Abstract: Background：The 1997 American College of Rheumatology (ACR) criteria, the 2012 Systemic Lupus Erythematosus International Collaborating Clinics (SLICC) criteria, and the 2019 European League Against Rheumatism/American College of Rheumatology (EULAR/ACR) criteria are widely used for diagnosing systemic lupus erythematosus (SLE) in both adults and children. However, these criteria were initially developed based on adult cohorts, and their diagnostic value in pediatric SLE requires further validation. Objective：To evaluate the diagnostic accuracy of the three classification criteria in adult and pediatric SLE. Design：Meta-analysis. Methods：A literature search was conducted in PubMed, EMBASE, Cochrane Library, China National Knowledge Infrastructure (CNKI), Wanfang, and China Biomedical Database. Studies were selected according to predefined inclusion and exclusion criteria. Sensitivity, specificity, and the area under the curve (AUC) of the three classification criteria were compared. Heterogeneity sources were analyzed. GRADE approach was used to evaluate the certainty of evidence. Main outcome measures：Sensitivity and specificity of SLE classification criteria in pediatric cohorts. Results：Thirty-six studies were included, with 22 on adult SLE, 13 on pediatric SLE, and 1 on both. The 1997 ACR criteria were applied in 23 adult and 13 pediatric studies, the 2012 SLICC criteria in 22 adult and 13 pediatric studies, and the 2019 EULAR/ACR criteria in 17 adult and 10 pediatric studies. Except for sensitivity and AUC for the 2012 SLICC criteria, there were no statistically significant differences in diagnostic accuracy and AUCResults： between the three criteria in both adult and pediatric SLE. For pediatric SLE, the 1997 ACR criteria showed 5 true positives (4-5) and 1 false positive (1-2) per 100,000 people, while the 2012 SLICC and 2019 EULAR/ACR criteria showed 6 true positives (6-6) and 0 false positives (0-0) per 100,000 people. Based on the GRADE evaluation, the evidence body was rated as low quality by downgrading two levels in the meta-analysis for adult SLE using the three criteria while the GRADE result was moderate quality by downgrading 1 level in the meta-analysis for pediatric SLE. Conclusion：In both pediatric and adult SLE, the 2012 SLICC and 2019 EULAR/ACR criteria demonstrated superior diagnostic performance compared to the 1997 ACR criteria, with similar diagnostic accuracy. The misdiagnosis rate of the 2012-SLICC and 2019-EULAR/ACR classification criteria in pediatric SLE is zero.

Key words: Systemic lupus erythematosus, Classification criteria, Children, Diagnostic value, Sensitivity, Specificity, Meta-analysis