Validation of Inverse-Probability of Missing Data Approach for the Inclusion of Laboratory Data in Healthcare Database Research

Validation of Inverse-Probability of Missing Data Approach for the Inclusion of Laboratory Data in Healthcare Database Research. Aim 1) To construct rich predictive models for the availability of each of the laboratory test results of interest. Aim 2) To compute a distribution of laboratory results for each of the laboratories of interest in a sample in both an unweighted and an inverse-probability re-weighted sample. Aim 3) To compare distributions of laboratory values in the reweighted sample the distribution of lab results from NHANES.

Principal Investigator: Alan Brookhart, PhD.
Funding: AHRQ
Total Project Period: 07/19/11-07/18/12