REmote moBile Outpatient mOnitoring in Transplant (Reboot) 2.0

Mobile health technologies such as smartphones and wearable devices can remotely monitor health. These technologies hold promise to improve health outcomes in a spectrum of patients by providing health care teams with better connectivity which may prompt more timely responses to questions and improvements to care.

The purpose of this study is to evaluate if solid organ transplant (SOT) recipients benefit from improved monitoring and removal of communication barriers as the most common reasons for readmission and mortality may be mitigated by clinical intervention. Additionally, medication adherence is critical in transplant patients to prevent graft rejection. We anticipate that remote monitoring will improve medication adherence/adjustments, and will allow for identification of early health issues, reducing preventable hospital readmissions. Thus, this study will determine if an innovative mobile health intervention, designed to improve patient-clinician communication, reduces unnecessary hospital readmission and visits to the emergency department and transplant clinic when utilized in addition to the standard of care telephone communication system. We will also incorporate clinical and continuous ambulatory physiologic data collected as part of the mobile health intervention to develop machine learning algorithms capable of identifying early indicators of adverse outcomes in adult heart, kidney, and liver transplant patients.

We hypothesize that: the delivery of personalized communication using a mobile health application will improve patient self-management resulting in a 50% reduction in preventable hospital readmission, and unscheduled visits to the emergency department and transplant clinic. With tailored communication through the mobile health application, we expect fewer standard of care phone messages for patients in the intervention group and patients with higher activity levels (average daily step-count) pre-transplantation will have lower index hospitalization length of stay. Finally, the large dataset collected from this study will allow novel machine learning-derived risk prediction models to more accurately predict adverse outcomes (e.g., organ rejection, infection, and death), compared to conventional regression models.