Search

BackgroundPatients with poorly controlled type 2 diabetes (T2D) experience increased morbidity, increased mortality, and higher cost of care. Self-monitoring of blood glucose (SMBG) is a critical component of diabetes self-management with established diabetes outcome benefits. Technological advancements in blood glucose meters, including cellular-connected devices that automatically upload SMBG data to secure cloud-based databases, allow for improved sharing and monitoring of SMBG data. Real-time monitoring of SMBG data presents opportunities to provide timely support to patients that is responsive to abnormal SMBG recordings. Such diabetes remote monitoring programs can provide patients with poorly controlled T2D additional support needed to improve critical outcomes. ObjectiveTo evaluate 6 months of a diabetes remote monitoring program facilitated by cellular-connected glucose meter, access to a diabetes coach, and support responsive to abnormal blood glucose recordings greater than 400 mg/dL or below 50 mg/dL in adults with poorly controlled T2D. MethodsPatients (N=119) receiving care at a diabetes center of excellence participated in a two-arm, 12-month randomized crossover study. The intervention included a cellular-connected glucose meter and phone-based diabetes coaching provided by Livongo Health. The coach answered questions, assisted in goal setting, and provided support in response to abnormal glucose levels. One group received the intervention for 6 months before returning to usual care (IV/UC). The other group received usual care before enrolling in the intervention (UC/IV) for 6 months. Change in hemoglobin A1c (HbA1c) was the primary outcome, and change in treatment satisfaction was the secondary outcome. ResultsImprovements in mean HbA1c were seen in both groups during the first 6 months (IV/UC −1.1%, SD 1.5 vs UC/IV −0.8%, SD 1.5; P

BackgroundAtrial fibrillation (AF) is the most common arrhythmia during critical illness, representing a sepsis-defining cardiac dysfunction associated with adverse outcomes. Large burdens of premature beats and noisy signal during sepsis may pose unique challenges to automated AF detection. ObjectiveThe objective of this study is to develop and validate an automated algorithm to accurately identify AF within electronic health care data among critically ill patients with sepsis. MethodsThis is a retrospective cohort study of patients hospitalized with sepsis identified from Medical Information Mart for Intensive Care (MIMIC III) electronic health data with linked electrocardiographic (ECG) telemetry waveforms. Within 3 separate cohorts of 50 patients, we iteratively developed and validated an automated algorithm that identifies ECG signals, removes noise, and identifies irregular rhythm and premature beats in order to identify AF. We compared the automated algorithm to current methods of AF identification in large databases, including ICD-9 (International Classification of Diseases, 9th edition) codes and hourly nurse annotation of heart rhythm. Methods of AF identification were tested against gold-standard manual ECG review. ResultsAF detection algorithms that did not differentiate AF from premature atrial and ventricular beats performed modestly, with 76% (95% CI 61%-87%) accuracy. Performance improved (P=.02) with the addition of premature beat detection (validation set accuracy: 94% [95% CI 83%-99%]). Median time between automated and manual detection of AF onset was 30 minutes (25th-75th percentile 0-208 minutes). The accuracy of ICD-9 codes (68%; P=.002 vs automated algorithm) and nurse charting (80%; P=.02 vs algorithm) was lower than that of the automated algorithm. ConclusionsAn automated algorithm using telemetry ECG data can feasibly and accurately detect AF among critically ill patients with sepsis, and represents an improvement in AF detection within large databases.

