A Scoping Review of Prediction Models for Renal Injury Risk in Children with Henoch-Sch?nlein Purpura postprint

Abstract: Background　Henoch-Schönlein purpura (HSP) is a common systemic vasculitis in childhood, with a relatively high incidence of renal involvement. In recent years, an increasing number of studies have focused on risk prediction models for HSP-related renal injury; however, substantial heterogeneity exists in predictor selection and modeling approaches. Objective　This study aims to systematically summarize the characteristics of existing predictive models for kidney injury risk in children with HSP, as well as their common predictors and methodological quality, thereby providing a basis for researchers to conduct related studies and for clinicians to identify high-risk children at an early stage. Methods　A scoping review was conducted. PubMed, Embase, the Cochrane Library, Web of Science, China National Knowledge Infrastructure (CNKI), Wanfang Data, VIP Database, and China Biology Medicine disc (CBM) were systematically searched from inception to June 1, 2025. Literature screening and data extraction were performed independently by two reviewers. Model-related information was extracted using the CHARMS checklist, and the risk of bias and applicability were assessed using the Prediction Model Risk of Bias Assessment Tool (PROBAST). Results　A total of 13 studies were included, comprising 6 Chinese-language and 7 Englishlanguage studies, with sample sizes ranging from 165 to 1 294 participants. The incidence of renal injury in children with HSP ranged from 26.67% to 63.75%, and the area under the receiver operating characteristic curve (AUC) ranged from 0.55 to 0.956. The most frequently reported predictors were age, recurrent purpura, persistent purpura, D-dimer, and serum albumin. PROBAST assessment indicated a high risk of bias across the included studies, mainly attributable to retrospective study designs, inadequate handling of missing data, and the lack of external validation. Conclusion　Existing prediction models for renal injury in children with HSP are at high risk of bias, which compromises their reliability and clinical implementation. Future studies should prioritize prospective study designs, standardized reporting, optimization of predictor selection, and multicenter external validation to develop more robust and generalizable risk prediction tools.