Development of a claims-based flare algorithm for systemic lupus erythematosus.

Area of Science:

• Rheumatology
• Health Informatics
• Data Science

Background:

• Systemic lupus erythematosus (SLE) is a chronic autoimmune disease requiring continuous monitoring.
• Accurate identification of SLE flares is crucial for timely intervention and disease management.
• Existing methods for flare identification can be resource-intensive and may not fully leverage available data.

Purpose of the Study:

• To develop and validate a claims-based algorithm for identifying SLE flares.
• To utilize linked administrative claims and electronic medical record (EMR) data for enhanced accuracy.
• To create a more streamlined approach for SLE flare detection in large patient populations.

Main Methods:

• Retrospective analysis of linked claims and EMR data from 2003-2019.
• Inclusion of adult SLE patients with continuous enrollment and clinical activity.
• Development of a proxy SLEDAI-2K score from EMR data and logistic regression modeling to predict flares based on claims data.

Main Results:

• The final algorithm demonstrated good performance with a c-statistic of 0.76 and a Brier score of 0.07.
• Key predictors for SLE flares included inpatient admissions, outpatient visits, MRI, ER visits, and rheumatology visit frequency.
• The model identified significant associations between specific claims data and SLE flares (p < .01).

Conclusions:

• The developed claims-based algorithm shows promise for identifying SLE flares.
• This approach offers a streamlined and alternative method compared to traditional flare assessment.
• The algorithm can advance SLE research by facilitating the study of disease flares using administrative claims data.