GOLD Classification of COPD
The GOLD classification system stratifies COPD using two complementary frameworks: spirometric grades (1-4) based on FEV1 % predicted, and ABCD groups that integrate symptoms, spirometry, and exacerbation history to guide treatment decisions 1.
Spirometric GOLD Grades (Airflow Limitation Severity)
The spirometric classification requires post-bronchodilator FEV1/FVC <0.70 plus:
- GOLD 1 (Mild): FEV1 ≥80% predicted
- GOLD 2 (Moderate): FEV1 50-79% predicted
- GOLD 3 (Severe): FEV1 30-49% predicted
- GOLD 4 (Very Severe): FEV1 <30% predicted
ABCD Assessment Groups (Combined Risk Assessment)
This multidimensional system combines three elements to assign patients to groups A, B, C, or D:
1. Symptom Assessment (Choose Higher Result)
- mMRC Dyspnea Scale:
- Low symptoms: mMRC 0-1
- High symptoms: mMRC ≥2
- CAT Score:
- Low symptoms: CAT <10
- High symptoms: CAT ≥10
2. Exacerbation Risk (Choose Higher Risk)
- Low risk: 0-1 moderate exacerbations per year (not requiring hospitalization)
- High risk: ≥2 moderate exacerbations OR ≥1 severe exacerbation requiring hospitalization in past year
3. Spirometric Grade
- Low risk: GOLD 1-2
- High risk: GOLD 3-4
The final exacerbation risk category is determined by whichever criterion (exacerbation history or spirometry) places the patient at higher risk 2.
Final ABCD Group Assignment
- Group A: Low symptoms + Low risk (fewer symptoms, low exacerbation risk)
- Group B: High symptoms + Low risk (more symptoms, low exacerbation risk)
- Group C: Low symptoms + High risk (fewer symptoms, high exacerbation risk)
- Group D: High symptoms + High risk (more symptoms, high exacerbation risk)
Clinical Implications and Caveats
Important discordances exist between classification criteria. The equivalence between mMRC ≥2 and CAT ≥10 for identifying "high symptoms" shows poor agreement—CAT ≥10 has 82% sensitivity but only 24% specificity for mMRC ≥2 3. Similarly, spirometric grades and exacerbation history show substantial disagreement (kappa 0.12), with 45% of patients classified discordantly 3.
Group C is clinically rare in practice. Studies consistently show only 4-12% of patients fall into Group C (low symptoms but high exacerbation risk), while Group D predominates at 59-60% 4, 5. This suggests the theoretical four-group model may not reflect real-world phenotypes.
Mortality prediction differs between systems. Spirometric GOLD grades (1-4) predict mortality substantially better than ABCD groups, with clear separation of survival curves across grades 6. Paradoxically, Group B patients (high symptoms, low risk) show worse mortality than Group C patients despite better lung function, likely due to cardiovascular disease and cancer 4. The ABCD system performs better for predicting exacerbations but worse for mortality and lung function decline 7.
Treatment recommendations vary by country but generally follow this pattern: Group A receives short-acting bronchodilators; Group B receives long-acting bronchodilators (LABA or LAMA); Groups C and D receive LAMA or ICS+LABA combinations, with triple therapy (ICS+LABA+LAMA) reserved for Group D patients with persistent symptoms or exacerbations 1.