BackgroundeCohort studies offer an efficient approach for data collection. However, eCohort studies are challenged by volunteer bias and low adherence. We designed an eCohort embedded in the Framingham Heart Study (eFHS) to address these challenges and to compare the digital data to traditional data collection. ObjectiveThe aim of this study was to evaluate adherence of the eFHS app-based surveys deployed at baseline (time of enrollment in the eCohort) and every 3 months up to 1 year, and to compare baseline digital surveys with surveys collected at the research center. MethodsWe defined adherence rates as the proportion of participants who completed at least one survey at a given 3-month period and computed adherence rates for each 3-month period. To evaluate agreement, we compared several baseline measures obtained in the eFHS app survey to those obtained at the in-person research center exam using the concordance correlation coefficient (CCC). ResultsAmong the 1948 eFHS participants (mean age 53, SD 9 years; 57% women), we found high adherence to baseline surveys (89%) and a decrease in adherence over time (58% at 3 months, 52% at 6 months, 41% at 9 months, and 40% at 12 months). eFHS participants who returned surveys were more likely to be women (adjusted odds ratio [aOR] 1.58, 95% CI 1.18-2.11) and less likely to be smokers (aOR 0.53, 95% CI 0.32-0.90). Compared to in-person exam data, we observed moderate agreement for baseline app-based surveys of the Physical Activity Index (mean difference 2.27, CCC=0.56), and high agreement for average drinks per week (mean difference 0.54, CCC=0.82) and depressive symptoms scores (mean difference 0.03, CCC=0.77). ConclusionsWe observed that eFHS participants had a high survey return at baseline and each 3-month survey period over the 12 months of follow up. We observed moderate to high agreement between digital and research center measures for several types of surveys, including physical activity, depressive symptoms, and alcohol use. Thus, this digital data collection mechanism is a promising tool to collect data related to cardiovascular disease and its risk factors.

BackgroundAccumulation of excess body fluid and autonomic dysregulation are clinically important characteristics of acute decompensated heart failure. We hypothesized that transthoracic bioimpedance, a noninvasive, simple method for measuring fluid retention in lungs, and heart rate variability, an assessment of autonomic function, can be used for detection of fluid accumulation in patients with acute decompensated heart failure. ObjectiveWe aimed to evaluate the performance of transthoracic bioimpedance and heart rate variability parameters obtained using a fluid accumulation vest with carbon black–polydimethylsiloxane dry electrodes in a prospective clinical study (System for Heart Failure Identification Using an External Lung Fluid Device; SHIELD). MethodsWe computed 15 parameters: 8 were calculated from the model to fit Cole-Cole plots from transthoracic bioimpedance measurements (extracellular, intracellular, intracellular-extracellular difference, and intracellular-extracellular parallel circuit resistances as well as fitting error, resonance frequency, tissue heterogeneity, and cellular membrane capacitance), and 7 were based on linear (mean heart rate, low-frequency components of heart rate variability, high-frequency components of heart rate variability, normalized low-frequency components of heart rate variability, normalized high-frequency components of heart rate variability) and nonlinear (principal dynamic mode index of sympathetic function, and principal dynamic mode index of parasympathetic function) analysis of heart rate variability. We compared the values of these parameters between 3 participant data sets: control (n=32, patients who did not have heart failure), baseline (n=23, patients with acute decompensated heart failure taken at the time of admittance to the hospital), and discharge (n=17, patients with acute decompensated heart failure taken at the time of discharge from hospital). We used several machine learning approaches to classify participants with fluid accumulation (baseline) and without fluid accumulation (control and discharge), termed with fluid and without fluid groups, respectively. ResultsAmong the 15 parameters, 3 transthoracic bioimpedance (extracellular resistance, R0; difference in extracellular-intracellular resistance, R0 – R∞, and tissue heterogeneity, α) and 3 heart rate variability (high-frequency, normalized low-frequency, and normalized high-frequency components) parameters were found to be the most discriminatory between groups (patients with and patients without heart failure). R0 and R0 – R∞ had significantly lower values for patients with heart failure than for those without heart failure (R0: P=.006; R0 – R∞: P=.001), indicating that a higher volume of fluids accumulated in the lungs of patients with heart failure. A cubic support vector machine model using the 5 parameters achieved an accuracy of 92% for with fluid and without fluid group classification. The transthoracic bioimpedance parameters were related to intra- and extracellular fluid, whereas the heart rate variability parameters were mostly related to sympathetic activation. ConclusionsThis is useful, for instance, for an in-home diagnostic wearable to detect fluid accumulation. Results suggest that fluid accumulation, and subsequently acute decompensated heart failure detection, could be performed using transthoracic bioimpedance and heart rate variability measurements acquired with a wearable vest.

