analyses

Observational outcome analyses appear frequently in the health research lit erature. For such analyses, clinical registries are preferred to administra tive databases. Missing data are a common problem in any clinical registry, and pose a threat to the validity of observational outcomes analyses. Face d with missing data in a new clinical registry, we compared three possible responses: exclude cases with missing data; assume that the missing data in dicated absence of risk; or merge the clinical database with an existing ad ministrative database. The predictive model derived using the merged data s howed a higher C statistic (C = 0.770), better model goodness-of-fit as mea sured in a decile-of-risk analysis, the largest gradient of risk across dec iles (46.3), and the largest decrease in deviance (-2 log likelihood = 406. 2). The superior performance of the enhanced data model supports the use of this "enhancement" methodology and bears consideration when researchers ar e faced with nonrandom missing data. (C) 2000 Elsevier Science Inc. All rig hts reserved.