LAUSR

Background: GOLD 2017 proposed not to consider airflow limitation (FEV1) to stratify COPD patients in order to guide their pharmacological treatment. Objective: To investigate the temporal stability (3 years follow-up) and association with all-cause mortality of the four 2017 GOLD groups (A-D) as compared to that of 2011. Methods: We used the ECLIPSE cohort to investigate these questions. Results: We found that GOLD 2017: (1) increased the proportion of patients in A and B groups (from 24 to 39% and from 14 to 39%, respectively), with a parallel decrease in that of C and D patients (from 23 to 7% and from 40 to 15%, respectively; (2) the proportion of patients who remained available for analysis at the end of follow-up (excluding deaths, lost for follow-up and/or missing data) was similar (A: 87 vs. 83%; B: 72 vs. 70%; C: 78 vs. 81%; and, D: 68 vs. 71%, respectively (2011 vs. 2017)); (3) at three years follow-up, the proportion of patients who remained in the same group as baseline using GOLD 2011 or 2017, respectively, were 56 vs. 57% (A), 33 vs. 52% (B), 47 vs. 26% (C) and 78 vs. 52% (D); (4) all-cause mortality at three years was 7% (A), 16% (B), 9% (C) and 18% (D). With the 2011 classification, these figures were, respectively, 4%, 10%, 8% and 14%. Conclusions: Changes in the baseline and follow-up distribution of groups reflect the “anchoring” effect of FEV1 (present in 2011 and absent in 2017). This is in keeping with recent analyses (Cabrera Lopez AJRCCM 2017). The mortality rate seen in GOLD 2017 B has also been reported recently (Gedebjerg TLRM 2018) and highlights the clinical relevance of symptom assessment as an important therapeutic goal in COPD patients.