Text message reminders to improve questionnaire response rates in RCTs: findings from three randomised sub-studies

OBJECTIVES To assess whether short message service (SMS) text messages sent as prenotification or postnotification reminders improve questionnaire response rates in randomized controlled trial (RCT) follow-up. STUDY DESIGN AND SETTING Three "trials within a trial" assessed text message notifications sent before or after receipt of a follow-up questionnaire within an RCT for patients with depression. Consenting patients (n = 523) were randomized to receive a prenotification or no notification at 3 months, prenotification or postnotification at 6 months, and a postnotification or no notification at 12-month follow-up. Unadjusted and adjusted questionnaire return rates and time to return were compared. RESULTS The two trials comparing prenotification or postnotification with no notification at 3- and 9-month follow-up found no evidence of an effect on questionnaire response rates (3-month response rate: 82.9% vs. 84.7% (difference 1.79%, 95% confidence interval [CI] -4.53% to 8.11%, P = 0.58); 9-month response rate: 77.1% vs. 78.5% (difference = -1.44%, 95% CI -8.56% to 5.67%, P = 0.69). For the trial at 6-month follow-up, there was a statistically significant difference in response rates for postnotification (83.2%) compared with prenotification (75.2%), (difference 7.95%, 95% CI 1.00% to 14.91%, P = 0.02). CONCLUSION SMS as a prenotification device seems ineffective. For postnotification, the evidence is unclear.