| Popis: |
Backgroud Diabetic foot is a common complication of diabetes patients, most of whom are seriously ill with rapid disease progression. A well-performing risk prediction model for the development of diabetic foot can help healthcare professionals to identify high-risk patients and take early interventions. Objective To systematically review the risk prediction models for diabetic foot, and provide reference for the construction and optimization of the model. Methods PubMed, Cochrane Library, Embase, Web of Science, CNKI and Wanfang Data were searched to collect the related studies on risk prediction models for diabetic foot from inception to May 15th, 2022. Two reviewers independently screened the literature, extracted data and evaluated the quality of models using prediction model risk of bias assessment tool (PROBAST). Meta-analysis of the predictors in the model was performed using Stata 17.0 software. Results A total of 13 papers were included, containing 13 models, 12 of which had AUC>0.7. Model calibration was performed on 7 models and 8 models were validated. PROBAST results showed that 1 of the 13 included papers was at low risk of bias and the remaining 12 were at high risk of bias; for model applicability, only 1 was of low applicability. The results of Meta-analysis showed that age (OR=1.13, 95%CI=1.04-1.24), glycated hemoglobin (OR=1.56, 95%CI=1.26-1.94), foot ulcer history (OR=5.93, 95%CI=2.85-12.37), previous amputation (OR=7.79, 95%CI=2.74-22.17), diminished sensitivity of the monofilament test (OR=1.59, 95%CI=1.42-1.78), foot fungal infection (OR=6.14, 95%CI=1.71-22.01), and kidney disease (OR=2.09, 95%CI= 1.65-2.65) were independent influencing factors for diabetic foot (P |
