Establishment And Verification Of Risk Prediction Model Of Peristomal Moisture-associated Skin Damage In Patients With Intestinal Stoma

OBJECTIVETo investigate the incidence and risk factors of peristomal moisture-associated skin damage in patients with intestinal stoma,and construct nomogram and application to predict the occurrence in early postoperative period.To verificate the prediction model internally and externally.To evaluate the clinical usefulness of the prediction model.METHODSA total of 305 patients in the wound stoma clinic of a third-grade a hospital were selected as the research subjects in Guangzhou from October 2019 to November 2020,and 292 patients were eventually included as the model training group.Use the questionnaire to obtain the data of patients.The standardized behavior of patients using stomy products was obtained by field observation of stomy nurses.Logistic regression analysis to explore independent risk factors for peristomal moisture-associated skin damage and establish a risk prediction model.Verificate the peristomal moisture-associated skin damage prediction model internally via the Bootstrap self-sampling method.After that,verificate the peristomal moisture-associated skin damage prediction model externally.A total of 94 patients from two other third-grade a hospitals in Guangzhou were selected for external validation.C-statistics were used to evaluate the prediction performance of the model.Calibration of the model was evaluated by calibration curve diagram and Hosmer-Lemeshow test.Decision Curve Analysis(DCA)was used to evaluate the clinical usefulness of the prediction model.RESULTSIn this study,the incidence of moisture-associated skin damage around enterostomy early after surgery(<6 months)in colorectal cancer patients was 31.2%.Multivariate logistic regression analysis showed that the height of stoma opening<1.4cm,skin folds around the stoma,receiving radiotherapy and chemotherapy,and irregular replacement of stoma products are independent risk factors for peristomal moisture-associated skin damage.Accept online continuous care is an independent protective factor.The C-statistics of the model constructed based on the above 5 factors is 0.910,means the prediction performance of the model were good.The Hosmer-Lemeshow test value is 8.644(P=0.195>0.05),means the calibration of the model were good.Sensitivity and specificity is 85.7%and 85.6%.The Youden index is 0.713 at the cut-off point of 0.303.The corrected C-statistic obtained by Bootstrap method was 0.900(95%CI:0.886～0.915),and Hosmer-Lemeshow test P=0.789>0.05.External test results showed that the C-statistic of the histograph based on the data of patients in the validation group was 0.821(95%CI:0.756～0.887),and Hosmer-Lemeshow test P=0.236>0.05,indicating that the model had good differentiation and calibration.The DCA decision curve of the risk prediction model shows that the curve(blue line)of the model is not closed to the two extreme case lines(black line and gray line),and the overall curve is close to the upper right corner,indicating that the risk prediction model has good clinical usefulness.CONCLUSIONThe risk prediction model for peristomal moisture-associated skin damage in patients with colorectal cancer in the early postoperative period(<6 months):logitP=-4.802+3.865 skin wrinkles around the stomy+2.995 stomy opening height less than 1.4 cm+7.281 receiving radiotherapy or chemotherapy+6.051 improper behavior in the use of stomy products+7.978 non-network continuous care model.And the prediction performance,calibration,and clinical usefulness of the risk prediction model were good.