Objective To track discharged patients with chronic obstructive pulmonary disease (COPD) based on the Health Social Determinants (SDH) theory, construct a predictive risk model for readmission in patients with acute exacerbation of COPD (AECOPD), and provide a basis for implementing precise prevention and control to reduce readmission.

Methods A total of 166 patients with AECOPD were selected, and followed up for 12 months. The occurrence of readmission was recorded. The SDH, behavioral psychology and clinical indicators were collected. Univariate and multivariate logistic regression were used to screen rhe risk factors, and a nomogram model was constructed. The model efficacy was verified by using the ROC curve and calibration curve. Clinical benefits were evaluated using the decision curve (DCA).

Results A total of 50 cases (30.12%) were readmitted due to AECOPD. The rersults of Multivariate logistic regression showed that no home-based oxygen therapy, low family care, insufficient health literacy, fear of disease progression, history of acute exacerbation within one year in the past, mMRC grade elevating and ≥4 comorbidities were the independent risk factors of readmission (P ＜ 0.05). The AUC of the model was 0.935 (95%CI: 0.893−0.978), with extremely high discrimination, a sensitivity of 86.0% and a specificity of 89.7%. The calibration curve fited well. The DCA evaluation model had good clinical practicability.

Conclusions The AECOPD readmission risk model constructed based on SDH has good predictive efficacy.