Gates 2009.

Study characteristics
Methods	Quasi‐randomisation, individuals randomised
Data	UK, secondary care setting. The sample included all host trial participants were being sent a follow‐up questionnaire at 4 months or 8 months. Total n = 2144; mean age 36.9 (SD 13.3) years; 56% females
Comparisons	Intervention group received a £5 gift voucher, redeemable at a range of shops with their questionnaire, and a covering letter including a sentence explaining that the voucher is to thank participants for their time and effort. Control group received no gift voucher, and a standard covering letter
Outcomes	Questionnaire return
Notes	Retention period: 4 and 8 months
Risk of bias
Item	Authors' judgement	Support for judgement
Allocation concealment?	No	There was lack of concealment of allocations before randomisation.
Adequate sequence generation?	No	Allocation to trial arms was according to whether a specific digit of the participant's trial number was odd or even.
Blinding of participants and personnel?	Yes	Trial office staff were unblinded. Unclear about other personnel or participants. Not clear if participants were blinded to effect of embedded trial intervention on retention but unblinding not likely to impact objective outcome.
Blinding of outcome assessment?	Yes	Trial office staff were unblinded. However, objective outcome, participants blind (did not know there is a study) staff have no plausible additional opportunity to influence postal response rate once questionnaires sent.
Incomplete outcome data addressed? All outcomes	No	There is a difference between the CONSORT diagram and Table 2 regarding the number of non‐responders.
Free of selective outcome reporting?	Unclear	Concerns raised due to inconsistency of outcome reporting (see above).
Other sources of bias	Unclear	No further concerns raised.
Overall Risk of Bias	No	Low