ObjectiveTo evaluate the existing prediction models for post-ischemic stroke depression risks. MethodsA comprehensive search was performed to identify studies on prediction models for post-ischemic stroke depression risks in the following databases, CINAHL, Embase, Medline, The Cochrane Library, Web of Science, PubMed, CNKI, VIP, China Biomedical Literature Database, and Wanfang Database, and the date of publication was set from the inception to July 13th, 2022. Two reviewers independently screened the studies, extracted data under the guidance of a checklist on Critical Appraisal and Data Extraction for Systematic Reviews of Prediction Modelling Studies, and analyzed the risk of bias and applicability of the included studies using Prediction Model Risk of Bias Assessment Tool. ResultsA total of 9 studies were enrolled, including 9 prediction models for post-ischemic stroke depression risks, and their areas under the receiver operating characteristic curves varied from 0.780 to 0.928. The predictors of the 9 prediction models mainly included National Institutes of Health Stroke Scale score, Barthel Index, age, and hypertension. The overall applicability of 5 studies was quite good, while the overall applicability of the other 4 studies was unclear; the risks of bias of the 9 studies were high. ConclusionsThe study on prediction models for post-ischemic stroke depression risks is still in the developing stage; multi-center research with large sample sizes should be carried out in the future to develop models with both accuracy and simplicity and optimize the models by external validation.

无