Validation of exacerbation history grading by GOLD for predicting future COPD outcomes

<bold>Background:</bold> Individuals with COPD often experience exacerbations (ECOPD). For over a decade, high ECOPD risk is stratified according to a history of ≥2 moderate ECOPD and/or ≥1 severe ECOPD in <12 months. To date, these cutoffs lack validation. <bold>Objectives:</bold> To validate the ECOPD history frequency cutoff values suggested by GOLD, and to explore potential other cutoff strategies to predict ECOPD and all-cause mortality. <bold>Methods:</bold> Data from patients with COPD of the German COSYCONET cohort study were analyzed. Risk of ECOPD, hospitalizations and all-cause mortality during 4.5 years were assessed using binomial logistic regressions, and area under the ROC curves (AUC-ROC) with 95% confidence intervals (CI). <bold>Results:</bold> 2,291 patients with COPD GOLD 1-4 (mean age 65±8 years, 61% male, FEV1 53±19% predicted) were included. ECOPD history cutoff points by GOLD showed an AUC of 0.63 (95% CI 0.60-0.65) and 0.62 (95% CI 0.58-0.66) to predict ECOPD and hospitalizations, respectively. One previous ECOPD <12 months (AUC 0.66, 95% CI 0.64-0.69) more accurately predicted ECOPD and hospitalizations, and in line with GOLD, one previous hospitalization <12 months (AUC 0.63, 95% CI 0.60-0.67) predicted ECOPD and hospitalizations. The 4-year mortality rate was 9.6%. Patients with ≥3 previous ECOPD or ≥1 previous hospitalizations in <12 months were 2.18 (95% CI 1.27-3.72) and 1.57 (95% CI 1.29-1.91) times more likely to die, respectively. <bold>Conclusions:</bold> This study highlights the limited predictive performance of the current COPD risk assessment by GOLD. Novel cutoffs were suggested, categorizing patients as non-exacerbators or high-risk exacerbators based on a history of ≥1 ECOPD and/or ≥1 hospitalization <12 months.