Text message link to online survey: a new highly effective method of longitudinal data collection

Abstract Objectives To evaluate response rate to a text message link for an online survey as a method of data collection over the course of a medication abortion. Study Design This is a secondary analysis of a randomized, double-blind, placebo-controlled trial of 110 women initiating a medication abortion up to 70 days gestation. We sent a text message containing a link to an online survey collecting pain scores, analgesic use, and adverse effects at six time points over 72 hours from time of misoprostol administration (0-, 2-, 6-, 12-, 24-, and 72-hours). Our primary outcome was the proportion of all text messages sent for which an online survey response was received. Secondary outcomes included the proportion of complete responses (all six surveys), responses received more than two hours after text prompt (defined as late), and differences in response rate by time of day. Results From June 2015 to October 2016, we screened 241 women, and 110 were randomized. We excluded three (1.2%) due to lack of a cellular phone with text capability, and three (2.7%) were lost to follow-up after enrollment. We received a response to 95.9% (633/660) of the delivered surveys, and 93.6% (103/110) of participants completed all six surveys. Over three-quarters of all responses were received within two hours of the requested time. Surveys sent before 08:00 were more likely to be returned late. Conclusions In this population of women seeking medication abortion, text message link to online survey response appears to be an effective mode of data collection. Implications Text message prompts for an online survey may optimize longitudinal real-time data collection response rates of sensitive data.