Supplemental text message support with the national diabetes prevention program: Pragmatic comparative effectiveness trial

Abstract

Background: The evidence-based National Diabetes Prevention Program (NDPP) is now widely disseminated, yet strategies to increase its effectiveness are needed, especially for underserved populations. The yearlong program promotes lifestyle changes for weight loss and can be offered in-person, online, via distance learning, or a combination of modalities. Less is known about which delivery features are optimal and may help address disparities in outcomes for subgroups. We previously demonstrated the efficacy of a stand-alone text messaging intervention based on the NDPP (SMS4PreDM) in a randomized controlled trial in a safety net health care system. Upon broader dissemination, we then showed that SMS4PreDM demonstrated high retention and modest weight loss at a relatively low cost, suggesting the potential to improve in-person NDPP delivery. Objective: In this study, we aim to compare the effectiveness of in-person NDPP classes with and without supplementary SMS4PreDM on attendance and weight loss outcomes to determine whether text messaging can enhance in-person NDPP delivery for a safety net patient population. Methods: From 2015 to 2017, patients with diabetes risks were identified primarily from provider referrals and enrolled in NDPP classes, SMS4PreDM, or both per their preference and availability. Participants naturally formed three groups: in-person NDPP with SMS4PreDM (n=236), in-person NDPP alone (n=252), and SMS4PreDM alone (n=285). This analysis compares the first two groups to evaluate whether supplemental text messaging may improve in-person NDPP outcomes. Outcomes for SMS4PreDM-only participants were previously reported. NDPP classes followed standard delivery guidelines, including weekly-to-monthly classes over a year. SMS4PreDM delivery included messages promoting lifestyle change and modest weight loss, sent 6 days per week for 12 months. Differences in characteristics between intervention groups were assessed using chi-square and t tests. Differences in NDPP attendance and weight loss outcomes were analyzed with multivariable linear and logistic regressions. Results: The mean age was 50.4 years (SD 13.9). Out of a total of 488 participants, 76.2% (n=372) were female and 59.0% (n=288) were Hispanic. An additional 17.2% (n=84) were non-Hispanic white and 12.9% (n=63) were non-Hispanic black. A total of 48.4% (n=236) of participants elected to receive supplemental text message support in addition to NDPP classes. Participants who chose supplemental text message support were on average 5.7 (SD 1.2) years younger (P