Presentation (CTI)

OBJECTIVES: For studies requiring time-sensitive data (e.g., seasonality, new product launch) and known pregnancy status at index, existing pregnancy identification algorithms may not be suitable as enough time may not have elapsed for an end-of-pregnancy outcome to occur. This research explored the use of codes indicative of pregnancy using weeks gestational age in a real-world claims database.
METHODS: Using HealthVerity’s Inovalon Closed Claims Database, we identified females aged 12-55 years during the study period of January 1 to December 31, 2023. International Classification of Diseases, 10th revision Clinical Modification (ICD-10-CM) codes denoting weeks or trimester of gestation were enumerated.
RESULTS: Among 7.8 million women of reproductive age in 2023, 435,971 (6%) had at least one ICD-10-CM code indicating pregnancy. The most frequent codes observed were for 39, 36, 37, 38, and <8 weeks gestation. Of 393,580 females with at least one Z3A code, 102,190 (26%) had at least one diagnosis code that indicated a conflicting temporal relationship for weeks of gestation. Discrepancies, such as a claim for 8 weeks gestation occurring after a claim for 12 weeks gestation, occurred most often prior to 15 weeks (9-16% of persons, average length 3 weeks) and then again after 25 weeks (10-61%, 8 weeks). Use of trimester codes without a corresponding Z3A code were infrequent (10%). Among 350,218 (80%) with both a Z3A and trimester code on the same day, Z3A and trimester codes were conflicted in 99,625 (28%) of pregnancies.
CONCLUSIONS: The majority of pregnancies in 2023 in this US-based claims dataset were enumerated with ICD-10-CM Z3A.XX codes rather than trimester-specific indicators. Disagreements within a pregnancy on weeks of gestation were common, and most extreme towards the end of pregnancy. Further work is needed to develop and validate an algorithm to identify pregnancies that does not rely on end of pregnancy events.