SLE-Derived Algorithms May Be Inconsistent When Assessing Adverse Pregnancy Outcomes Among Racial Groups

Systemic lupus erythematous (SLE)-rooted diagnostic algorithms yield inconsistent results when accounting for adverse pregnancy outcomes (APOs) among different racial groups, according to study results published in Rheumatology.

Over the last few decades, incident APOs have occurred among women with SLE at a higher rate than among those without, particularly Black women.  As large, diverse datasets of pregnancy outcomes in women with SLE remain understudied, researchers aimed to identify disparities among racial groups.

The PLiCC Pregnancies in Lupus in the Carolinas Collaborative (PLiCC) cohort was created using electronic medical record (EMR)-based datasets within 3 sites of the Carolinas Collaborate. Race, as well as APOs, were identified within the cohort. Women diagnosed with SLE with singleton pregnancy between 2014 and 2019 were included in the study.

Pregnancy outcomes were stratified into 4 cohorts based on 3 separate diagnostic algorithms for SLE, as well as one group with a confirmed diagnosis: a group with the highest sensitivity (sensitive cohort), a group with the highest positive predictive value (predictive cohort), a group which corrected errors from the summation of multiple models (XG Boost cohort), and a group with a confirmed SLE diagnosis through chart review (confirmed cohort).

A total of 172 women were included in the study, with 53% of pregnancies occurring in Black women, 36% in White women, 5% in Asian women, and 3% in Hispanic women.

APOs were identified in 40% to 52% of patients overall, with severe APOs occurring in 22% to 30% of pregnancies. Preeclampsia was identified in 18% to 29% of pregnancies.

Researchers found the algorithms used to identify women with SLE underestimated APOs in White women compared with Black women, and overestimated the disparities between the two racial groups.

APOs and severe APOs were significantly underestimated in White Women, with estimates up to 40% to 75% lower in the algorithm cohorts compared with the confirmed cohort. Twenty-nine percent of White women with confirmed SLE had preeclampsia, compared with only 11% of the algorithm-based cohorts.

Comparably, the algorithms more accurately estimated APO frequencies among Black women, with estimates being 12% to 20% lower when compared with the confirmed cohort.

Due to the overestimation of SLE diagnosis in White women, it was suggested that Black women experience APOs at rates 2.5 to 6 times higher. However, when confirmed pregnancies were adjusted by maternal age, no statistical differences in outcomes by race were observed.

This study was limited by potential referral bias, as all deliveries took place in academic medical centers.  Additionally, generalizability of study results was limited due to similar patient populations across all study sites.

“In conclusion, the PLiCC study revealed high rates of adverse pregnancy outcomes for both White and Black women with confirmed SLE delivered at major academic medical

centres in the Southeastern USA, demonstrating a specific population with an enduring need for ongoing research and improved care,” the study authors stated.

Disclosure: One of the study authors declared affiliations with biotech, pharmaceutical, and/or device companies. Please see the original reference for a full list of authors’ disclosures.