PRECISE-X模型开发与验证：预测COPD首次重度急性加重

原文

OBJECTIVES: In patients with chronic obstructive pulmonary disease (COPD), severe exacerbations (ECOPDs) impose significant morbidity and mortality. Current guidelines emphasise using ECOPD history to inform preventive treatments but offer limited guidance for risk stratification for the first severe ECOPD. METHODS: We developed and validated PRECISE-X using a cohort of newly diagnosed COPD patients from the UK's Clinical Practice Research Datalink (2004-2022), to predict first severe ECOPD over 5 years (primary outcome) and 12 months (secondary outcome). Predictors were selected via clinical expertise and data-driven methods. Internal-external cross-validation was performed across practice regions to evaluate the model's out-of-sample performance in terms of discrimination (c-statistic), calibration and net benefit. RESULTS: The study included 2 19 015 patients (mean age 66.0; 42.4% female). Observed risk of first severe ECOPD was 29.5% at 5 years (4.2% at 1 year). The final model included four mandatory predictors (sex, age, Medical Research Council dyspnoea score and forced expiratory volume in 1 second) and 28 optional predictors. In internal-external cross-validation, the average out-of-sample c-statistic was 0.836 (95% CI 0.827 to 0.846) for 5-year prediction and 0.756 (95% CI 0.746 to 0.766) for 1-year prediction. Calibration across regions was robust, and the model showed positive NB across a wide range of risk thresholds. In a secondary validation assessment among those with available spirometry data with confirmed airflow obstruction, the model was well calibrated and had only a modest decline in discriminatory performance. CONCLUSIONS: PRECISE-X accurately predicts the first severe COPD exacerbation using routine clinical data, supporting earlier risk stratification and proactive disease management.