基于决策树法构建肠造口周围潮湿相关性皮肤损伤的风险预测模型

打开文本图片集

Construction of a risk prediction model for peristomal moisture-associated skin damage based on adecisiontreemethod

ZHANGDongju，GUOSiqin，WANGWei*

Suqian Hospital，NanjingDrum Tower Hospital Group， 2238oo China

*Corresponding Author WANG Wei，E-mail： 247731799@ qq.com

AbstractObjective：Toconstructariskpredictionmodelforperistomalmoistureassociated skindamage（PMASD）basedonadecision treemethod.Methods：ClinicaldatafrompatientswhounderwentcolostomysurgeryatourhospitalbetweenJune2O22andJune2024 werecolleted.UnivariateandmultivariateLogisticregressonwereusedtoanalyetheriskfactorsforPMASDinpatientswith colostomy.Basedontheidentifiedriskfactors，adecisiontreemodelwasconstructedtopredictheriskofPMASDinpatientswith colostomy.Thepredictivevalueof theriskpredictionmodelsconstructedusingLogisticregressionanddecisiontreealgorithsfor PMASD wascompared.Results：Among 220patientswhounderwentcolostomysurgery，75developedPMASD，withanincidencrate of 34.09% .Multivariate Logistic regresionanalysis showed thatstoma site，stoma height，stoma leakage，exposure toradiotherapy/ chemotherapy，and stoma self-care ability were influencing factors for PMASD in patients with colostomies ΔP<0.05 ）.A decision tree modelwasconstructedbasedontheseriskfactors.Themodelconsistedoffourlayersand1lnodes，withstomaleakage，exposureto radiotherapy/mothaposelfaeabiltoht，dolocasafcatiooestolageeot important predictor.Receiveroperating characteristic（ROC）curves showed thattheareaunder theROCcurve（AUC）forpredicting PMASD inpatients withcolostomies was O.847forthedecisiontree model，whiletheAUCforpredicting PMASDinpatients with colostomieswasO.844fortheLogisticregressionmodel.Therewasnostatisticallysignificantdiferencebetweenthetwomodel P> 0.05）. Conclusions： Ileostomy，stoma height <1.4cm ，stoma leakage，exposure to radiotherapy/chemotherapy，and poor stoma self-care arerisk factors forPMASDinpatients withcolostomies.Theconstructeddecisiontreemodelhasahigh predictive value.

Keywordscolostomy;peristomalmostureassociatedskindamage;decisiontree;Logisticregresion;predictionmodel;influencing factors

摘要目的：基于决策树法构建肠造口周围潮湿相关性皮肤损伤（PMASD）的风险预测模型。（剩余7761字）