LAUSR

OBJECTIVE Our aim was to design and evaluate a novel behaviour change approach to increase response rates to an annual postal questionnaire in three randomised Studies within a Trial (SWATs) and replicate the most promising SWAT. STUDY DESIGN AND SETTING SWAT1 tested a trial logo sticker on questionnaire envelopes vs no sticker; SWAT2 tested a theoretically informed letter sent with the questionnaire versus a standard letter; SWAT3 tested a theoretically informed newsletter sent prior to the questionnaire versus no newsletter. The SWATs were conducted within a large dental trial (N=1,877 adults) and SWAT2 replicated in a different trial in a similar setting (N=2,372). RESULTS SWAT1 improved response rates by 1.4%, 95% confidence interval (CI) (-7.2%, 10.0%). SWAT2 improved response rates by 7.0%, 95% CI (1.7%, 12.3%). SWAT3 improved response rates by 0.8%, 95% CI (-5.1%, 6.7%). Replication of SWAT2 as the most promising SWAT showed improvement in response rates of 1.0%, 95% CI (-3.2%, 5.3%). Pooled results from SWAT2 showed an overall improvement in response rates of 3.4%, 95% CI (0.1%, 6.7%). CONCLUSION A theory-based behavioural approach to design interventions to improve trial response rates showed small, but meaningful improvements. The approach presented here can be easily implemented and adapted to address other identified barriers to trial retention.