UBC Theses and Dissertations
Examining patient’s mobile phone access and planning a virtual care intervention using mHealth and conversation analytics Ghaseminejad-Tafreshi, Niloufar
Introduction: Unplanned hospital readmissions create stress for patients and their families while placing individuals at risk for negative outcomes and increasing healthcare system costs. Development of effective interventions to reduce readmissions involves timely discharge planning, transitional care, and stakeholder uptake. Mobile health (mHealth) and machine learning technology may help improve coordination of care, identify the underlying reasons for complications, and potentially reduce readmissions. Methods: To determine whether mHealth can help streamline and improve transitional care after discharge from the hospital, we will utilize a two-way text messaging virtual care platform to be piloted at the medical wards in the Vancouver General Hospital (VGH) Clinical Teaching Unit (CTU). Prior to launching the program, we conducted a survey of patients admitted to the CTU to determine mobile phone access, usage, and preferences to better understand the population we wish to serve. Using this information, we designed an mHealth intervention protocol that is patient-centered and collaborative. Results: We found that a two-way text messaging mHealth platform would likely be well-placed to facilitate better transitional care and to understand the underlying reasons for readmissions. Our survey results indicated that 86% of participants had access to a mobile phone, 63% of whom owned their own device and 23% of whom had access via a proxy (e.g., family or caregiver). These findings indicate that most patients can participate in mHealth interventions that rely on mobile phones and that engaging a proxy may further expand inclusivity. Lastly, we conducted training sessions and consulted with hospital staff to ensure the study protocol meets end-user needs and preferences. Using these findings, we developed a framework that utilizes natural language processing (NLP) and machine learning to analyze patient text message conversations with their health care provider (HCP). Conclusion: Our findings suggest that mHealth virtual care platforms are feasible and accessible in a hospital setting, which may help in reducing the burden of hospital readmission on patients, their families, and the health care system.
Item Citations and Data
Attribution-NonCommercial-NoDerivatives 4.0 International