Establishment of risk prediction model for medical adhesive-related skin injuries in ICU patients and its validation
Objective To establish a prediction model of the risk of medical adhesive-related skin injury(MARSI)in ICU patients,and verify it.Methods A total of 261 patients admitted to a top-three hospital in Guizhou province from March to December 2022 were selected for data monitoring,and logistic regression analysis was used to screen relevant factors to establish a nomogram prediction model for MARSI,and receiver operating characteristic(ROC)curve,calibration curve,and decision curve were used to test the fitting effect of the prediction model.Results The results of logistic regression analysis revealed that age,mechanical ventilation,APACHE Ⅱ score,history of aller-gy,fever,hypoproteinemia,skin edema,and use of sedatives were the factors influencing the development of MAR-SI.Using these 8 influencing factors identified,a well-fitting nomogram model for MARSI risk prediction was estab-lished,with the area under the ROC curve of 0.85(95％CI was 0.824-0.886)and the C-index of 0.85,and there was no significant difference between the model-predicted data and actual data by model fit degree detection(Hosmer-Lemeshow test;x2=4.621,P=0.797)and the model had a good prediction effect.Conclusion The nomogram pre-diction model established in this study is of good prediction effect,which can provide a reference for clinical nurses to predict the risk of MARSI in ICU patients and formulate personalized prevention strategies as soon as possible.
intensive caremedical adhesive-related skin injurynomographrediction model
万方数据
维普