BackgroundScalable and accurate health outcome prediction using electronic health record (EHR) data has gained much attention in research recently. Previous machine learning models mostly ignore relations between different types of clinical data (ie, laboratory components, International Classification of Diseases codes, and medications). ObjectiveThis study aimed to model such relations and build predictive models using the EHR data from intensive care units. We developed innovative neural network models and compared them with the widely used logistic regression model and other state-of-the-art neural network models to predict the patient’s mortality using their longitudinal EHR data. MethodsWe built a set of neural network models that we collectively called as long short-term memory (LSTM) outcome prediction using comprehensive feature relations or in short, CLOUT. Our CLOUT models use a correlational neural network model to identify a latent space representation between different types of discrete clinical features during a patient’s encounter and integrate the latent representation into an LSTM-based predictive model framework. In addition, we designed an ablation experiment to identify risk factors from our CLOUT models. Using physicians’ input as the gold standard, we compared the risk factors identified by both CLOUT and logistic regression models. ResultsExperiments on the Medical Information Mart for Intensive Care-III dataset (selected patient population: 7537) show that CLOUT (area under the receiver operating characteristic curve=0.89) has surpassed logistic regression (0.82) and other baseline NN models (

BackgroundOnline support groups for atrial fibrillation (AF) and apps to detect and manage AF exist, but the scientific literature does not describe which patients are interested in digital disease support. ObjectiveThe objective of this study was to describe characteristics associated with Facebook use and interest in digital disease support among older patients with AF who used the internet. MethodsWe used baseline data from the Systematic Assessment of Geriatric Elements in Atrial Fibrillation (SAGE-AF), a prospective cohort of older adults (≥65 years) with AF at high stroke risk. Participants self-reported demographics, clinical characteristics, and Facebook and technology use. Online patients (internet use in the past 4 weeks) were asked whether they would be interested in participating in an online support AF community. Mobile users (owns smartphone and/or tablet) were asked about interest in communicating with their health care team about their AF-related health using a secure app. Logistic regression models identified crude and multivariable predictors of Facebook use and interest in digital disease support. ResultsOnline patients (N=816) were aged 74.2 (SD 6.6) years, 47.8% (390/816) were female, and 91.1% (743/816) were non-Hispanic white. Roughly half (52.5%; 428/816) used Facebook. Facebook use was more common among women (adjusted odds ratio [aOR] 2.21, 95% CI 1.66-2.95) and patients with mild to severe depressive symptoms (aOR 1.50, 95% CI 1.08-2.10) and less common among patients aged ≥85 years (aOR 0.27, 95% CI 0.15-0.48). Forty percent (40.4%; 330/816) reported interest in an online AF patient community. Interest in an online AF patient community was more common among online patients with some college/trade school or Bachelors/graduate school (aOR 1.70, 95% CI 1.10-2.61 and aOR 1.82, 95% CI 1.13-2.92, respectively), obesity (aOR 1.65, 95% CI 1.08-2.52), online health information seeking at most weekly or multiple times per week (aOR 1.84, 95% CI 1.32-2.56 and aOR 2.78, 95% CI 1.86-4.16, respectively), and daily Facebook use (aOR 1.76, 95% CI 1.26-2.46). Among mobile users, 51.8% (324/626) reported interest in communicating with their health care team via a mobile app. Interest in app-mediated communication was less likely among women (aOR 0.48, 95% CI 0.34-0.68) and more common among online patients who had completed trade school/some college versus high school/General Educational Development (aOR 1.95, 95% CI 1.17-3.22), sought online health information at most weekly or multiple times per week (aOR 1.86, 95% CI 1.27-2.74 and aOR 2.24, 95% CI 1.39-3.62, respectively), and had health-related apps (aOR 3.92, 95% CI 2.62-5.86). ConclusionsAmong older adults with AF who use the internet, technology use and demographics are associated with interest in digital disease support. Clinics and health care providers may wish to encourage patients to join an existing online support community for AF and explore opportunities for app-mediated patient-provider communication.

