Systematic review of risk prediction models for the progression of diabetic nephropathy in type 2 diabetes mellitus
10.3760/cma.j.cn115682-20231229-02869
- VernacularTitle:2型糖尿病肾病进展风险预测模型的系统评价
- Author:
Chengcheng LI
1
;
Xin SUN
;
Shiye ZENG
;
Xin DUAN
;
Rong XU
;
Jin HUANG
Author Information
1. 中南大学湘雅二医院临床护理学教研室，长沙 410011
- Keywords:
Diabetes mellitus, type 2;
Diabetic nephropathy;
Risk prediction;
Prognostic model;
Systematic review
- From:
Chinese Journal of Modern Nursing
2024;30(30):4119-4127
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To systematically evaluate the risk of bias and applicability of risk prediction models for the progression of diabetic nephropathy (DN) .Methods:A systematic search was conducted in CNKI, CBMdisc, Wanfang, VIP, PubMed, Web of Science, Embase, and CINAHL for literature related to DN progression prediction models, with a search timeline up to April 30, 2023. Two researchers independently screened the literature and extracted data according to a checklist for key assessments of prediction model studies and the PROBAST tool for assessing risk of bias in prediction models.Results:A total of nine articles encompassing 15 models were included. Of these, eight studies were retrospective study, and one was a randomized controlled trial. The area under the receiver operating characteristic curve ( AUC) for these models ranged from 0.626 to 0.986. Three studies conducted external validation, and seven studies conducted internal validation. Commonly repeated predictive factors included eGFR, cystatin C, and glycated hemoglobin (HbA1c). While the overall applicability of the models was good, methodological issues such as inappropriate data acquisition, selection of predictive factors, and neglect of model performance evaluation contributed to a certain risk of bias. Conclusions:The current DN progression risk prediction models demonstrate good discrimination and applicability. However, most models lack comprehensive calibration assessments and exhibit methodological flaws. Future research should focus on developing models with better applicability and lower bias, coupled with effective internal and external validation.
