Previous studies have demonstrated that up to 30 percent of systemic lupus erythematosus (SLE) pregnancies result in preterm birth and more than half have complications. To further test these findings and to assess the impact of race on adverse pregnancy outcomes, researchers at three tertiary care centers in the Southeastern U.S. assembled a multi-center, EHR-based cohort of SLE births.
“There hasn’t really been prior work to assemble a group across multiple centers,” said April Barnado, M.D., assistant professor and research director at the Vanderbilt Lupus Center. “Most studies are traditional, single-center cohort studies that include patients who are typically doing better; there is some bias there. We wanted to take advantage of EHRs that have rich longitudinal data to capture more real-world data.”
“Too little is known about how to improve pregnancy outcomes in women with lupus. We hope our work can be the foundation to build a large cohort of SLE patients with pregnancy data that can help identify the determinants of poor outcomes and ultimately prevent them,” added Ashley Blaske, M.D., rheumatology fellow at Vanderbilt University Medical Center.
Barnado, Blaske and colleagues from Duke University and the Medical University of South Carolina (MUSC) presented their preliminary findings at the American College of Rheumatology annual meeting in November.
“Too little is known about how to improve pregnancy outcomes in women with lupus.”
Digging into the EHR
The collaborators examined SLE births from 2013 to 2018 using an algorithm developed by the Vanderbilt team for analyzing the EHR systems. “The methodology was a key component of the study,” Barnado said.
The algorithm was developed using the Synthetic Derivative, a deidentified database containing clinical information from Vanderbilt’s EHR. The newly developed algorithm was then validated at the Duke and MUSC sites.
All three centers selected subjects with at least four or more counts of the SLE ICD-9 or ICD-10 codes and at least one ICD-9 or ICD-10 code for delivery-related diagnoses. A subject was defined as a case if diagnosed with SLE by a rheumatologist, nephrologist or dermatologist.
Across the three centers, there were 131 women with SLE with 146 pregnancies. The mean age at delivery was 29 years. The researchers observed higher than previously reported rates of preeclampsia (24 percent), preterm birth (42 percent), and Caesarean section (54 percent) across all three centers; 79 percent of preeclampsia occurred preterm.
The demographics were different between centers, with more African American women at Duke and MUSC (71 percent and 66 percent respectively) compared to Vanderbilt (38 percent); there were few Hispanic women.
“We expected to see higher rates of preeclampsia and preterm births in African American compared to white women,” Barnado said. “Race, however, was not associated with adverse outcomes, which may reflect referral bias of the sickest SLE patients to delivery at tertiary care centers.”
Barnado says the results suggest there is still an unmet need for guidelines and management strategies to reduce adverse outcomes – even at sophisticated tertiary care centers.
“We’re still trying to develop tools to help providers have meaningful conversations with patients about contraception and pregnancy in lupus,” she said. “We don’t always do the best job as physicians in meeting patients where they are and helping them understand where they need to be.”
“We’re still trying to develop tools to help providers have meaningful conversations with patients about contraception and pregnancy in lupus.”
The researchers hope to use the algorithms developed for this study to enlarge their lupus pregnancy cohort. “The methodology will be deployed at other institutions outside of the Southeast,” Blaske said. “A larger cohort will give us the power to look at the determinants of these outcomes – patient factors, disease factors, provider factors – to help better inform monitoring and medication management.”