Construction and validation of a model for predicting the risk of potentially inappro-priate medication in elderly patients with chronic diseases
Objective To construct a Nomogram prediction model for the risk of potentially inappropriate medication(PIM)in elderly patients with chronic diseases and provide references for clinical decision-making.Methods The clinical data of 358 elderly patients with chronic diseases in the First People's Hospital of Chuzhou from January to December 2023 were retrospectively analyzed,and they were randomly divided into the training set and the validation set according to the ratio of 7∶3.According to the presence of PIM,they were categorized into PIM group and non-PIM group.Risk factors for PIM in the elderly patients with chronic diseases were screened in the train-ing set using univariate and multivariate regression analysis,and the optimal Nomogram model was chosen using the AIC criteria and a nomogram was plotted.The performance of the model was evaluated by area under the curve(AUC),calibration curve and decision curve analysis(DCA).Results The AUC of receiver operator characteristic curve of the subjects in the training set and validation set was 0.786 and 0.768,respectively.The calibration plots of the training set and validation set showed that the Brier scores were 0.164 and 0.165,respectively.The results of the Hosmer-Lemeshow goodness-of-fit test were x2=4.405(P=0.883)andx2=6.645(P=0.674),respectively,which indi-cated that the actual and ideal curves of the two groups overlapped well.The analysis of the decision curves of training set and validation set showed that when the threshold probability of the occurrence of PIM in elderly patients with chronic diseases was in the range of 20％～94％and 30％～94％,there was a clinical practical value.Conclusion The constructed Nomogram model has good differentiation,calibration and clinical applicability,which can effectively and conveniently predict PIM in the elderly patients with chronic diseases,and provide a reference for early clinical identifi-cation and targeted intervention.
ElderlyPotentially inappropriate medicationNomogramPrediction model
万方数据
维普
