Abstract: Objective: To construct a mental health prediction model for spouses of patients with gynecological malignant tumors. Methods: A convenient sampling was used to select the spouses of 292 patients with gynecological malignant tumor in a tertiary grade A hospital from February 2019 to January 2022. Based on the total score of the Symptom Checklist 90 exceeding 160, the training set was divided into the mental health group (n=96) and the mental health abnormal group (n=109). R language and Logistic regression were used to construct a predictive model, and 87 patients were included for model validation. Results: Logistic regression analysis showed that the protective factors included social support, level of hope, education level, and the risk factors were medical payment method, birth demand and tumor metastasis. The sensitivity of the model was 0.954, the specificity was 0.844, the C-index of consistency was 0.951 (95%CI: 0.920-0.982), and the validation accuracy was good. Conclusion: The predictive model of mental health status for spouses of patients with gynecological malignant tumor drawn by the above factors has a good correlation with the actual observation results, which is worthy popularization.

Key words: gynecological malignant tumor; spouse; mental health; prediction model