: The 2-dimensional, 4-quadrant 2017 Global Initiative for Chronic Obstructive Lung Disease (GOLD) COPD A-D assessment tool (GOLD2017) does not include lung function variables to classify patients into different risk groups. The previous 2011 tool (GOLD2011) classified cases in the upper-quadrants (higher risk groups) regardless of whether they had a history of exacerbations or worse lung function. We hypothesized that a modified, three-dimensional classification (GOLD3D) that separately includes assessment of lung function and exacerbations history would improve the ability to predict adverse events. Methods: A total of 1303 COPD patients were included in a historical cohort study. The ability of GOLD3D to predict outcomes (all-cause death and hospitalizations due to severe exacerbation) was compared with GOLD2017 and GOLD2011. Results: Mean follow-up was 45.0 ± 28.0 months. Two hundred and twenty-eight patients (17.5%) died and 337 (25.9%) subjects suffered at least a severe exacerbation that required hospital admission. The area under the receiver-operating characteristics curve for mortality prediction was slightly but significantly higher for GOLD3D than for GOLD2011. The area under the curve for prediction of severe exacerbations was significantly higher for GOLD3D than for GOLD2011 and GOLD2017. A worse ventilatory obstruction was associated in most cases with a higher mortality risk and a higher exacerbation risk for the GOLD2017 A-D groups. Conclusions: The proposed GOLD3D classification system upgrades the previous ones, and is advantageous in predicting future adverse events.Le texte complet de cet article est disponible en PDF.
Key words : Pulmonary disease, Chronic obstructive, Prognosis, Mortality, Exacerbation, Gold