Background: Engagement is essential for the effectiveness of digital behavior change interventions. Existing systematic reviews examining hypertension self-management interventions via mobile apps have primarily focused on intervention efficacy and app usability. Engagement in the prevention or management of hypertension is largely unknown. Objective: This systematic review explores the definition and role of engagement in hypertension-focused mobile health (mHealth) interventions, as well as how determinants of engagement (ie, tailoring and interactivity) have been implemented. Methods: A systematic review of mobile app interventions for hypertension self-management targeting adults, published from 2013 to 2020, was conducted. A total of 21 studies were included in this systematic review. Results: The engagement was defined or operationalized as a microlevel concept, operationalized as interaction with the interventions (ie, frequency of engagement, time or duration of engagement with the program, and intensity of engagement). For all 3 studies that tested the relationship, increased engagement was associated with better biomedical outcomes (eg, blood pressure change). Interactivity was limited in digital behavior change interventions, as only 7 studies provided 2-way communication between users and a health care professional, and 9 studies provided 1-way communication in possible critical conditions; that is, when abnormal blood pressure values were recorded, users or health care professionals were notified. The tailoring of interventions varied at different aspects, from the tailoring of intervention content (including goals, patient education, advice and feedback from health professionals, reminders, and motivational messages) to the tailoring of intervention dose and communication mode. Tailoring was carried out in a number of ways, considering patient characteristics such as goals, preferences, disease characteristics (eg, hypertension stage and medication list), disease self-management experience levels, medication adherence rate, and values and beliefs. Conclusions: Available studies support the importance of engagement in intervention effectiveness as well as the essential roles of patient factors in tailoring, interactivity, and engagement. A patient-centered engagement framework for hypertension self-management using mHealth technology is proposed here, with the intent of facilitating intervention design and disease self-management using mHealth technology.

Despite recent advances in cancer treatments that prolong patients' lives, treatment-induced cardiotoxicity (i.e., the various heart damages caused by cancer treatments) emerges as one major side effect. The clinical decision-making process of cardiotoxicity is challenging, as early symptoms may happen in non-clinical settings and are too subtle to be noticed until life-threatening events occur at a later stage; clinicians already have a high workload focusing on the cancer treatment, no additional effort to spare on the cardiotoxicity side effect. Our project starts with a participatory design study with 11 clinicians to understand their decision-making practices and their feedback on an initial design of an AI-based decision-support system. Based on their feedback, we then propose a multimodal AI system, CardioAI, that can integrate wearables data and voice assistant data to model a patient's cardiotoxicity risk to support clinicians' decision-making. We conclude our paper with a small-scale heuristic evaluation with four experts and the discussion of future design considerations.

The COVID-19 outbreak was declared a pandemic by the World Health Organization and claimed hundreds of thousands of lives in the U.S. However, the adoption of protective behaviors in response to COVID-19 varies among sociodemographic groups. To better understand the disparities in U.S. adults’ adoption of protective behaviors in relation to sociodemographic factors, we conducted a two-wave panel survey in the U.S. Although sociodemographic factors were not found to be significant predictors in lagged regression analysis, results from a cross-sectional analysis showed that the White, younger, and higher educated individuals were more likely to take protective behaviors. Furthermore, based on the structural equation modeling, indirect effects were observed. Specifically, online COVID-19 information seeking mediated the effects of age and education on performance of protective behaviors, and subjective norms and online COVID-19 information seeking were found to be significant serial mediators of these relationships. The study not only extends the health disparities and health information seeking literature, but also offers practical insights to health campaigns and interventions promoting protective behaviors during a pandemic and reducing health disparities.

Sepsis is an organ dysfunction caused by a deregulated immune response to an infection. Early sepsis prediction and identification allow for timely intervention, leading to improved clinical outcomes. Clinical calculators (e.g., the six-organ dysfunction assessment of SOFA in Figure 1) play a vital role in sepsis identification within clinicians' workflow, providing evidence-based risk assessments essential for sepsis diagnosis. However, artificial intelligence (AI) sepsis prediction models typically generate a single sepsis risk score without incorporating clinical calculators for assessing organ dysfunctions, making the models less convincing and transparent to clinicians. To bridge the gap, we propose to mimic clinicians' workflow with a novel framework SepsisCalc to integrate clinical calculators into the predictive model, yielding a clinically transparent and precise model for utilization in clinical settings. Practically, clinical calculators usually combine information from multiple component variables in Electronic Health Records (EHR), and might not be applicable when the variables are (partially) missing. We mitigate this issue by representing EHRs as temporal graphs and integrating a learning module to dynamically add the accurately estimated calculator to the graphs. Experimental results on real-world datasets show that the proposed model outperforms state-of-the-art methods on sepsis prediction tasks. Moreover, we developed a system to identify organ dysfunctions and potential sepsis risks, providing a human-AI interaction tool for deployment, which can help clinicians understand the prediction outputs and prepare timely interventions for the corresponding dysfunctions, paving the way for actionable clinical decision-making support for early intervention.

Background: Cardiotoxicity is a limitation of several cancer therapies and early recognition improves outcomes. Symptom-tracking mobile health (mHealth) apps are feasible and beneficial, but key elements for mHealth symptom-tracking to indicate early signs of cardiotoxicity are unknown. Objective: We explored considerations for the design of, and implementation into a large academic medical center, an mHealth symptom-tracking tool for early recognition of cardiotoxicity in patients with cancer after cancer therapy initiation. Methods: We conducted semistructured interviews of >50% of the providers (oncologists, cardio-oncologists, and radiation oncologists) who manage cancer treatment-related cardiotoxicity in the participating institution (n=11), and either interviews or co-design or both with 6 patients. Data were coded and analyzed using thematic analysis. Results: Providers indicated that there was no existing process to enable early recognition of cardiotoxicity and felt the app could reduce delays in diagnosis and lead to better patient outcomes. Signs and symptoms providers recommended for tracking included chest pain or tightness, shortness of breath, heart racing or palpitations, syncope, lightheadedness, edema, and excessive fatigue. Implementation barriers included determining who would receive symptom reports, ensuring all members of the patient’s care team (eg, oncologist, cardiologist, and primary care) were informed of the symptom reports and could collaborate on care plans, and how to best integrate the app data into the electronic health record. Patients (n=6, 100%) agreed that the app would be useful for enhanced symptom capture and education and indicated willingness to use it. Conclusions: Providers and patients agree that a patient-facing, cancer treatment-related cardiotoxicity symptom-tracking mHealth app would be beneficial. Additional studies evaluating the role of mHealth as a potential strategy for targeted early cardioprotective therapy initiation are needed.