Informative censoring by health plan disenrollment among commercially insured adults

Health plan disenrollment in commercial insurance databases occurs frequently and for heterogeneous reasons, but whether disenrollment is informatively censored by health status had not been systematically quantified in a large US commercial claims database.

Target RWE and collaborators analyzed 2 million patient-years of commercial insurance data to characterize predictors of within-year and end-of-year disenrollment, finding that within-year disenrollment is driven by health status while end-of-year disenrollment reflects plan characteristics.

Establishing that within-year disenrollment is informatively censored by health status provides Target RWE and pharma partners with the evidence base for selecting and implementing appropriate censoring adjustment methods in commercial claims studies.

Informative censoring characterization in commercial insurance for appropriate IPCW adjustment in claims studies, Evidence base for censoring adjustment method selection in Target RWE's commercial claims analyses