Use of a text message program to raise type 2 diabetes risk awareness and promote health behavior change (part I): assessment of participant reach and adoption

Abstract

Background: There are an estimated 25.8 million American children and adults, equivalent to 8.3% of the US population, living with diabetes. Diabetes is particularly burdensome on minority populations. The use of mobile technologies for reaching broad populations is a promising approach, given its wide footprint and ability to deliver inexpensive personalized messages, to increase awareness of type 2 diabetes and promote behavior changes targeting risk factors associated with type 2 diabetes. As a part of the Beacon Community Cooperative Agreement Program, txt4health, a public-facing mobile health information service, was launched in 3 Beacon Communities: the Southeast Michigan Beacon Community in Detroit, MI, the Greater Cincinnati Beacon Community in Cincinnati, OH, and the Crescent City Beacon Community in New Orleans, LA. Txt4health is a mobile health information service designed to help people understand their risk for type 2 diabetes and become more informed about the steps they can take to lead healthy lives.

Objective: The purpose of this investigation was to use the RE-AIM framework to document txt4health reach and adoption by focusing on enrollment and participant engagement in program pilots in Southeast Michigan and Greater Cincinnati.

Methods: We conducted a retrospective records analysis of individual-level txt4health system data from participants in Southeast Michigan and Greater Cincinnati to determine participant usage of txt4health and engagement with the program.

Results: Results from the retrospective records analysis revealed that 5570 participants initiated the 2-step enrollment process via 1 of 3 enrollment strategies: text message, website, or directly with Beacon staff who signed participants up via the website. In total, 33.00% (1838/5570) of participants completed the 2-step enrollment process and were fully enrolled in the program. All participants (100.00%, 1620/1620) who enrolled via text message completed the entire 2-step enrollment process versus 5.52% (218/3950) of participants who enrolled via website or a Beacon staff member. Of those who fully enrolled, 71.00% (1305/1838) completed the diabetes risk assessment and 74.27% (1365/1838) set an initial weight loss goal. Overall, 39.06% (718/1838) of participants completed all 14 weeks of the program and 56.26% (1034/1838) dropped out before completing all 14 weeks, with the bulk of dropouts occurring in the first 4 weeks. Length of participation varied greatly, ranging from 0-48.7 weeks (median 8.6, mean 15.8, SD 15.8). Wide variability of participant engagement in regards to weekly weight and physical activity was documented.

Conclusions: Although broadly focused public health text message interventions may have the potential to reach large populations and show high levels of engagement among some users, the level of individual engagement among participants varies widely, suggesting that this type of approach may not be appropriate for all.

Keywords: cellular phone; diabetes mellitus, type 2; mobile health; program evaluation; risk reduction behavior; text messaging.

Figure 1

Participant flow through txt4health.

Figure 2

Participant dropout (n=1034) by week out of total number of participants (N=1838).