Risk factors of postoperative non-compliance behavior in elderly patients with gastrointestinal malignant tumors

Elderly patients with gastrointestinal malignancies frequently exhibit postoperative non-compliance behaviors, which includes failure to persist with adjuvant therapy, attend regular follow-up appointments, and take medications strictly as prescribed. These non-compliant behaviors seriously affect the prognosis of patients.

We conducted a large-scale retrospective study in China, which enrolled 284 elderly patients who underwent surgery for gastrointestinal malignancies at Jingzhou Hospital Affiliated to Yangtze University between February 2018 and February 2023. The patients were randomly divided into a training group (198 patients, 70%) and a validation group (86 patients, 30%). Univariate and multivariate logistic regression analyses were performed to identify independent risk factors for postoperative non-compliance behaviors (P < 0.05), based on which a predictive nomogram was constructed. The model's performance was evaluated using receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis (DCA). Furthermore, all patients were followed up for 2 years, and survival curves were plotted.

The results showed that education level, disease type, history of heart disease, tumor differentiation, net family income, and living alone were independent risk factors for postoperative non-compliance behaviors in elderly patients with gastrointestinal malignancies (P < 0.05). These factors were incorporated into the construction of a nomogram prediction model. The areas under the curve (AUC) for the training and validation groups were 0.865 (95% CI: 0.814-0.915) and 0.871 (95% CI: 0.797-0.946), respectively. Calibration curve results indicated that the predicted probabilities of the model were in close agreement with the actual observed frequencies in both the training and validation groups, with mean absolute errors of 0.037 and 0.023, respectively. DCA demonstrated that the model provided significant net clinical benefit within threshold probability ranges of 0.11-0.99 for the training group and 0.41-0.91 for the validation group. Kaplan-Meier analysis showed significantly better survival in the compliance group. Disease-free survival (DFS) rates at 6, 12, and 24 months were 90.1%, 85.9%, and 79.6% versus 73.2%, 66.9%, and 62.0% in the non-adherence group (all P < 0.05). Overall survival (OS) was also significantly higher (P < 0.001), with a hazard ratio of 0.28.

Our research suggests that early intervention for elderly patients with low education levels, lesions located in the colon, comorbid heart disease, low tumor differentiation, low monthly family net income, and living alone who do not follow medical advice after gastrointestinal malignant tumor surgery can change the patient's prognosis.