Detection of COVID-19 using multimodal data from a wearable device: results from the first TemPredict Study.

Early detection of diseases such as COVID-19 could be a critical tool in reducing disease transmission by helping individuals recognize when they should self-isolate, seek testing, and obtain early medical intervention. Consumer wearable devices that continuously measure physiological metrics hold promise as tools for early illness detection. We gathered daily questionnaire data and physiological data using a consumer wearable (Oura Ring) from 63,153 participants, of whom 704 self-reported possible COVID-19 disease. We selected 73 of these 704 participants with reliable confirmation of COVID-19 by PCR testing and high-quality physiological data for algorithm training to identify onset of COVID-19 using machine learning classification. The algorithm identified COVID-19 an average of 2.75 days before participants sought diagnostic testing with a sensitivity of 82% and specificity of 63%. The receiving operating characteristic (ROC) area under the curve (AUC) was 0.819 (95% CI [0.809, 0.830]). Including continuous temperature yielded an AUC 4.9% higher than without this feature. For further validation, we obtained SARS CoV-2 antibody in a subset of participants and identified 10 additional participants who self-reported COVID-19 disease with antibody confirmation. The algorithm had an overall ROC AUC of 0.819 (95% CI [0.809, 0.830]), with a sensitivity of 90% and specificity of 80% in these additional participants. Finally, we observed substantial variation in accuracy based on age and biological sex. Findings highlight the importance of including temperature assessment, using continuous physiological features for alignment, and including diverse populations in algorithm development to optimize accuracy in COVID-19 detection from wearables.

Inclusive innovation in telehealth.

It has been 30 years since the passage of the Americans with Disabilities Act and technological development has drastically changed the future for those with disabilities. As healthcare evolves toward promoting telehealth and patient-centered care, leaders must embrace persons with disabilities and caregivers as valued partners in design and implementation, not as passive "end-users". We call for a new era of inclusive innovation, a term proposed in this publication to describe accessible technological design for all. The next 30 years of the ADA leading to year 2050, should reflect a new era of access, whereby digital health surmounts geographic, social, and economic barriers toward an inclusive virtual society.

Tele-transitions of care (TTOC): a 12-month, randomized controlled trial evaluating the use of Telehealth to achieve triple aim objectives.

BACKGROUND: Poor transitions of care leads to increased health costs, over-utilization of emergency room departments, increased re-hospitalizations and causes poor patient experiences and outcomes. This study evaluated Telehealth feasibility in improving transitions of care. METHODS: This is a 12-month randomized controlled trial, evaluating the use of telehealth (remote patient monitoring and video visits) versus standard transitions of care with the primary outcomes of hospital readmission and emergency department utilization and secondary outcomes of access to care, medication management and adherence and patient engagement. Electronic Medical Record data, Health Information Exchange data and phone survey data was collected. Multi-variable logistic regression models were created to evaluate the effect of Telehealth on hospital readmission, emergency department utilization, medication adherence. Chi-square tests or Fisher's exact tests were used to compare the percentages of categorical variables between the Telehealth and control groups. T tests or Wilcoxon rank sum tests were used to compared means and medians between the two randomized groups. RESULTS: The study conducted between June 2017 and 2018, included 102 patients. Compared with the standard of care, Telehealth patients were more likely to have medicine reconciliation (p = 0.013) and were 7 times more likely to adhere to medication than the control group (p = 0.03). Telehealth patients exhibited enthusiasm (p = 0.0001), and confidence that Telehealth could improve their healthcare (p = 0.0001). Telehealth showed no statistical significance on emergency department utilization (p = 0.691) nor for readmissions (p = 0.31). 100% of Telehealth patients found the intervention to be valuable, 98% if given the opportunity, reported they would continue using telehealth to manage their healthcare needs, and 94% reported that the remote patient monitoring technology was useful. CONCLUSIONS: Telehealth can improve transitions of care after hospital discharge improving patient engagement and adherence to medications. Although this study was unable to show the effect of Telehealth on reduced healthcare utilization, more research needs to be done in order to understand the true impact of Telehealth on preventing avoidable hospital readmission and emergency department visits. TRIAL REGISTRATION: ClinicalTrials.Gov ID: NCT03528850 Date Registered (Retrospective): 5/18/2018. Status: Completed. IRB #: 970227.

