Researchers in the Johns Hopkins Department of Gynecology and Obstetrics are leveraging the power of big data to empower studies that can lead to better population health outcomes and individualized patient care.

The Johns Hopkins Precision Medicine Analytics Platform (PMAP) is a data repository that brings together Johns Hopkins electronic medical records with various other large clinical databases, allowing physicians and researchers to use data analytics, biostatistical methods and machine learning to push research forward and improve patient care.

“It brings together all that information so that study teams can see the whole picture of the patient they’re looking at,” says Marie Bielman, a senior software engineer with the Gyn/Ob department.

Using PMAP, researchers in the department have published retrospective studies on bacterial species responsible for recurrent urinary tract infections (UTIs); maternal, obstetric and neonatal factors associated with delayed cord clamping; surgical site infection severity in patients receiving non-cephalosporin antibiotics at the time of a cesarean section; and clinical outcomes in a high-risk cohort receiving aspirin prophylaxis for the prevention of preeclampsia.

Researchers were able to analyze large amounts of data that would be resource-intensive to acquire through their own trials, allowing them to explore the intricacies of Gyn/Ob issues. As part of Johns Hopkins’ inHealth Precision Medicine initiative, PMAP helps clinicians understand patient subgroups at a granular level, which opens up the opportunity to develop more precise care for patients. The department has been using the platform since 2022.

The study on recurrent UTIs, for example, drew on data from 181 women who experienced 496 UTIs. They were all seen at a Johns Hopkins gynecology practice between March 1, 2019, and March 1, 2024. Using PMAP, Victoria Handa, director of the Division of Urogynecology and Reconstructive Surgery, and her coauthors found that consistent bacterial species cause most sequential infections among women with recurrent UTIs, which supports the hypothesis that bacterial reservoirs may contribute to recurrent UTIs.

“We’re learning so much through retrospective data studies,” Bielman says. “The quality of the data is really amazing.”

Bielman maintains two registries for the department: an obstetrics registry created in 2022 capturing data from all Johns Hopkins maternal records, with data from mothers and babies, and a gynecology registry created in 2023 that includes all patients who have seen gynecologists across all Johns Hopkins locations. “Between those two parent registries, we have a complete picture of all Gyn/Ob at all Hopkins sites,” Bielman says.

PMAP can be accessed within Johns Hopkins’ Secure Analytics Framework Environment, a secure virtual desktop and storage network. Patient data is deidentified within PMAP registries, and protected health information is only accessible by certain study team members with IRB approval.

For each Gyn/Ob study using PMAP, Bielman creates subsets of patient cohorts tailored to researchers’ needs. She’s able to save these subsets, making them easy to reuse and access for future research projects.

She notes that the obstetrics registry was created in conjunction with maternal-fetal medicine specialist Jason Vaught, medical director of labor and delivery at The Johns Hopkins Hospital. For Vaught’s initial hypertension and pregnancy studies, Bielman created over 200 standard pregnancy and delivery variables in PMAP that can now be run automatically for any future obstetrics studies.

Currently, PMAP is aiding research in divisions throughout Gyn/Ob, including fetal therapy, gynecologic oncology, maternal-fetal medicine, urogynecology, and reproductive endocrinology and infertility.

Newer studies are using artificial intelligence to look at unstructured data — such as clinical notes and pathology reports — to utilize even more data that would be time-intensive to process and analyze manually.

“PMAP allows us to query data from the entire Johns Hopkins system to look at outcomes and improve our diagnostic and therapeutic options,” says Andy Satin, director of the Department of Gynecology and Obstetrics. “It significantly boosts our research, which will ultimately lead to better, more individualized patient care.”