Prediction models for central line associated bloodstream infections based on machine learning: a systematic review
10.3760/cma.j.cn115682-20230206-00378
- VernacularTitle:机器学习应用于中央导管相关血流感染预测模型的系统评价
- Author:
Kezhou YANG
1
;
Ning LIU
;
Ting WANG
;
Yang YANG
Author Information
1. 遵义医科大学珠海校区护理学系，珠海 519040
- Keywords:
Central line associated bloodstream infections;
Machine learning;
Prediction model;
Systematic review
- From:
Chinese Journal of Modern Nursing
2023;29(34):4677-4682
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To systematically review the central line associated bloodstream infection prediction models based on machine learning.Methods:The research on central line associated bloodstream infection prediction models based on machine learning was systematically searched in Cochrane Library, PubMed, Embase, WanFang Data, VIP, and China National Knowledge Infrastructure. The search period was from database establishment to March 18, 2022. Two researchers independently screened articles and extracted data. The data extraction was carried out using CHARMS checklist. The prediction model risk of bias assessment tool (PROBAST) was used to evaluate the bias risk of included studies.Results:A total of 9 related studies were ultimately included. The overall results indicated that modeling using machine learning algorithms had higher predictive accuracy compared to traditional Logistic regression. There was a risk of multiple biases in the modeling process, but the overall applicability evaluation of the article performed well.Conclusions:Existing evidence suggests that machine learning can accurately predict the occurrence of central line associated bloodstream infections, but there is a certain bias in development and reporting. In the future, emphasis should be placed on external validation and model updates, continuously improving its predictive performance.
