Article Text

S57 Short-term clinically important deterioration predicts long-term clinical outcome in COPD patients: A post-hoc analysis of the TORCH trial
  1. I Naya1,
  2. L Tombs2,
  3. P Jones3
  1. 1GlaxoSmithKline, Stockley Park, Uxbridge, UK
  2. 2Precise Approach Ltd on Assignment at GlaxoSmithKline, Uxbridge, UK
  3. 3Institute for Infection and Immunity, St George’s University of London., London, UK


Background COPD is a progressive disease leading to adverse outcomes such as exacerbations and death. Numerous predictors of these outcomes have been identified, based on observations at single time points, but little is known about disease trajectory as a predictor of long-term outcome. We hypothesised that the occurrence of a composite measure of clinically important deterioration (CID) made up of moderate/severe exacerbations, worsening of FEV1 or St George’s Respiratory Questionnaire (SGRQ) total score measured over 6 months may predict future long-term adverse outcomes.

Method A post hoc analysis of the TORCH data, in all four treatment arms, was performed in 5292 (86.5%) of the 6112 COPD patients in the study at 6-months (day 182). CID was defined as: decrease of ≥100 mL in post-bronchodilator FEV1, or increase of ≥4 units in the SGRQ, or a moderate/severe exacerbation. Using day 182 status, we tested the association between the occurrence of any CID type at or before 6 months and outcomes over the next 30 months including: sustained deterioration in FEV1 and SGRQ scores, moderate/severe exacerbations and mortality. A Cox’s proportional hazards model used day 182 deterioration status with covariates collected at day 182, smoking status and geographical region to estimate future risk.

Results By day 182, 2870 [54%] patients had experienced a CID (CID+) and 2422 [46%] had not (CID-). 30 months later, the CID- group had a LS mean post-bronchodilator FEV1 117 ml (95% CL 100,134; p < 0.001) higher compared to the CID+ group and the SGRQ total score was 6.4 units (95% CL 5.4, 7.5; p < 0.001) better. Over the same period, post CID+ patients had a 61% (95% CL 50, 72%; p < 0.001) increased risk of a new moderate/severe exacerbation and a 41% (95% CL 15, 72%; p < 0.001) increased risk of all-cause death vs. the CID- group (see Figure 1).

Conclusion Patients experiencing a clinically important deterioration early in the TORCH trial appeared to be set on a clinical path of sustained deterioration in both health status and FEV1 with an increased medium/long-term future risk of exacerbations and all-cause death.

Statistics from

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.