Introduction: Systemic lupus erythematosus (SLE) is a complex autoimmune disease with heterogeneous behavior. Currently, there are no sufficient clinical tools for its diagnosis, there are only classifying criteria associated with the development of the disease, but no pathognomonic. Objectives: To evaluate the diagnostic and classifying performance of the systemic lupus erythematosus risk probability index (SLERPI) in a sample of the Colombian population. Methodology: 435 patients with SLE and 430 controls were included. SLERPI >7 indicated SLE, compared with ACR-1997, SLICC-2012 and EULAR/ACR-2019 criteria by estimating sensitivity (SN) and specificity (ES). The impact of polyautoimmunity (PolyA) on SLERPI scores and the contribution of its components was evaluated using multivariate regression. Results: Patients with SLE had higher SLERPI scores compared to the SLE control group (P < 0.0001), classifying close to 90% of patients with a definitive diagnosis. The main factors associated to SLERPI scores in the population included immunological disorder, malar/maculopapular rash, and antinuclear antibodies. Interstitial lung disease and subacute cutaneous/discoid lupus were non-significant (P > 0.0500). SLERPI had high SN in PolyA and non-PolyA (95.4% and 94.6%), with lower ES (92.8% and 93.7%) and comparable precision levels. Furthermore, it stood out in early SLE, nephritis and hematological SLE (SN >96%), but it had a lower performance in neuropsychiatric SLE (SN 92%). Conclusion: SLERPI identified efficiently SLE disease in a Colombian cohort with high SN but lower ES. It had excellent performance in detecting early SLE and in certain sub phenotypes but requires caution in neuropsychiatric SLE.
