Validating Claims-Based Algorithms Determining Pregnancy Outcomes and Gestational Age Using a Linked...

Area of Science:

• Pharmacovigilance
• Medical Informatics
• Reproductive Health

Background:

• Accurate pregnancy outcome identification and gestational age (GA) estimation are vital for drug safety studies in pregnant populations.
• Validated algorithms utilizing the International Classification of Diseases, Tenth Revision, Clinical Modification/Procedure Coding System (ICD-10-CM/PCS) for these purposes have been lacking.

Purpose of the Study:

• To develop and validate algorithms for classifying pregnancy outcomes and estimating GA using ICD-10-CM/PCS and service codes.
• To compare the performance of these claims-based algorithms against expert physician adjudication of electronic medical records (EMRs).

Main Methods:

• Algorithms were developed using 2016-2018 IBM MarketScan claims data.
• Obstetric services were grouped into episodes, and GA was determined using the most accurate clinical evidence.
• Physician adjudication of sampled EMRs (full-term live births, preterm live births, spontaneous abortions, stillbirths) was performed using standardized definitions.

Main Results:

• The claims-based algorithms identified 34,204 pregnancies, with 9.9% having available obstetric EMRs for adjudication.
• High percent agreement was observed for full-term live births (97.8%) and spontaneous abortions (100.0%).
• Agreements for preterm live births (62.4%) and stillbirths (70.8%) were also assessed, along with GA estimation accuracy within 7 and 28 days.

Conclusions:

• The developed pregnancy outcome algorithms demonstrate high agreement with physician adjudication of EMRs.
• These validated algorithms can be utilized for post-market maternal safety surveillance.
• The findings support the use of claims data for robust pregnancy outcome and GA assessment in pharmacoepidemiology.