BackgroundNew electronic cohort (e-Cohort) study designs provide resource-effective methods for collecting participant data. It is unclear if implementing an e-Cohort study without direct, in-person participant contact can achieve successful participation rates. ObjectiveThe objective of this study was to compare 2 distinct enrollment methods for setting up mobile health (mHealth) devices and to assess the ongoing adherence to device use in an e-Cohort pilot study. MethodsWe coenrolled participants from the Framingham Heart Study (FHS) into the FHS–Health eHeart (HeH) pilot study, a digital cohort with infrastructure for collecting mHealth data. FHS participants who had an email address and smartphone were randomized to our FHS-HeH pilot study into 1 of 2 study arms: remote versus on-site support. We oversampled older adults (age ≥65 years), with a target of enrolling 20% of our sample as older adults. In the remote arm, participants received an email containing a link to enrollment website and, upon enrollment, were sent 4 smartphone-connectable sensor devices. Participants in the on-site arm were invited to visit an in-person FHS facility and were provided in-person support for enrollment and connecting the devices. Device data were tracked for at least 5 months. ResultsCompared with the individuals who declined, individuals who consented to our pilot study (on-site, n=101; remote, n=93) were more likely to be women, highly educated, and younger. In the on-site arm, the connection and initial use of devices was ≥20% higher than the remote arm (mean percent difference was 25% [95% CI 17-35] for activity monitor, 22% [95% CI 12-32] for blood pressure cuff, 20% [95% CI 10-30] for scale, and 43% [95% CI 30-55] for electrocardiogram), with device connection rates in the on-site arm of 99%, 95%, 95%, and 84%. Once connected, continued device use over the 5-month study period was similar between the study arms. ConclusionsOur pilot study demonstrated that the deployment of mobile devices among middle-aged and older adults in the context of an on-site clinic visit was associated with higher initial rates of device use as compared with offering only remote support. Once connected, the device use was similar in both groups.

The technological evolution and widespread availability of wearables and handheld ECG devices capable of screening for atrial fibrillation (AF), and their promotion directly to consumers, has focused attention of health care professionals and patient organizations on consumer-led AF screening. In this Frontiers review, members of the AF-SCREEN International Collaboration provide a critical appraisal of this rapidly evolving field to increase awareness of the complexities and uncertainties surrounding consumer-led AF screening. Although there are numerous commercially available devices directly marketed to consumers for AF monitoring and identification of unrecognized AF, health care professional-led randomized controlled studies using multiple ECG recordings or continuous ECG monitoring to detect AF have failed to demonstrate a significant reduction in stroke. Although it remains uncertain if consumer-led AF screening reduces stroke, it could increase early diagnosis of AF and facilitate an integrated approach, including appropriate anticoagulation, rate or rhythm management, and risk factor modification to reduce complications. Companies marketing AF screening devices should report the accuracy and performance of their products in high- and low-risk populations and avoid claims about clinical outcomes unless improvement is demonstrated in randomized clinical trials. Generally, the diagnostic yield of AF screening increases with the number, duration, and temporal dispersion of screening sessions, but the prognostic importance may be less than for AF detected by single-time point screening, which is largely permanent, persistent, or high-burden paroxysmal AF. Consumer-initiated ECG recordings suggesting possible AF always require confirmation by a health care professional experienced in ECG reading, whereas suspicion of AF on the basis of photoplethysmography must be confirmed with an ECG. Consumer-led AF screening is unlikely to be cost-effective for stroke prevention in the predominantly young, early adopters of this technology. Studies in older people at higher stroke risk are required to demonstrate both effectiveness and cost-effectiveness. The direct interaction between companies and consumers creates new regulatory gaps in relation to data privacy and the registration of consumer apps and devices. Although several barriers for optimal use of consumer-led screening exist, results of large, ongoing trials, powered to detect clinical outcomes, are required before health care professionals should support widespread adoption of consumer-led AF screening.

Atrial fibrillation (AF) is a major cause of morbidity and mortality globally, and much of this is driven by challenges in its timely diagnosis and treatment. Existing and emerging mobile technologies have been used to successfully identify AF in a variety of clinical and community settings, and while these technologies offer great promise for revolutionizing AF detection and screening, several major barriers may impede their effectiveness. The unclear clinical significance of device-detected AF, potential challenges in integrating patient-generated data into existing healthcare systems and clinical workflows, harm resulting from potential false positives, and identifying the appropriate scope of population-based screening efforts are all potential concerns that warrant further investigation. It is crucial for stakeholders such as healthcare providers, researchers, funding agencies, insurers, and engineers to actively work together in fulfilling the tremendous potential of mobile technologies to improve AF identification and management on a population level.