Expanding Telehealth Competencies in Primary Care: A Longitudinal Interdisciplinary Simulation to Train Internal Medicine Residents in Complex Patient Care.

BACKGROUND: Despite increasing use of telehealth, there are limited published curricula training primary care providers in utilizing telehealth to deliver complex interdisciplinary care. OBJECTIVE: To describe and evaluate a telehealth curriculum with a longitudinal objective structured clinical examination (OSCE) to improve internal medicine residents' confidence and skills in coordinating complex interdisciplinary primary care via televisits, electronic consultation, and teleconferencing. METHODS: In 2019, 56 first- and third-year residents participated in a 3-part, 5-week OSCE training them to use telehealth to manage complex primary care. Learners conducted a standardized patient (SP) televisit in session 1, coordinated care via inter-visit e-messaging, and led a simulated interdisciplinary teleconference in session 2. Surveys measured confidence before session 1 (pre), post-session 1 (post-1), and post-session 2 (post-2). SP televisit checklists and investigators' assessment of e-messages evaluated residents' telehealth skills. RESULTS: Response rates were pre 100%, post-1 95% (53 of 56), and post-2 100%. Post-intervention, more residents were "confident/very confident" in adjusting their camera (33%, 95% CI 20-45 vs 85%, 95% CI 75-95, P < .0001), e-messaging (pre 36%, 95% CI 24-49 vs post-2 80%, 95% CI 70-91, P < .0001), and coordinating interdisciplinary care (pre 35%, 95% CI 22-47 vs post-2 84%, 95% CI 74-94, P < .0001). More residents were "likely/very likely" to use telemedicine in the future (pre 56%, 95% CI 43-69, vs post-2 79%, 95% CI 68-89, P = .001). CONCLUSIONS: A longitudinal, interdisciplinary telehealth simulation is feasible and can improve residents' confidence in using telemedicine to provide complex patient care.

Deep Learning on Electronic Health Records to Improve Disease Coding Accuracy.

Characterization of a patient's clinical phenotype is central to biomedical informatics. ICD codes, assigned to inpatient encounters by coders, is important for population health and cohort discovery when clinical information is limited. While ICD codes are assigned to patients by professionals trained and certified in coding there is substantial variability in coding. We present a methodology that uses deep learning methods to model coder decision making and that predicts ICD codes. Our approach predicts codes based on demographics, lab results, and medications, as well as codes from previous encounters. We are able to predict existing codes with high accuracy for all three of the test cases we investigated: diabetes, acute renal failure, and chronic kidney disease. We employed a panel of clinicians, in a blinded manner, to assess ground truth and compared the predictions of coders, model and clinicians. When disparities between the model prediction and coder assigned codes were reviewed, our model outperformed coder assigned ICD codes.

Tele-transitions of care. A 12-month, parallel-group, superiority randomized controlled trial protocol, evaluating the use of telehealth versus standard transitions of care in the prevention of avoidable hospital readmissions.

INTRODUCTION: Comprehensive transitions of care, reduce dangerous hospital readmissions. Telehealth offers promise, however few guidelines aid clinicians in introducing it in a feasible way while addressing the needs of a multi-comorbid population. Physician adoptability remains a significant barrier to the use of Telehealth due to data overload, concerns for disruptive workflows and uncertain practices. The methods proposed aid clinicians in implementing Telehealth training and research with limited resources to reach patients who need clinical surveillance most. This study introduces a new workflow for addressing tele-transitions of care, using risk stratification, remote patient monitoring, and patient-centered virtual visits. We propose a new communication tool which facilitates adoption. We take a clinically meaningful approach in assessing avoidable hospital readmissions, which can lead to further quality improvements and improved patient care. METHODS: This study design is a parallel-group, superiority, randomized controlled trial in which 180 patients are enrolled in the standard of care or Telehealth arms and evaluated for 30-days post hospitalization. The Telehealth group receives daily vitals surveillance with a "teledoc", a senior resident physician, who performs weekly virtual visits. The endpoint is 30-day hospital readmission. Patient data is collected on hospital utilization, patient self-management, physician and patient experience. DISCUSSION: Our protocol introduces a novel study design with existing clinical trainees, to provide comprehensive tele-transitions of care to reduce avoidable readmissions.