Understanding mHealth Engagement Among Patients With 30-Day Hospital Revisits: Secondary Analysis of a Randomized Clinical Trial

Background Reducing 30-day hospital readmissions has been a long-standing goal across health systems in the United States. While nurse-led phone outreach has been widely adopted to support transitional care, its reach is constrained by staffing and time limitations. Mobile health (mHealth) interventions, such as automated SMS text messaging and patient portals, offer scalable alternatives but have shown mixed effectiveness in reducing readmissions. Understanding how patients engage with mHealth after discharge may help optimize these tools for postdischarge care. Objective This study aimed to characterize patients in an mHealth transitional care program who experienced hospital revisits within 30 days of discharge, comparing demographic and clinical characteristics, intervisit interactions, and revisit features between those who engaged with mHealth and those who did not. Methods We conducted a secondary analysis of patients in the intervention arm of the Mobile Outreach to Reduce Emergencies-Primary Care randomized clinical trial. Participants received automated SMS text messages for 30 days after discharge alongside usual transitional care. We identified patients with a 30-day hospital revisit and conducted manual chart reviews to assess mHealth engagement and other forms of health care contact. We compared patient characteristics, intervisit interactions, and revisit features (time to revisit, relatedness to index hospitalization, and predictability of revisit) between mHealth users and nonusers. Results Among 496 patients with a 30-day revisit, 185 (37%) engaged with mHealth before their return. mHealth users were younger (n=47, 26% aged <50 years vs n=41, 14% among nonusers; P=.004) and more likely to have commercial insurance (n=43, 23% mHealth users vs n=35, 11% mHealth nonusers; P=.005). Revisits were more likely to be rated highly or somewhat predictable among mHealth users compared to nonusers (n=105, 56% vs n=152, 49%; P=.04), while relatedness to the index hospitalization was similar (n=98, 53% vs n=173, 55%; P=.09). mHealth users had a longer mean time to revisit than nonusers (15.2, SD 8.1 vs 11.3, SD 8.4 days; P<.001) and were more likely to contact their practices via telephone (n=100, 54% vs n=136, 44%; P=.03) or attend a clinic visit (n=112, 61% vs n=131, 42%; P<.001). Conclusions Among patients enrolled in an mHealth postdischarge program who experienced hospital revisits, fewer than half engaged with mHealth prior to their return. Revisits among mHealth users occurred later and were more predictable, suggesting that engagement may enhance situational awareness but not necessarily prevent revisits. Future work should focus on strategies to increase engagement across groups and integrate mHealth with existing transitional care infrastructure.