Systematic evaluation of risk prediction model for intensive care unit-acquired weakness
10.3760/cma.j.cn115682-20200325-02174
- VernacularTitle:ICU获得性衰弱风险预测模型的系统评价
- Author:
Yang LIU
1
;
Jian LUO
;
Lin XIE
;
Miao LIU
;
Xingting ZHOU
;
Yunhan DING
Author Information
1. 华中科技大学同济医学院附属协和医院护理部，武汉 430022
- Keywords:
ICU-acquired weakness;
Prediction model;
Systematic review;
Risk factors
- From:
Chinese Journal of Modern Nursing
2020;26(34):4769-4774
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To systematically evaluate the predictive model of intensive care unit (ICU) -acquired weakness so as to provide objective basis for clinical workers to choose appropriate predictive model and provide reference for future model update and new model development.Methods:Sixth databases including PubMed, Embase, web of science, The Cochrane Library, CNKI and Wanfang were searched by the computer. The retrieval time was the construction of the database to October 15, 2019, and the language was limited to Chinese and English. A total of two researchers conducted independent screening of literature, data extraction and evaluated the quality of the included literature in turn, and then used prediction model risk of bias assessment tool (PROBAST) to evaluate the quality of the models included in the literature.Results:A total of 8 articles with high quality were included. The area under the ROC curve of the five models were all greater or equal to 0.7. The model risk of bias assessment showed that only Witteveen's model was rated as low bias, and the remaining 7 models all had a higher risk of bias, but all models had good applicability.Conclusions:The predictive performance of ICU acquired weakness model is good, but there are some biases in development and report. In the future, the whole process of model development and verification should be reported in a standardized way to reduce methodological bias and provide high-quality evidence for clinical practice. Future studies should focus on external validation and updating of models to continuously improve model prediction performance and provide practical models for clinical practice.
