Design: Prespecified subgroup analysis of a multicentre randomised cessation trial
Population: 172 smokers with COPD followed for one year
Primary outcome: One-year smoking cessation, verified by exhaled carbon monoxide
Effect: 25.8% (immediate referral) vs 10.1% (usual care); usual-care odds ratio 0.25 (95% CI 0.09-0.64)

Brief advice in the consultation is the usual way general practice addresses smoking in COPD, but this prespecified analysis of a multicentre cessation trial suggests referral does much better.

Among 172 COPD smokers analysed, 18.6% had quit at one year. The split by allocation was striking: 25.8% of those given immediate referral to a cessation clinic had quit against 10.1% under usual care, and usual care carried markedly lower odds of cessation in the adjusted model. Absence of comorbidity was associated with quitting; disease severity was not independently predictive once other factors were accounted for. Quitting was verified by exhaled carbon monoxide.

The actionable change is to make a direct referral to structured cessation support the default for a smoker with COPD, rather than relying on opportunistic advice. It is the single most effective thing to offer, and the pathway, not the patient's disease severity, is what moved the numbers.

Prespecified analysis of 172 COPD smokers in a multicentre cessation trial.
One-year quitting was 25.8% with immediate clinic referral versus 10.1% with usual care.
Usual care carried much lower adjusted odds of cessation (odds ratio 0.25).
Cessation was verified by exhaled carbon monoxide, not self-report alone.
Disease severity did not independently predict quitting once adjusted.

Why it matters

It shows the lever for quitting is the referral pathway, not the severity of the patient's lung disease.

The statistics, in plain English

A quit rate of 25.8% versus 10.1% is more than a doubling, and the adjusted odds ratio of 0.25 for usual care means roughly a quarter of the odds of quitting without referral. The factor analysis beyond allocation is exploratory, so read those predictors cautiously.