Is it feasible to recruit problem drinkers into a trial to compare an online alcohol treatment against face-to-face treatment?
Over seven million adults in England drink alcohol at “hazardous or harmful” levels. Identification and brief advice (IBA) is effective at reducing consumption, with onward referral to community drug and alcohol teams (CDATs) for specialist treatment if more help is needed. However, only a small proportion of people who could benefit go on to access alcohol services due to stigma, and cuts to services/counsellors. Hence, alternatives to face-to-face treatments are needed. Web-based programmes may improve access, and may be effective, but most trials have compared online programmes against assessment only and few have compared against face-to-face treatment. Trials of web-based alcohol treatment programmes against face-to-face treatment are needed, because if equally effective this could increase patient choice and improve access to evidence-based, cost-effective therapy.
Previous alcohol studies have found it is difficult to recruit from primary care, so we hypothesised that recruitment would be more successful from a population of help-seeking clients referred, or self-referred, to CDATs. We therefore conducted a feasibility RCT to determine if recruitment was possible in this setting. The DIAMOND (Digital Alcohol Management on Demand) study examined recruitment and retention, and also online randomisation and data-collection methods. Four CDATs in north London (Camden, Islington and Haringey) agreed to take part. Recruitment ran from 1 January to 31 August 2015.
Over the study period alcohol counsellors screened 1253 patients, the majority of whom were not eligible for our study. 579 were dependent drinkers and 548 had complex needs: severe comorbid drug use or physical or mental health problems (n=548); social service issues (n=41); homelessness (n = 11); 10 were ineligible for other reasons including not being computer literate (n=10).
Of the 64 eligible patients, 54 declined to participate: preferred face-to-face treatment (n = 36); offered no reason (n=13); had social service issues (n=2); did not want to use a computer (n=2); going on holiday (n=1). We recruited 10 participants to the study, 7 were randomised and we collected follow up data on 4 of these.
These results show the importance of conducting feasibility studies before undertaking fully-powered RCTs of complex interventions. We found that our original recruitment strategy was not feasible, and that it would not be possible to recruit in CDATs, as they see mainly dependent drinkers or hazardous and harmful drinkers with additional severe and complex needs, and not people for whom an online intervention would be appropriate. We now plan to extend the feasibility study by recruiting via Emergency Departments (EDs) at five north London Hospitals with posters and leaflets in waiting rooms, and active recruitment by researchers in one ED, with further recruitment via links from the IBA website ‘Don’t Bottle it Up.’