Impact of an Electronic Health Record-Integrated Personal Health Record on Patient Participation in Health Care: Development and Randomized Controlled Trial of MyHealthKeeper.

BACKGROUND: Personal health record (PHR)-based health care management systems can improve patient engagement and data-driven medical diagnosis in a clinical setting. OBJECTIVE: The purpose of this study was (1) to demonstrate the development of an electronic health record (EHR)-tethered PHR app named MyHealthKeeper, which can retrieve data from a wearable device and deliver these data to a hospital EHR system, and (2) to study the effectiveness of a PHR data-driven clinical intervention with clinical trial results. METHODS: To improve the conventional EHR-tethered PHR, we ascertained clinicians' unmet needs regarding PHR functionality and the data frequently used in the field through a cocreation workshop. We incorporated the requirements into the system design and architecture of the MyHealthKeeper PHR module. We constructed the app and validated the effectiveness of the PHR module by conducting a 4-week clinical trial. We used a commercially available activity tracker (Misfit) to collect individual physical activity data, and developed the MyHealthKeeper mobile phone app to record participants' patterns of daily food intake and activity logs. We randomly assigned 80 participants to either the PHR-based intervention group (n=51) or the control group (n=29). All of the study participants completed a paper-based survey, a laboratory test, a physical examination, and an opinion interview. During the 4-week study period, we collected health-related mobile data, and study participants visited the outpatient clinic twice and received PHR-based clinical diagnosis and recommendations. RESULTS: A total of 68 participants (44 in the intervention group and 24 in the control group) completed the study. The PHR intervention group showed significantly higher weight loss than the control group (mean 1.4 kg, 95% CI 0.9-1.9; P<.001) at the final week (week 4). In addition, triglyceride levels were significantly lower by the end of the study period (mean 2.59 mmol/L, 95% CI 17.6-75.8; P=.002). CONCLUSIONS: We developed an innovative EHR-tethered PHR system that allowed clinicians and patients to share lifelog data. This study shows the effectiveness of a patient-managed and clinician-guided health tracker system and its potential to improve patient clinical profiles. TRIAL REGISTRATION: ClinicalTrials.gov NCT03200119; https://clinicaltrials.gov/ct2/show/NCT03200119 (Archived by WebCite at http://www.webcitation.org/6v01HaCdd).

Impact of Personal Health Records and Wearables on Health Outcomes and Patient Response: Three-Arm Randomized Controlled Trial.

BACKGROUND: Although using the technologies for a variety of chronic health conditions such as personal health record (PHR) is reported to be acceptable and useful, there is a lack of evidence on the associations between the use of the technologies and the change of health outcome and patients' response to a digital health app. OBJECTIVE: This study aimed to examine the impact of the use of PHR and wearables on health outcome improvement and sustained use of the health app that can be associated with patient engagement. METHODS: We developed an Android-based mobile phone app and used a wristband-type activity tracker (Samsung Charm) to collect data on health-related daily activities from individual patients. Dietary record, daily step counts, sleep log, subjective stress amount, blood pressure, and weight values were recorded. We conducted a prospective randomized clinical trial across 4 weeks on those diagnosed with obstructive sleep apnea (OSA) who had visited the outpatient clinic of Seoul National University Bundang Hospital. The trial randomly assigned 60 patients to 3 subgroups including 2 intervention groups: (1) mobile app and wearable device users (n=20), (2) mobile app-only users (n=20), and (3) controls (n=20). The primary outcome measure was weight change. Body weights before and after the trial were recorded and analyzed during clinic visits. Changes in OSA-related respiratory parameters such as respiratory disturbance, apnea-hypopnea, and oxygenation desaturation indexes and snoring comprised the secondary outcome and were analyzed for each participant. RESULTS: We collected the individual data for each group during the trial, specifically anthropometric measurement and laboratory test results for health outcomes, and the app usage logs for patient response were collected and analyzed. The body weight showed a significant reduction in the 2 intervention groups after intervention, and the mobile app-only group showed more weight loss compared with the controls (P=.01). There were no significant changes in sleep-related health outcomes. From a patient response point of view, the average daily step counts (8165 steps) from the app plus wearable group were significantly higher than those (6034 steps) from the app-only group because they collected step count data from different devices (P=.02). The average rate of data collection was not different in physical activity (P=.99), food intake (P=.98), sleep (P=.95), stress (P=.70), and weight (P=.90) in the app plus wearable and app-only groups, respectively. CONCLUSIONS: We tried to integrate PHR data that allow clinicians and patients to share lifelog data with the clinical workflow to support lifestyle interventions. Our results suggest that a PHR-based intervention may be successful in losing body weight and improvement in lifestyle behavior. TRIAL REGISTRATION: ClinicalTrials.gov NCT03200223; https://clinicaltrials.gov/ct2/show/NCT03200223 (Archived by WebCite at http://www.webcitation.org/74baZmnCX).

Ultrasound image analysis using deep learning algorithm for the diagnosis of thyroid nodules.

Fine needle aspiration (FNA) is the procedure of choice for evaluating thyroid nodules. It is indicated for nodules >2 cm, even in cases of very low suspicion of malignancy. FNA has associated risks and expenses. In this study, we developed an image analysis model using a deep learning algorithm and evaluated if the algorithm could predict thyroid nodules with benign FNA results.Ultrasonographic images of thyroid nodules with cytologic or histologic results were retrospectively collected. For algorithm training, 1358 (670 benign, 688 malignant) thyroid nodule images were input into the Inception-V3 network model. The model was pretrained to classify nodules as benign or malignant using the ImageNet database. The diagnostic performance of the algorithm was tested with the prospectively collected internal (n = 55) and external test sets (n = 100).For the internal test set, 20 of the 21 FNA malignant nodules were correctly classified as malignant by the algorithm (sensitivity, 95.2%); and of the 22 nodules algorithm classified as benign, 21 were FNA benign (negative predictive value [NPV], 95.5%). For the external test set, 47 of the 50 FNA malignant nodules were correctly classified by the algorithm (sensitivity, 94.0%); and of the 31 nodules the algorithm classified as benign, 28 were FNA benign (NPV, 90.3%).The sensitivity and NPV of the deep learning algorithm shown in this study are promising. Artificial intelligence may assist clinicians to recognize nodules that are likely to be benign and avoid unnecessary FNA.