Background: Tradeoffs between registries and insurance claims for U.S. post-approval safety studies in pregnancy include smaller registry enrollment with deep clinical and patient-reported information versus larger sample sizes from claims data that can have missing/incomplete key variables and limited infant follow-up. Existing integrations with Health Information Exchanges (HIEs) may offer a solution that includes large numbers of pregnancies with rich EMR data across disparate medical providers to provide clinically meaningful medical history, pregnancy, and medication exposures with the relative efficiency of a retrospective database study.
Objectives: To assess key data available in pregnant women with systemic lupus erythematosus (SLE) obtained through Dorsata’s Ob/Gyn EMR network with additional linkage to non-Ob/Gyn EMRs.A
Methods: A retrospective cohort of pregnant women with SLE were identified using SNOMED terms within a clinically recognized pregnancy within the Dorsata EMR provider network in the U.S. [Jan 2018 – Feb 2022]. Data relevant to the pregnant woman, SLE condition, pregnancy, and infant follow-up were obtained. Dorsata enables linkage to non-Dorsata EMRs via integration (Zus Health), whereby a common patient record can be generated for technology companies and providers to use in the provision of patient care. Linkage proportions of the Dorsata EMR to non-Ob/Gyn EMRs and key clinical and demographic factors were quantified and summarized descriptively.
Results: During the analysis period, 380,109 pregnancies were identified in the Dorsata EMR; 703 (0.19%) had SLE. To evaluate linkage feasibility, a representative subsample of 83,915 pregnancies from one provider network was used, resulting in 166 pregnancies (0.2%) in 144 women with SLE. In this subsample, 115 (80%) women had Dorsata EMRs linkable to non-Ob/Gyn EMRs, while 13 infants ( < 10%) from live birth pregnancies had EMR linkages available. Dorsata’s EMR provided complete (100%) capture of key maternal variables such as age, race, ethnicity, pregnancy gestation, multiple pregnancy, ultrasound results, and gestational age at delivery; significant capture of pregnancy outcome (84%) and birthweight (76.9%); and partial capture of pre-pregnancy BMI (56.6%) and APGAR score (44.3%). Of the 115 SLE pregnancies with linkage to the mother’s non-Dorsata EMRs, 74 (64.3%) had data on SLE treatment, with 9 (12.2%) pregnancies indicating treatment from 90 days prior to LMP through end of pregnancy.
Conclusions: Provisioning of linked Dorsata Ob/Gyn EMRs to non-Ob/Gyn-EMRs enabled by providers under HIEs for improving patient care also provides a rich source of longitudinal clinical and demographic data that may be considered as a tool to assess post-approval product safety in pregnancy.
