GOLD COPD exacerbation history groups ‘mediocre’ in exacerbation risk estimation
Key takeaways:
- Risk estimation was based on area under the receiver operating characteristic curve, sensitivity and specificity.
- Suggested cutoff for moderate exacerbations in this study lowers the current one.
The Global Initiative for Chronic Obstructive Lung Disease, or GOLD, COPD exacerbation history groups were “mediocre” in estimating moderate and severe exacerbation risk at 1 and 4 years, according to results published in JAMA Network Open.
“This study revealed poor performance of GOLD’s current cutoffs of two or more moderate [exacerbations of COPD (ECOPD)], one or more severe ECOPD or both for estimating future COPD-related outcomes,” Kiki Waeijen-Smit, MSc, postdoctoral researcher at Maastricht University’s department of respiratory medicine, and colleagues wrote.
Using the German COPD and Systemic Consequences-Comorbidities Network study, Waeijen-Smit and colleagues evaluated 2,291 patients (mean age, 65 years; 60.9% men; mean estimated FEV1, 52.5%) with COPD to determine if the current GOLD groupings using COPD exacerbation history (A, B and E) are accurate in estimating three outcomes: moderate exacerbation risk at 1 and 4 years, future severe exacerbation risk at 1 and 4 years and all-cause mortality risk at 4 years.
During a 4.5-year follow-up period, 1,996 patients had data for visit 2, 1,724 had data for visit 3, 1,182 had data for visit 4 and 878 had data for visit 5.
Based on area under the receiver operating characteristic curve (AUROC), researchers reported “poor performance” of the COPD exacerbation history categories in estimating 1-year moderate exacerbation risk (AUROC, 0.63; sensitivity, 39.8%; specificity, 85.5%). When estimating severe exacerbation risk, a comparable AUROC was found (0.62; sensitivity, 47.3%; specificity, 76.8%).
Switching to estimating 4-year COPD exacerbation risk, AUROCs continued to stay around 0.6 for moderate exacerbation risk estimation (0.6; sensitivity, 35.1%; specificity, 85.3%) and severe exacerbation risk estimation (0.61; sensitivity, 42.7%; specificity 78.8%), according to the study.
“The cutoff value with the optimal sensitivity and specificity scores to estimate 1-year moderate ECOPD was 1 moderate and 1 severe ECOPD event within 12 months,” Waeijen-Smit and colleagues wrote.
Researchers reported that the above cutoff finding held true for 4-year moderate exacerbation risk estimation, 1-year severe exacerbation risk estimation and 4-year severe exacerbation risk estimation.
“Novel cutoffs were suggested, categorizing patients as without exacerbations or with high-risk exacerbations based on a history of 1 or more moderate ECOPD, 1 or more severe ECOPD or both within 12 months,” Waeijen-Smit and colleagues wrote.
For comparison, the AUROC to estimate future moderate exacerbations when using one previous moderate COPD exacerbation within 12 months was 0.66 (sensitivity, 62.6%; specificity, 70.1%). The AUROC to estimate future severe exacerbations when using one previous severe COPD exacerbation within 12 months was 0.63 (sensitivity, 68.4%; specificity, 58.2%), which researchers wrote was “in line with GOLD 1.”
Death occurred for 9.6% of the population over 4.5 years. When estimating 4-year all-cause mortality risk, researchers again reported low performance with use of the GOLD COPD exacerbation history categories (AUROC, 0.55; sensitivity, 36.6%; specificity, 73.7%).
The study noted elevated odds for 4-year all-cause mortality with at least three previous moderate COPD exacerbations vs. no previous exacerbation within 12 months (OR = 2.18; 95% CI, 1.27-3.72). In terms of severe exacerbation history, higher odds for all-cause mortality were found with at least one previous severe exacerbation vs. no previous serious exacerbation within 12 months (OR = 1.57; 95% CI, 1.29-1.91).
“Future studies are needed to validate the proposed cutoffs and elaborate further on other determinants aside from ECOPD history to estimate ECOPD risk, as well as their optimal, clinically applicable combination with the proposed novel approach,” Waeijen-Smit and colleagues wrote. “Moreover, the association of this lower cutoff with potential overtreatment of patients and subsequent financial implications needs to be evaluated.”
Perspective
Back to Top
COPD affects 50 million Americans with an additional 15 million likely undiagnosed. It was the ninth leading cause of death in 2016 and is expected to rise to fourth by 2040. This disease makes breathing harder, lowers quality of life and costs a lot to manage. Current guidelines (GOLD criteria) use past flare-ups to predict risk, but new research shows the benchmarks may be outdated, leaving patients vulnerable to exacerbations.
This study of 2,291 patients revealed gaps in GOLD’s threshold (≥ 2 moderate or ≥ 1 severe flare-up annually) that poorly predicted future crises with an accuracy of 63% or less. Lowering the high-risk threshold to just one moderate flare-up improved detection but still missed critical details. Patients with at least three moderate or at least one severe flare-ups faced double the risk of death proving that every flare-up accelerates decline. However, relying only on flare-up history ignores other key factors such as heart disease and poverty.
In clinical practice even one moderate flare-up should trigger action by starting optimized medical therapy, pulmonary rehab and managing heart and comorbid conditions. Emphasis on prevention efforts should be increased by expanding low-cost vaccines, smoking cessation programs and clean air policies to protect vulnerable lungs.
To stop the rise of COPD, we need global validation of new thresholds across diverse populations, AI tools that combine flare-up data with social and environmental factors, and policies to reduce health care disparities.
Collapse
Read more about