重症监护病房患者压力性损伤严重程度预测模型的构建及验证OA

Objective To analyze the independent risk factors influencing the severity of pressure injury(PI)in patients in the department of intensive care unit(ICU),and to construct and validate a nomogram model for predicting the risk of irreversible PI.Methods A retrospective study design was adopted.A total of 143 ICU patients with PI admitted to Heyuan People's Hospital(Guangdong Provincial People's Hospital Heyuan Hospital)from 2019 to 2023 were included as the modeling set,and 73 patients from Guangdong Provincial People's Hospital Affiliated to Southern Medical University(Guangdong Academy of Medical Sciences)were selected as the external validation set.According to the National Pressure Injury Advisory Panel(NPIAP)guidelines,stage 1 and stage 2 PI were defined as reversible PI,while stage 3 and above were defined as irreversible PI.Demographic characteristics,disease status,and clinical indicators were collected.Independent predictors that affect the develpment of ICU patients'PI into irreversible PI were screened by univariate and multivariate Logistic regression analyses,based on which a nomogram prediction model was constructed.The discrimination,goodness-of-fit,and clinical utility of the model were evaluated by receiver operator characteristic curve(ROC curve),calibration curve,and decision curve analysis(DCA).Internal validation was performed via Bootstrap resampling,and external validation was conducted using external data.Results Multivariate Logistic regression analysis showed that length of ICU stay[odds ratio(OR)=1.096,P=0.002],mechanical ventilation use(OR=3.635,P=0.046),urinary incontinence(OR=32.688,P<0.001),and have a history diabetes mellitus(OR=3.746,P=0.011)were independent risk factors for the progression of PI to irreversible injury in ICU patients.The nomogram model constructed based on the above factors performed excellently in the modeling set,with area under the curve(AUC)of 0.935[95%confidence interval(95%CI)was 0.892-0.978],sensitivity of 87.4%,and specificity of 87.3%,showing good calibration and clinical net benefit,the internal validation AUC was 0.833.External validation results showed an AUC of 0.880(95%CI was 0.875-0.974),sensitivity of 80.8%,and specificity of 81.6%,confirming the good generalization ability of the model.Conclusions Length of ICU stay,mechanical ventilator use,urinary incontinence,and have a history diabetes mellitus are independent risk factors in predicting the severity of PI in ICU patients.The nomogram model is simple,intuitive,and has good predictive performance.It can help clinical staff implement stratified management and targeted interventions early,thereby improving patient prognosis.