The Impact of Text Messaging Intervention as an In-Between Support to Diabetes Group Visits in Federally Qualified Health Centers: Cluster Randomized Control Study (Preprint)

BACKGROUND One in eleven people receive primary care from a federally qualified health center (FQHC) in the U.S. Text-messaging interventions (TMI) are accessible ways to deliver health information, engage patients, and improve health outcomes in the health center setting. OBJECTIVE We aimed to evaluate the impact of a TMI implemented with a group visit intervention among patients with type 2 diabetes (T2DM) at FQHCs on patient reported outcomes and clinical outcomes based on patient TMI engagement. METHODS A TMI was implemented for eleven health centers participating in a cluster randomized study of diabetes group visits (GV) in Midwestern FQHCs targeting adults with T2DM. FQHC patients participated in six monthly group visits either in-person or virtually and a concurrent 25-week TMI. Outcome measures included clinical markers such as hemoglobin A1c (HbA1c) and patient-reported diabetes distress, diabetes self-care, diabetes self-efficacy, diabetes care knowledge, diabetes quality of life, diabetes social support and TMI use and satisfaction. TMI response rate was calculated as responses to a text message requesting a response divided by total messages requesting a response sent. Patients were grouped as high responders if their response rate was ≥ the median response rate and low responders if their response rate was below the median. We conducted linear mixed models to compare high and low responders and/or within a group, adjusting for age, gender, GV attendance, and depression/anxiety at baseline. RESULTS 101 out of 124 GV patients enrolled in the TMI (81.5%). The average age of the population in the TMI was 53 years. Sixty percent of the cohort were racial/ethnic minorities, 53% had a high school diploma/GED or less, and 71% reported an annual income <$30,000. Seventy-nine percent owned a smartphone and 88% had an unlimited texting plan. The median response rate was 41% and the mean response rate was 41.6%. Adjusted models showed significantly improved diabetes knowledge (P < .001), foot care (P < .001), and exercise (P = .002) in high responders compared to low responders at six months. No group difference was found in HbA1c. Within the high responders group, diabetes distress (P = .001), diabetes social support (P < .001), diabetes quality of life (P < .001), diabetes care knowledge (P < .001), foot care (P < .001), and diet (P = .003) improved from baseline to six months. Low responders only improved in diabetes quality of life (P = .003) from baseline to six months CONCLUSIONS In a safety-net population receiving care at FQHCs participating in a combined text messaging and group visit intervention, our study showed improved diabetes distress, social support, knowledge, self-care, self-efficacy, and quality of life among patients highly engaged in the text-messaging program. CLINICALTRIAL ClincialTrials.gov NCT03487692