Validity of Claims Data for the Identification of Male Infertility.

Abstract

We sought to determine whether infertile men can accurately be identified within a large insurance claims database to validate its use for reproductive health research. Prior literature suggests that men coded for infertility are at higher risk for chronic disease though it was previously unclear if these diagnostic codes correlated with true infertility. We found that the specificity of one International Classification of Disease (9th edition) code in predicting abnormal semen parameters was 92.4%, rising to 99.8% if a patient had three different codes for infertility. The positive predictive value was as high as 85%. The use of claims data for male infertility research has been rapidly progressing due to its high power and feasibility. The high specificity of ICD codes for men with abnormal semen parameters is reassuring and validates prior studies as well as future investigation into men's health.