Cao Yi, Zeng Xi, Gou Yangyang, Lu Yu, Zhu Dian, Wang Hui, Dai Yan, Tian Jie, Jian Liu, Min Peng
Department of Neurosurgery, The Affiliated Hospital of Guizhou Medical University, China; School of Nursing, Guizhou Medical University, China.
Department of Neurosurgery, The Affiliated Hospital of Guizhou Medical University, China.
Comput Biol Med. 2025 Sep;195:110612. doi: 10.1016/j.compbiomed.2025.110612. Epub 2025 Jun 25.
The heterogeneity of machine learning (ML) models predicting the risk of stroke-associated pneumonia (SAP) is considerable. This study aims to conduct a meta-analysis and comparison of published ML models that predict SAP risk.
A systematic search was conducted across eight databases, covering the period from their inception to August 16, 2024. Data extraction was performed based on the Critical Appraisal and Data Extraction for Systematic Reviews of Prediction Modelling Studies (CHARMS) framework. The Assess the Risk of Bias and Applicability of Prediction Model (PROBAST) tool was used to evaluate the risk of bias and applicability of the included models. Descriptive analysis was performed on the included studies, and Meta-Disc 1.4 and Stata 14.0 software were used for sensitivity analysis, subgroup analysis, and meta-regression.
A total of 18 studies comprising 46 SAP risk prediction models were included. The overall Area Under the Curve (AUC) was 0.8623, with a pooled sensitivity of 0.77 (95 % CI: 0.76-0.77, P < 0.001, I = 94.7 %) and a pooled specificity of 0.75 (95 % CI: 0.74-0.75, P < 0.001, I = 99.1 %). Logistic regression (LR) was the most commonly used ML method for SAP prediction, with an AUC of 0.8684, sensitivity of 0.77 (95 % CI: 0.75-0.78, P < 0.001, I = 94.7 %), and specificity of 0.74 (95 % CI: 0.73-0.74, P < 0.001, I = 98.6 %). In contrast, non-LR models had an AUC of 0.8591, sensitivity of 0.77 (95 % CI: 0.76-0.78, P < 0.001, I = 94.9 %), and specificity of 0.75 (95 % CI: 0.75-0.75, P < 0.001, I = 99.3 %). Sensitivity analysis indicated that the random-effects meta-analysis yielded an AUC of 0.8476, sensitivity of 0.77 (95 % CI: 0.76-0.78, P < 0.001, I = 93.5 %), and specificity of 0.72 (95 % CI: 0.72-0.72, P < 0.001, I = 98.1 %). Meta-regression analysis revealed that country/region, ML algorithms, participants, year, study source, and study design were not sources of heterogeneity (P = 0.183).
In the existing SAP prediction models, the LR model demonstrates relatively better prediction performance due to its good interpretability and adaptability to smaller sample sizes. However, there are significant limitations in the current research: the overall bias risk of the models is relatively high, the variable handling methods are inconsistent, and there is a scarcity of prediction studies for patients with hemorrhagic stroke. Moreover, the models generally lack external validation, which limits their clinical generalization ability. Future research should conduct prospective, multi-center data studies and carry out internal and external validations to enhance reliability. Strictly following the requirements of CHARMS and PROBAST will effectively reduce the bias risk, enhance the validation efficacy of the models and the clinical translation value.
预测卒中相关性肺炎（SAP）风险的机器学习（ML）模型存在相当大的异质性。本研究旨在对已发表的预测SAP风险的ML模型进行荟萃分析和比较。
对八个数据库进行了系统检索，涵盖从数据库创建至2024年8月16日的时间段。基于预测模型研究系统评价的关键评估和数据提取（CHARMS）框架进行数据提取。使用预测模型的偏倚风险和适用性评估（PROBAST）工具评估纳入模型的偏倚风险和适用性。对纳入研究进行描述性分析，并使用Meta-Disc 1.4和Stata 14.0软件进行敏感性分析、亚组分析和Meta回归。
共纳入18项研究，包含46个SAP风险预测模型。曲线下面积（AUC）总体为0.8623，合并敏感度为0.77（95%CI：0.76 - 0.77，P < 0.001，I = 94.7%），合并特异度为0.75（95%CI：0.74 - 0.75，P < 0.001，I = 99.1%）。逻辑回归（LR）是预测SAP最常用的ML方法，AUC为0.8684，敏感度为0.77（95%CI：0.75 - 0.78，P < 0.001，I = 94.7%），特异度为0.74（95%CI：0.73 - 0.74，P < 0.001，I = 98.6%）。相比之下，非LR模型的AUC为0.8591，敏感度为0.77（95%CI：0.76 - 0.78，P < 0.001，I = 94.9%），特异度为0.75（95%CI：0.75 - 0.75，P < 0.001，I = 99.3%）。敏感性分析表明，随机效应荟萃分析得出的AUC为0.8476，敏感度为0.77（95%CI：0.76 - 0.78，P < 0.001，I = 93.5%），特异度为0.72（95%CI：0.72 - 0.72，P < 0.001，I = 98.1%）。Meta回归分析显示，国家/地区、ML算法、参与者、年份、研究来源和研究设计不是异质性来源（P = 0.183）。
在现有的SAP预测模型中，LR模型因其良好的可解释性和对较小样本量的适应性而表现出相对较好的预测性能。然而，当前研究存在显著局限性：模型的总体偏倚风险相对较高，变量处理方法不一致，且针对出血性卒中患者的预测研究稀缺。此外，模型普遍缺乏外部验证，这限制了它们的临床推广能力。未来研究应开展前瞻性、多中心数据研究，并进行内部和外部验证以提高可靠性。严格遵循CHARMS和PROBAST的要求将有效降低偏倚风险，提高模型的验证效力和临床转化价值。
