Abstract
OBJECTIVES: To evaluate the performance of Systemic Lupus Erythematosus Risk Probability Index (SLERPI) in patients with SLE using a Chinese cohort. METHODS: The Chinese cohort included 352 patients with and 385 without SLE (control group). The clinical data of patients, including demographic data, clinical findings and serological profiles, were collected. Patients with an SLERPI score >7 were classified as SLE. The performance of 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 references. RESULTS: Of these four classification criteria, SLERPI has the highest sensitivity (98.3% (95% CI 96.3% to 99.4%)), but lowest specificity (89.4% (95% CI 85.8% to 92.2%)). In the control group, patients eligible for the classification criteria for SLE were mainly those with primary Sjogren's syndrome (pSS) and undifferentiated connective tissue disease (UCTD), which adversely affected the specificity of the classification criteria. Moreover, significantly more patients with pSS and UCTD met SLERPI than those who met other classification criteria. After excluding patients with pSS and UCTD from the control group, the specificity and accuracy of SLERPI improved to 94.3% (95% CI 91.0% to 96.6%) and 96.5% (95% CI 95.0% to 97.9%), respectively, and both outperformed the EULAR/ACR-2019 criteria. The time to SLERPI classification was the same as their clinical time to diagnosis in 261 patients, earlier than the clinical diagnosis in 23 patients and later than the clinical diagnosis in 9 patients. A total of 280 patients had the same time to SLERPI classification as EULAR/ACR-2019, 8 patients had earlier than EULAR/ACR-2019 and 1 patient had later than EULAR/ACR-2019. CONCLUSION: SLERPI performed well in patients with SLE, particularly for the earlier diagnosis of SLE.