Performance of the Systemic Lupus Erythematosus Risk Probability Index (SLERPI) in a cohort of Colombian population

Objective: To evaluate the performance of the Systemic Lupus Erythematosus Risk Probability Index (SLERPI) in Colombian patients with systemic lupus erythematosus (SLE). Methods: The Colombian cohort included 435 SLE patients and 430 controls with other autoimmune diseases (ADs). Clinical and serological data were collected, and SLE was indicated by SLERPI scores > 7. The American College of Rheumatology (ACR)-1997, Systemic Lupus International Collaborating Clinics (SLICC)-2012, and European League Against Rheumatism (EULAR)/ACR-2019 criteria were used as reference standards. The impact of overt polyautoimmunity (PolyA) on SLERPI performance was assessed. Additionally, multivariate lineal regression analysis was performed to evaluate the contribution of SLERPI features to the overall SLERPI score. Results: SLE patients had higher SLERPI scores (P < 0.0001), with almost 90% meeting "definite" lupus criteria. Main factors influencing SLERPI included immunological disorder (β:44.75, P < 0.0001), malar/maculopapular rash (β:18.43, P < 0.0001), and anti-nuclear antibody positivity (β:15.65, P < 0.0001). In contrast, subacute cutaneous lupus erythematosus/discoid lupus erythematosus (β:2.40, P > 0.05) and interstitial lung disease (β:-21.58, P > 0.05) were not significant factors to the overall SLERPI score. SLERPI demonstrated high sensitivity for SLE, both for the overall SLE group and for those without overt PolyA (95.4% and 94.6%, respectively), but had relatively low specificity (92.8% and 93.7%, respectively). The model showed high sensitivity for hematological lupus (98.8%) and lupus nephritis (96.0%), but low sensitivity for neuropsychiatric lupus (93.2%). Compared to the ACR-1997, SLICC-2012 and EULAR/ACR-2019 criteria, SLERPI yielded the highest sensitivity and lowest specificity. Conclusion: SLERPI efficiently identified SLE patients in a Colombian cohort, showing high sensitivity but low specificity. The model effectively distinguishes SLE patients, even in the presence of concurrent overt PolyA. (Table presented.)