Literature Collection

BACKGROUND: Integrating advanced machine-learning (ML) algorithms into clinical practice is challenging and requires interdisciplinary collaboration to develop transparent, interpretable, and ethically sound clinical decision support (CDS) tools. We aimed to design a ML-driven CDS tool to predict opioid overdose risk and gather feedback for its integration into the University of Florida Health (UFHealth) electronic health record (EHR) system. METHODS: We used user-centered design methods to integrate the ML algorithm into the EHR system. The backend and UI design sub-teams collaborated closely, both informed by user feedback sessions. We conducted seven user feedback sessions with five UF Health primary care physicians (PCPs) to explore aspects of CDS tools, including workflow, risk display, and risk mitigation strategies. After customizing the tool based on PCPs' feedback, we held two rounds of one-on-one usability testing sessions with 8 additional PCPs to gather feedback on prototype alerts. These sessions informed iterative UI design and backend processes, including alert frequency and reappearance circumstances. RESULTS: The backend process development identified needs and requirements from our team, information technology, UFHealth, and PCPs. Thirteen PCPs (male = 62%, White = 85%) participated across 7 user feedback sessions and 8 usability testing sessions. During the user feedback sessions, PCPs (n = 5) identified flaws such as the term "high risk" of overdose potentially leading to unintended consequences (e.g., immediate addiction services referrals), offered suggestions, and expressed trust in the tool. In the first usability testing session, PCPs (n = 4) emphasized the need for natural risk presentation (e.g., 1 in 200) and suggested displaying the alert multiple times yearly for at-risk patients. Another 4 PCPs in the second usability testing session valued the UFHealth-specific alert for managing new or unfamiliar patients, expressed concerns about PCPs' workload when prescribing to high-risk patients, and recommended incorporating the details page into training sessions to enhance usability. CONCLUSIONS: The final backend process for our CDS alert aligns with PCP needs and UFHealth standards. Integrating feedback from PCPs in the early development phase of our ML-driven CDS tool helped identify barriers and facilitators in the CDS integration process. This collaborative approach yielded a refined prototype aimed at minimizing unintended consequences and enhancing usability.

Parkinson's Disease (PD) is a progressive neurodegenerative condition that requires timely intervention to manage symptoms and prevent deterioration. This study investigates the essential requirements for a speech-based decision support tool to monitor PD progression in a community-integrated care setting. While still at an early design stage, the envisioned tool may take the form of mobile or desktop software accessible to patients, carers, and professionals in home and clinical settings. A mixed-methods approach, including surveys (n=31), focus groups (n=12), interviews (n=11), and policy document analysis, was used to gather insights from health and social care staff in Wales. Four major themes emerged: essential technical requirements (reliability, ease of use), workforce needs (training, analytic transparency), patient considerations (preferences, privacy), and systemic integration (interoperability, funding). Findings highlight the potential of speech-based AI systems for early, objective detection of PD deterioration. However, clinician trust, digital literacy, and user-centered design remain critical for adoption. Co-design with people with PD (PwPD), carers, and staff is strongly recommended for future development and evaluation. This study contributes to the growing field of intelligent systems in digital health and decision support.

Electronic health care records offer big data to mine and analyze towards improving public health outcomes. The information extracted, specifically social network data, could help us understand the primary care referrals for patients experiencing alcohol use disorder and wield that knowledge to better inform the engagement of this patient population. Network exposure and affiliation exposure models are two metrics that can be utilized to analyze the influence of social networks. We developed a core software library that address the scalability issue of our previous work. Our library computed high volume, randomly generated network graphs that range from 500-10,000 nodes (~126,000-40 million edges). This C library can be integrated with our previous work to handle high volume network data. Future plans include providing support for variant network exposure models and interfaces towards big network data analytics.

Efforts to develop an individualized treatment rule (ITR) to optimize major depressive disorder (MDD) treatment with antidepressant medication (ADM), psychotherapy, or combined ADM-psychotherapy have been hampered by small samples, small predictor sets, and suboptimal analysis methods. Analyses of large administrative databases designed to approximate experiments followed iteratively by pragmatic trials hold promise for resolving these problems. The current report presents a proof-of-concept study using electronic health records (EHR) of n = 43,470 outpatients beginning MDD treatment in Veterans Health Administration Primary Care Mental Health Integration (PC-MHI) clinics, which offer access not only to ADMs but also psychotherapy and combined ADM-psychotherapy. EHR and geospatial databases were used to generate an extensive baseline predictor set (5,865 variables). The outcome was a composite measure of at least one serious negative event (suicide attempt, psychiatric emergency department visit, psychiatric hospitalization, suicide death) over the next 12 months. Best-practices methods were used to adjust for nonrandom treatment assignment and to estimate a preliminary ITR in a 70% training sample and to evaluate the ITR in the 30% test sample. Statistically significant aggregate variation was found in overall probability of the outcome related to baseline predictors (AU-ROC = 0.68, S.E. = 0.01), with test sample outcome prevalence of 32.6% among the 5% of patients having highest predicted risk compared to 7.1% in the remainder of the test sample. The ITR found that psychotherapy-only was the optimal treatment for 56.0% of patients (roughly 20% lower risk of the outcome than if receiving one of the other treatments) and that treatment type was unrelated to outcome risk among other patients. Change in aggregate treatment costs of implementing this ITR would be negligible, as 16.1% fewer patients would be prescribed ADMs and 2.9% more would receive psychotherapy. A pragmatic trial would be needed to confirm the accuracy of the ITR.

Patient-generated health data (PGHD) encompass health-related information created, recorded, and gathered by patients in their daily lives, and are distinct from data collected in clinical settings. PGHD can offer insight into patients' everyday health behaviors and conditions, supporting health management and clinical decision-making. The Veterans Health Administration (VHA) has developed a robust infrastructure to collect PGHD, including automatically collected data from digital sensors and patient-entered data. This effort is guided by comprehensive policy and strategy documents to ensure the secure storage and effective use of PGHD. This paper describes the development and implementation of an infrastructure to support PGHD within the VHA and highlights envisioned clinical and research uses of PGHD to advance health care for US veterans. The PGHD database was built to Fast Healthcare Interoperability Resources standards, facilitating secure data storage and exchange of PGHD. Clinical tools, such as the provider-facing dashboards, make PGHD accessible from the electronic health records. Research and evaluation efforts focus on evaluating PGHD's impact on patient engagement, clinical outcomes, and health care equity. The VHA's comprehensive PGHD infrastructure represents a significant advancement in personalized health care and patient engagement. The integration of PGHD into clinical practice can enhance shared decision-making and self-management, while research and evaluation efforts can address how to maximize the benefits of PGHD for veterans. The VHA's approach sets a benchmark for other US health care systems in leveraging PGHD to achieve the broad aims of enhancing stakeholder health care experiences, improving population health and health equity, and reducing costs.