Development and evaluation of a clinical prediction model for in-hospital mortality in patients with acute exacerbation of chronic obstructive pulmonary disease.

BACKGROUND: Acute exacerbation of chronic obstructive pulmonary disease (AECOPD) is a major cause of hospitalization and is associated with high in-hospital mortality; consequently, early identification of high-risk patients is crucial. This study aimed to develop and validate a clinical prediction model for in-hospital mortality among patients with AECOPD using readily available admission indicators to aid in risk stratification.

METHODS: A retrospective analysis was conducted on 287 patients with AECOPD admitted to the Emergency Department and Respiratory and Critical Care Medicine Department of The Sixth People's Hospital of Nantong between February 2022 and February 2025. The complete patient dataset was randomly and equally divided into five mutually exclusive subsets (or folds). Each subset was sequentially used as the validation set, while the remaining four served as the training set. The training datasets were used for variable screening and model construction, and this process was repeated five times. Ultimately, the model's performance metrics were calculated as the average of the five validation results.

RESULTS: In the mortality group, parameters such as partial pressure of carbon dioxide (PaCO2) (mmHg) (P<0.001), lactate (mmol/L) (P<0.001), creatinine (μmol/L) (P<0.001), high-sensitivity troponin T (hscTnT) (ng/mL) (P<0.001), and heart rate (beats/min) (P<0.001) were significantly higher in the AECOPD patients compared to the improvement group. Although diastolic blood pressure (P=0.07), gender (P>0.99), and age (P=0.38) showed no statistical significance in univariate analysis, they were included in the multivariate model for adjustment based on their known clinical importance and to control for potential confounding effects. Subsequently, these parameters were incorporated as independent variables in the multivariate analysis, and a predictive model was constructed using Firth penalized logistic regression. According to the Youden index, the optimal cutoff value for this model was determined to be 0.080, corresponding to a sensitivity of 87.5% and specificity of 90.1%.

CONCLUSIONS: Using prognosis (in-hospital outcome) as the dependent variable (death =1 and improvement =0), we ultimately identified eight variables-age, sex, PaCO2 (mmHg), lactate (mmol/L), creatinine (μmol/L), hscTnT (ng/mL), heart rate (beats/min), and diastolic blood pressure (mmHg)-that can serve as valuable prognostic predictors for patients with AECOPD.