Construction and validation of a risk prediction model for pressure injury in emergency patients

LYU Honglei; YUE Chenqi; YAO Ying; GAO Ning; CHAI Qianwen; LU Minghui; WEI Li

doi:10.12102/j.issn.1009-6493.2025.02.003

Abstract：Objective:To construct a risk prediction model for pressure injury in emergency patients and verify the predictive effectiveness of the model.Methods:A total of 312 emergency patients admitted to a tertiary grade A hospital in Tianjin was selected as the research subjects by convenience sampling method from May 2022 to March 2023.The patients were divided into modeling group and validation group in ratio of 7∶3,with 218 patients in the modeling group and 94 patients in the validation group.A survey questionnaire on the influencing factors of pressure injury in emergency patients was formed through literature review,group discussions,and expert inquiries to investigate the patients.Based on the influencing factors,a predictive model was constructed,then the model was validated.Results:Among 312 emergency patients,28 patients developed pressure injuries,with the total incidence rate of 8.97%.The results of multiple factor analysis showed that serum albumin levels,incontinence status,perceptual ability,and mobility ability were the influencing factors for pressure injury in emergency patients(P<0.05).Based on the influencing factors,a prediction model was constructed.The Hosmer-Lemeshow fitting test results showed that χ2=4.648(P=0.590),the area under the curve(AUC)of receiver operator characteristic of the modeling group was 0.946[95%CI(0.914,0.978)],the sensitivity was 95.5%,and the specificity was 84.7%.The AUC of the validation group was 0.905[95%CI(0.728,1.000)],with a sensitivity of 83.3%and a specificity of 77.0%.Conclusions:The incidence of pressure injury in emergency patients is relatively high,and serum albumin levels,incontinence status,perceptual ability,and mobility ability have impact on it.The predictive model constructed has good predictive effectiveness,it is suitable for individualized risk prediction of pressure injury in emergency patients.