Factors predicting recruitment to a UK wide primary care smoking cessation study (the ESCAPE trial)

Abstract

Background: Recruiting smokers to smoking cessation trials is challenging and participation rates are often low. Consequently, the interventions evaluated may fail to reach a broad spectrum of the target population, thus compromising the generalizability of the findings. Brief interventions, using proactive recruitment, are likely to attract a broader and more representative proportion of the population.

Objective: We explored the factors that influenced recruitment into a trial evaluating computer-tailored feedback reports that aimed to help smokers to quit [the ESCAPE (Effectiveness of computer-tailored Smoking Cessation Advice in Primary Care) study] in order to investigate the possibilities for increasing recruitment into smoking cessation trials.

Methods: Current cigarette smokers, identified from GP records, were invited to participate in the study. The main outcome measure was the recruitment rate, i.e. the proportion of participants who responded and were randomized to one of the intervention groups. Predictor variables included geographical region, level of deprivation, practice characteristics and the number and timing of mailings of questionnaires.

Results: The recruitment rate varied by practice (2.5-19.8%) and differed significantly between regions (from 16.3% in Scotland to 8.4% in London, P < 0.001). Recruitment decreased significantly by 1.1% between the lowest and highest quintiles of deprivation (P = 0.012), measured by Index of Multiple Deprivation scores, and decreased by 1.33% for every extra 10% smokers identified within a practice population (P = 0.010). Sending reminders increased recruitment by 7.5% (P < 0.001). Multivariable analysis showed region and length of time between mailings were the main predictors of recruitment.

Conclusions: Proactive recruitment methods can increase participation in smoking cessation trials and weighting the target sample in favour of more deprived areas will recruit a more representative sample. The number and timing of mailings to potential participants can also increase recruitment.