Enrolling men in clinical trials: An analysis of recruitment strategies and processes in the T4DM diabetes prevention trial

Abstract

Participant recruitment is a crucial component of clinical trial conduct yet an estimated 50% of trials fail to recruit to target. Trialists seeking strategies to boost participant recruitment are faced with a lack of good-quality evidence to guide their decision-making. In this context, recruitment activities are often conducted in a trial-and-error manner, resulting in wasted time and resources, and risking trial failure. Men have a lower life expectancy than women, and are also at increased risk of a number of chronic diseases including diabetes. Nevertheless, men have been underrepresented in diabetes prevention trials, and the published literature on how best to engage men in trials is limited. The aim of this research was to identify and evaluate strategies to recruit men into clinical trials. The research was set with the Testosterone for the prevention of Type 2 Diabetes (T4DM) study, a large, multi-centre, phase III, double-blind, placebo-controlled two-year trial of testosterone therapy combined with a lifestyle intervention (Weight Watchers®) compared to the lifestyle intervention alone for the prevention of Type 2 diabetes. Between January 2013 and February 2017, 19,022 participants were screened and 1007 randomised to the T4DM study. The final study visit was conducted in May 2019. A systematic review of strategies to recruit men aged 50 years and over to clinical trials was conducted. In the included studies, the most effective strategies for identifying prospective participants were referral from an affiliated health service, media coverage and mass mailing. Identification of participants through community outreach activities such as displaying posters and attending local community events was not effective. Site staff training that focused on trial-specific recruitment challenges was also effective in increasing recruitment. Of the 16 included studies, only one was assessed to be of good quality. The most common sources of bias in fair and poor quality studies were inadequate description of recruitment interventions, and failure to report cost data. Next, a program of three recruitment evaluations was conducted within the T4DM study. First, a range of promotional strategies to identify prospective T4DM participants was developed and evaluated. Repeated, high frequency bursts of radio advertising, infrequent but high reach television news coverage, and direct mass mail outs by a government health agency were the most effective recruitment strategies. Promotions through community groups and businesses, online advertising, referral by general practitioner, coverage on newspaper and radio news bulletins, and newspaper advertising were also implemented but were evaluated to be less effective. Promotional activities cost an average of AUD$594 per randomised participant (ranging from no direct cost for television, radio and newspaper news coverage to AUD$1941 per randomised participant for newspaper advertising). Second, a centralised, semi-automated approach to screening and enrolling T4DM participants was implemented and evaluated. The sequential screening process comprised: (i) web-based prescreening, (ii) laboratory eligibility screening at a network of third-party pathology centres, (iii) final on-site clinical screening. This approach delivered high-volume screening at low cost and required few staff. Efficiencies were achieved by pre-screening the majority of subjects (95%) online, automating email communications with prospective participants, outsourcing blood collection for laboratory screening to a third party pathology centre, and automating eligibility ascertainment up to the point of final clinical screening. Screening and enrolment cost on average AUD$1411 per randomised participant, including direct and indirect costs. This represented a considerable cost saving compared to previous studies in similar disease settings. Third, a ‘study within a trial’ (SWAT) was implemented to address an identified roadblock in trial recruitment. Phone call and Short Message Service (SMS) reminders were designed to increase screening uptake in participants who did not attend a further screening assessment after completing pre-screening. In a randomised comparison (N=709), there was no significant difference in screening uptake between those who received an SMS reminder and those who received a telephone call reminder (18% (62/354) and 23% (80/355) respectively) (RR=1.29, 95% CI 0.96–1.73, p=0.09). Both SMS (95% CI 14%–22%) and telephone call reminders (95% CI 18%–27%) increased screening uptake compared to neither reminder (12%). SMS reminders cost substantially less than telephone call reminders (AUD$0.53 versus AUD$6.21 per reminder), making them an adequate alternative to telephone call reminders to boost screening uptake. In total, this research provides a detailed and costed approach for recruiting men to diabetes prevention trials. The estimates of cost and effectiveness reported in this research will be of use to trialists seeking to recruit similar participant populations. More broadly, the methods for implementing and evaluating recruitment strategies presented in this thesis will be generalizable to many trial settings.