A Novel Remote Patient and Medication Monitoring Solution to Improve Adherence and PerSiStence With IBD Therapy

The investigators hypothesize that use of a remote monitoring digital health system that supports medication taking and monitoring of symptoms will improve adherence, clinical outcomes, and decrease healthcare utilization compared to standard care in participants with inflammatory bowel disease (IBD) initiating oral or subcutaneous treatment. The investigators are conducting a 12-month multicenter, randomized, controlled trial to assess the feasibility and effectiveness of a remote monitoring digital health system on adherence, clinical outcomes, and healthcare utilization. Adult participants with IBD with regular access to a mobile device or tablet initiating therapy with an oral or subcutaneous treatment for IBD at one of the five sites will be eligible to participate. Participants will be randomized 2:1 to the intervention or standard care. Eligible participants will complete an informed consent and baseline survey gathering demographic and clinical information. TapptTM digital health system (developed by Synchronyx) is a remote therapeutic monitoring and digital engagement solution that monitors real-time medication adherence patterns through smart label technologies, capture patient reported outcomes (PROs) and barriers to care, and process patient data through algorithms that trigger personalized digital and human touchpoints between clinical visits. The research team will input information into the system on the intervention participant's medication to be tracked, dose, and frequency of dosing. Participants in the intervention group will be shipped the smart labels to be affixed to the pill bottle, pen, or syringe of the newly prescribed medication. Prior to receiving their medication, participants will receive virtual training on how to attach and use the proprietary labels, set up their participant profile, use the participant-facing web app, and seek technical helpdesk support. At the time of a medication dose, participants will scan the label by tapping it with their mobile device to verify that they are taking the medication. Upon scanning, participants immediately will receive a notification on their device indicating that the label was successfully scanned, and that their medication adherence was updated in their profile. The app offers participants visibility into their medication adherence patterns and upcoming doses, as well as the opportunity to respond to in-app questions that capture barriers towards adherence, IBD symptoms, and patient report outcomes (PROs). This information will also be available to the research and clinical team in real-time via the provider dashboard. Most importantly, the dashboard's artificial intelligence-based algorithm will send email alerts to the clinical team if participant's adherence, symptoms, or PROs fall below predetermined thresholds. For oral medications, mean adherence \<86% in a 2-week period will trigger an alert. For SC medications, an alert will be generated if a dose is 10 days late for administration. A PRO 2 score for UC or CD of 2 or more will trigger an alert. If alerts are triggered for non-adherence, a clinical nurse, pharmacist, or social worker will contact the participant to identify barriers to adherence; remediation will be initiated if possible. The primary outcome of the proposed study will be the difference in mean medication possession ratio (MPR) between the intervention and control group during the 12-month study. Secondary outcomes will include self-reported adherence (MARS-5), clinical response and remission (Harvey Bradshaw Index for CD and partial Mayo score of UC), steroid-free response and remission, PROMIS measures of Fatigue, Sleep Disturbance, Pain Interference, Anxiety, Depression, and Quality of Life, self-efficacy (IBD Self-Efficacy Scale), new steroid use, and health care utilization (urgent care or emergency room visit, unplanned office visit, hospitalization, and/or surgery). Assuming 2 intervention participants for every 1 control with an adherence rate among controls of 0.65 and 0.9 in intervention participations with a Type 1 error rate of 0.05 and power of 0.9, we will need to enroll 82 intervention participants and 41 controls (n=123). All analyses will be completed using intention to treat principles. For categorical variables, the groups will be compared using the Chi Square test (Fisher's Exact if not normally distributed). For continuous variables, the groups will be compared using t tests (Wilcoxon signed rank if not normally distributed). We will also build logistic and linear regression models to adjust for confounding variable for the outcomes of interest. Possible confounding variables include but are not limited to route of administration, gender, insurance type, disease type, age, concurrent psychiatric disease, smoking, and concurrent steroid and/or narcotic use.