Introduction
Retaining participants within a trial and successfully recording the outcome measures are key requirements for a high-quality clinical trial. Failure to retain individuals within a trial can threaten both internal and external validity of its findings.134,135 Loss of more individuals from one study arm than the other could lead to biased estimates of treatment effect. Selective loss of specific groups of individuals can limit the generalisability of the findings. Recent studies have emphasised the importance of developing a strategy to ensure high retention in randomised trials.136,137 Robinson et al.135 argue that, given the cost of mounting, running and analysing a trial, it would be only prudent to put effort into ensuring retention of individuals.
A rule of thumb for retention rates in clinical trials has been proposed by several authors;136,138 a < 5% loss to follow-up is of little concern, whereas a > 20% loss casts serious doubts on the validity of the trial. Achieving high retention rates may be more challenging in interventions delivered in community settings than those in clinical settings. One review of community trials found loss to follow-up rates between 1% and 56% with a median rate of 24%.134 If the 20% cut-off had been applied two-thirds of trials would be discarded. Trials of brief interventions for alcohol also experience high loss to follow-up. Reviews have found that the average attrition rates are usually well above the 20% threshold.8,9,139 This chapter reviews the development of a strategy to maximise retention within the trial and reports on its impact.
Developing a retention strategy
Recent review papers have identified a variety of techniques to increase retention in follow-up studies.135,140,141
Box 3 summarises the key findings. Rather than focusing on technical issues of achieving contact, the strategies emphasise cultural and person-sensitive approaches. They also emphasise that all stages of the study, from recruitment strategy to outcome assessment, should be designed to achieve a high recruitment.
Strategies to promote retention Create project identity
Application of recommended retention strategies
The retention strategy of this feasibility study applied the recommended strategies within three components of study design: the method of recruitment, the nature of the intervention and the outcome assessment (Box 4). Several of the techniques which were included in the recruitment strategy were also useful for retention. These included using the university logo to establish the credibility of the research team. Financial incentives were used to recruit and retain participants.
Trial strategies to maximise retention Use of university logo on all correspondence
Several techniques to promote retention were incorporated into the intervention. The messages were tailored to the target group, using preferred first name, and using quotes from focus groups to capture appropriate language. Questions were asked in the text messages throughout intervention delivery to promote interaction. Possibly the most important feature was the design of the intervention, which was enjoyable as well as informative. The behaviour change strategies were presented using texts and images designed to appeal to the target group. Texts on the harms of alcohol were designed to be gain-framed rather than loss-framed. Regular contact was maintained, and personal names were used.
Impact of the strategies
The retention rate in the trial was 96% (64/67). Three men from the intervention group were lost to follow-up, of whom two were recruited by RDS. Two were aged between 25 and 29 years and were single, while the third was aged > 40 years and was married. Two of the men were in employment and both had vocational training after leaving school. The men lost to follow-up drank infrequently but they all drank in excess of 16 units in each drinking session. These men were in pre-contemplation on the Readiness to Change Questionnaire and two reported that they had no plans to cut down on their drinking.
The retention rate in this study is substantially higher than most studies of brief alcohol interventions9 and most studies of community-based interventions.134 The result is more remarkable because it was achieved with disadvantaged men, a group recognised to be difficult to retain in research studies.
It was not possible to identify which component(s) of the recruitment strategy contributed to this success. However, there is clear evidence that many of the men valued the experience of taking part in the study. Following the final text message to the participants, which thanked the participants for taking part, some men spontaneously texted to thank the researcher. Their comments show how much they enjoyed participating, confirming the high level of engagement identified in the process evaluation:
No worries my pleasure.
Well really enjoyed some of the interesting facts so ur welcome Donald.
Thank for all the texts some were really funny good luck with your study thanks again look forward to chatting in a couple of months.
It was fun.speak soon bud.
Okay its been gud taking part in ur study!and i wul hear from you in 2 months then ok spk to then k from <<NAME>>!
Summary
Through the use of multiple evidence-based strategies embedded throughout the study, a retention rate of 96% was achieved with a difficult-to-reach group. Interventions delivered by mobile telephone which tackle adverse health behaviours typically have retention rates of 70–90%23,24,69 so the present study ranks among the highest. The finding was unexpected because disadvantaged individuals are more likely to be lost to follow-up.142,143
The use of incentives may have played a part, as one trial which gave a lower incentive to the intervention group at follow-up had higher attrition in that group.69 It seems likely that the features of the intervention, tailoring to the target group and the use of humour and interactivity led to a high level of retention. This suggests that future studies should ensure not only that their interventions are theoretically and empirically based, but also that they are sufficiently enjoyable and rewarding to promote engagement and retention in the study.
