Abstract
Background:
Low response rates in randomised controlled trials can compromise the reliability of the results, so ways to boost retention are often implemented. Although there is evidence to suggest that sending a text message to participants increases retention, there is little evidence around the timing or personalisation of these messages.
Methods:
A two-by-two factorial SWAT (study within a trial) was embedded within the MiQuit3 trial, looking at smoking cessation within pregnant smokers. Participants who reached their 36-week gestational follow-up were randomised to receive a personalised or non-personalised text message, either one week or one day prior to the telephone follow-up. Primary outcomes were completion rate of questionnaire via telephone. Secondary outcomes included: completion rate via any method, time to completion, and number of reminders required.
Results:
In total 194 participants were randomised into the SWAT; 50 to personalised early text, 50 to personalised late text, 47 to non-personalised early text, and 47 to non-personalised late text. There was no evidence that timing of the text message (early: one week before; or late: one day before) had an effect on any of the outcomes. There was evidence that a personalised text would result in fewer completions via telephone compared with a non-personalised text (adjusted OR 0.44, 95% CI 0.22–0.87, p=0.02). However, there was no evidence to show that personalisation or not was better for any of the secondary outcomes.
Conclusion:
Timing of the text message does not appear to influence the retention of participants. Personalisation of a text message may be detrimental to retention; however, more SWATs should be undertaken in this field.
Low response rates in randomised controlled trials can compromise the reliability of the results, so ways to boost retention are often implemented. Although there is evidence to suggest that sending a text message to participants increases retention, there is little evidence around the timing or personalisation of these messages.
Methods:
A two-by-two factorial SWAT (study within a trial) was embedded within the MiQuit3 trial, looking at smoking cessation within pregnant smokers. Participants who reached their 36-week gestational follow-up were randomised to receive a personalised or non-personalised text message, either one week or one day prior to the telephone follow-up. Primary outcomes were completion rate of questionnaire via telephone. Secondary outcomes included: completion rate via any method, time to completion, and number of reminders required.
Results:
In total 194 participants were randomised into the SWAT; 50 to personalised early text, 50 to personalised late text, 47 to non-personalised early text, and 47 to non-personalised late text. There was no evidence that timing of the text message (early: one week before; or late: one day before) had an effect on any of the outcomes. There was evidence that a personalised text would result in fewer completions via telephone compared with a non-personalised text (adjusted OR 0.44, 95% CI 0.22–0.87, p=0.02). However, there was no evidence to show that personalisation or not was better for any of the secondary outcomes.
Conclusion:
Timing of the text message does not appear to influence the retention of participants. Personalisation of a text message may be detrimental to retention; however, more SWATs should be undertaken in this field.
| Original language | English |
|---|---|
| Article number | 637 |
| Journal | F1000research |
| Volume | 10 |
| DOIs | |
| Publication status | Published - 22 Jul 2021 |
Bibliographical note
© 2021 Coleman E et al.Open peer review versions 1 & 2.