Outcome criteria in smoking cessation trials: proposal for a common standard

TLDR

Smoking cessation treatment is widely integrated into health‑care systems, yet randomized trial outcomes are reported inconsistently, hindering interpretation. The authors propose the Russell Standard, a set of six criteria to standardize smoking cessation trial outcomes. The standard applies to trials with a defined quit date and face‑to‑face contact, requiring 6‑ or 12‑month follow‑up, self‑reported abstinence allowing up to five cigarettes, biochemical verification at the final visit, intention‑to‑treat analysis, inclusion of protocol violators, and blinded follow‑up. The authors contend that the Russell Standard balances practicality and surrogate validity, facilitating meaningful cross‑study comparisons. The authors note that adaptations may be needed for special populations or trials lacking a target quit date or face‑to‑face contact.

Abstract

Smoking cessation treatment is now integrated into many health-care systems and a major research effort is under way to improve current success rates. Until now results from randomized clinical trials have been reported in many different ways, leading to problems of interpretation. We propose six standard criteria comprising the 'Russell Standard' (RS). These criteria are applicable to trials of cessation aids where participants have a defined target quit date and there is face-to-face contact with researchers or clinic staff, as follows. (1) Follow-up for 6 months (RS6) or 12 months (RS12) from the target quit date or the end of a predefined 'grace period'; (2) self-report of smoking abstinence over the whole follow-up period allowing up to five cigarettes in total; (3) biochemical verification of abstinence at least at the 6-month or 12-month follow-up point; (4) use of an 'intention-to-treat' approach in which data from all randomized smokers are included in the analysis unless they have died or moved to an untraceable address (participants who are included in the analysis are counted as smokers if their smoking status at the final follow-up cannot be determined); (5) following-up 'protocol violators' and using their true smoking status in the analysis; and (6) collecting follow-up data blind to smokers' allocation to trial group. We believe that these criteria provide the best compromise between practicability and surrogacy for long-term cessation and will enable meaningful comparison between studies. There may be good reasons why other outcome criteria would also be reported, and studies that involve interventions with special groups or where there is no designated target quit date or face to face contact would need to adapt these criteria accordingly.