The effect of using social pressure in cover letters to improve retention in a longitudinal health study: an embedded randomised controlled retention trial

The effect of using social pressure in cover letters to improve retention in a longitudinal health study: an embedded randomised controlled retention trial

Trials, July 2017

28728603

Sarah Cotterill, Kelly Howells, Sarah Rhodes, Peter Bower

Retention of participants in cohort studies is important for validity. One way to promote retention is by sending a persuasive cover letter with surveys. The study aimed to compare the effectiveness of a covering letter containing social pressure with a standard covering letter on retention in a health cohort study. Social pressure involves persuading people to behave in a certain way by the promise that their actions will be made know to others. We implemented a mild form of social pressure, where the recipient was told that information about whether they responded to the current survey would be noted by the research team and printed on future correspondence from the research team to the recipient. The design was an embedded randomised controlled retention trial, conducted between July 2015 and April 2016 in Salford, UK. Participants in the host health cohort study were eligible. They received either: (1) a covering letter with two consecutive surveys (sent six and twelve months after recruitment), containing a social pressure intervention; or (2) a matching letter without the social pressure text. The primary outcome was retention in the host study, defined as return of both surveys. Randomisation was computer-generated, with stratification by household size. Participants were blinded to group assignment. Researchers were blinded for outcome ascertainment. Adults (n = 4447) aged over 65 years, with a long-term condition and enrolled in the host study, were randomly allocated to receive a social pressure covering letter (n = 2223) or control (n = 2224). All 4447 participants were included in the analysis. Both questionnaires were returned by 1577 participants (71%) sent the social pressure letters and 1511 (68%) sent control letters, a risk difference of 3 percentage points (adjusted odds ratio = 1.16 (95% confidence interval = 1.02-1.33)). A mild form of social pressure made a small but significant improvement in retention of older adults in a health cohort study. Investigation of social pressure across other research contexts and stronger social pressure messages is warranted. The host cohort study, the Comprehensive Longitudinal Assessment of Salford Integrated Care (CLASSIC) study is associated with the CLASSIC PROTECTs trial, which is registered on the ISRCTN registry. ISRCTN12286422 . Date of registration 19 June 2014.