A Novel Patient Engagement Platform Using Accessible Text Messages and Calls (Epharmix): Feasibility Study

BACKGROUND: Patient noncompliance with therapy, treatments, and appointments represents a significant barrier to improving health care delivery and reducing the cost of care. One method to improve therapeutic adherence is to improve feedback loops in getting clinically acute events and issues to the relevant clinical providers as necessary (ranging from detecting hypoglycemic events for patients with diabetes to notifying the provider when patients are out of medications). Patients often don’t know which information should prompt a call to their physician and proactive checks by the clinics themselves can be very resource intensive. We hypothesized that a two-way SMS system combined with a platform web service for providers would enable both high patient engagement but also the ability to detect relevant clinical alerts. OBJECTIVE: The objectives of this study are to develop a feasible two-way automated SMS/phone call + web service platform for patient-provider communication, and then study the feasibility and acceptability of the Epharmix platform. First, we report utilization rates over the course of the first 18 months of operation including total identified clinically significant events, and second, review results of patient user-satisfaction surveys for interventions for patients with diabetes, COPD, congestive heart failure, hypertension, surgical site infections, and breastfeeding difficulties. METHODS: To test this question, we developed a web service + SMS/phone infrastructure (“Epharmix”). Utilization results were measured based on the total number of text messages or calls sent and received, with percentage engagement defined as a patient responding to a text message at least once in a given week, including the number of clinically significant alerts generated. User satisfaction surveys were sent once per month over the 18 months to measure satisfaction with the system, frequency and degree of communication. Descriptive statistics were used to describe the above information. RESULTS: In total, 28,386 text messages and 24,017 calls were sent to 929 patients over 9 months. Patients responded to 80% to 90% of messages allowing the system to detect 1164 clinically significant events. Patients reported increased satisfaction and communication with their provider. Epharmix increased the number of patient-provider interactions to over 10 on average in any given month for patients with diabetes, COPD, congestive heart failure, hypertension, surgical site infections, and breastfeeding difficulties. CONCLUSIONS: Engaging high-risk patients remains a difficult process that may be improved through novel, digital health interventions. The Epharmix platform enables increased patient engagement with very low risk to improve clinical outcomes. We demonstrated that engagement among high-risk populations is possible when health care comes conveniently to where they are.