Establishment and evaluation of a nomogram model for predicting postoperative frailty in elderly patients with gastrointestinal tumors / 中国实用护理杂志

ABSTRACT

Objective:To analyze the influencing factors of postoperative frailty in elderly patients with gastrointestinal tumors, establish a nomogram model for predicting postoperative frailty and evaluate its efficacy, so as to provide reference basis for formulating perioperative frailty management plans for elderly gastrointestinal tumor patients in the later stage.Methods:Convenience sampling method was used to select 376 elderly patients with gastrointestinal tumors who underwent surgical treatment in the First Affiliated Hospital of Anhui Medical University from December 2021 to August 2022 as the study objects. On the 5th day after surgery according to Tilburg Frailty Indicator, the patients were diagnosed whether they developed postoperative frailty and were divided into postoperative frailty group and postoperative non-frailty group. General data, laboratory indicators and clinical data of patients were collected. Univariate analysis and multivariate Logistic regression analysis were used to screen the independent influencing factors of postoperative frailty in elderly patients with gastrointestinal tumors. R software was used to establish a nomogram prediction model and conduct internal validation to evaluate the differentiation, calibration and clinical applicability of the model.Results:A total of 265 males and 111 females included aged (70.04 ± 5.89) years old, with 222 (59.0%) patients experienced postoperative frailty in this study. Multivariate analysis showed that low Barthel Index score ( OR=0.941, 95% CI 0.903-0.980), low hemoglobin ( OR=0.976, 95% CI 0.963-0.989), high Charison Comorbid Index score ( OR=1.457, 95% CI 1.128-1.882), preoperative frailty ( OR=4.369, 95% CI 1.486-12.841), and pathological stage Ⅲ-Ⅳ ( OR=2.053, 95% CI 1.253-3.364) were independent influencing factors for postoperative frailty of elderly gastrointestinal tumors (all P<0.05). The AUC before and after internal validation of the nomogram model was 0.811 (95% CI 0.768-0.854) and 0.803 (95% CI 0.762-0.856) respectively. The results of Hosmer-Lemeshow test showed good goodness of fit ( χ2=4.09, P>0.05). Decision curve analysis showed that the model had certain clinical applicability. Conclusions:Based on the risk factors of postoperative frailty in elderly patients with gastrointestinal tumors, the nomogram prediction model was established, which has good differentiation, consistency and clinical applicability, and can provide reference for clinical staff to make perioperative frailty management plan.