Background: Studies of drug safety during pregnancy often seek to identify MCMs as a study outcome and may be conducted within health insurance claims data. A range of ICD-10-CM (International Classification of Diseases, Tenth Revision, Clinical Modification) codes for MCMs exist and are used to identify MCMs among infant claims. However, the use of these codes for a non-live birth on a maternal claim is uncertain. Further, ICD-10-CM guidelines specify that codes for congenital malformations can be used throughout life, meaning the presence of a MCM code on a claim may refer to the mother rather than the fetus.
Objectives: To determine the frequency of codes for MCMs associated with non-live births in claims data.
Methods: Pregnancies with the last menstrual period between 01 January 2016 and 31 August 2022 from US women of reproductive age in the Optum Research Database were included. Medical claims from these pregnancies were used to identify non-live births using the Optum Dynamic Assessment of Pregnancies and Infants process. From these non-live births, we generated counts of ICD-10-CM codes that could be for MCMs within 30 days before and 30 days after the non-live births. We further restricted these counts to exclude codes that were also present outside of the ± 30 days surrounding the non-live birth.
Results: Of the 1,238,600 pregnancies identified, 251,539 (20.3%) ended in non-live birth (2.1% stillbirth, 84.6% abortion, 8.2% ectopic, and 5.1% molar). Of these non-live births, 3,204 (1.3%) had an MCM code within ± 30 days of the non-live birth. This reduces to 2,395 (1.0%) when restricted to MCM codes that were not also present > 30 days from the non-live birth. The most common MCM among this 2,395 was congenital malformations of uterus and cervix (n = 1,141; 47.6%); it was also the most common MCM code on a claim outside of the ± 30 days surrounding non-live births (68.2%).
Conclusions: Amongst pregnancies that ended in non-live births, only 1.0% had a code for an MCM within ± 30 days of the non-live birth that was not present on a maternal claim outside of this time frame. Since MCMs occur in approximately 2 to 4% of livebirths and are expected to be even more common among non-live births, this suggests MCMs among non-live births are not well-captured in claims data. The most common MCM code on claims surrounding non-live births was for congenital malformations of the uterus and cervix, which is a risk factor for non-live births, suggesting some of the codes represent maternal malformations. The code for congenital malformations of the uterus and cervix was also the most common MCM code on claims outside of the ± 30 days surrounding non-live births, further suggesting this is a common maternal MCM code among pregnancies that end in non-live births.