Understanding the Role of Virtual Outreach and Programming for LGBT Individuals in Later Life.

Due to health disparities LGBT older adults may have more health care needs, but they are likely to have less informal sources of support. While efforts have been made to serve LGBT older adults, traditional forms of in person outreach and service may still be inaccessible to those living in rural areas, with restricted mobility, due to lack of transportation, during inclement weather, or in public health situations as the Covid-19 pandemic. We conducted focus group discussions to understand the role of virtual outreach in serving LGBT individuals' needs in their later years of life. Study participants expressed a desire for dating, community, aging in place, and affirming health care. However, their experience of internalized and institutional homophobia and ageism may act as barriers in fulfilling those needs. A dedicated virtual space has the potential to overcome these barriers by facilitating online get-togethers, support groups, dating events, having coming out resources, and exchanging information on LGBT friendly health services. Having a space to express their generativity may make such virtual services more empowering. Lack of technological access and privacy concerns may hinder the use of virtual services but can be overcome with training and education.

Corrigendum and Editorial Warning Regarding Use of the MMAS-8 Scale (A Remote Medication Monitoring System for Chronic Heart Failure Patients to Reduce Readmissions: A Two-Arm Randomized Pilot Study).

[This corrects the article DOI: .].

A machine learning model to predict the risk of 30-day readmissions in patients with heart failure: a retrospective analysis of electronic medical records data.

BACKGROUND: Heart failure is one of the leading causes of hospitalization in the United States. Advances in big data solutions allow for storage, management, and mining of large volumes of structured and semi-structured data, such as complex healthcare data. Applying these advances to complex healthcare data has led to the development of risk prediction models to help identify patients who would benefit most from disease management programs in an effort to reduce readmissions and healthcare cost, but the results of these efforts have been varied. The primary aim of this study was to develop a 30-day readmission risk prediction model for heart failure patients discharged from a hospital admission. METHODS: We used longitudinal electronic medical record data of heart failure patients admitted within a large healthcare system. Feature vectors included structured demographic, utilization, and clinical data, as well as selected extracts of un-structured data from clinician-authored notes. The risk prediction model was developed using deep unified networks (DUNs), a new mesh-like network structure of deep learning designed to avoid over-fitting. The model was validated with 10-fold cross-validation and results compared to models based on logistic regression, gradient boosting, and maxout networks. Overall model performance was assessed using concordance statistic. We also selected a discrimination threshold based on maximum projected cost saving to the Partners Healthcare system. RESULTS: Data from 11,510 patients with 27,334 admissions and 6369 30-day readmissions were used to train the model. After data processing, the final model included 3512 variables. The DUNs model had the best performance after 10-fold cross-validation. AUCs for prediction models were 0.664 ± 0.015, 0.650 ± 0.011, 0.695 ± 0.016 and 0.705 ± 0.015 for logistic regression, gradient boosting, maxout networks, and DUNs respectively. The DUNs model had an accuracy of 76.4% at the classification threshold that corresponded with maximum cost saving to the hospital. CONCLUSIONS: Deep learning techniques performed better than other traditional techniques in developing this EMR-based prediction model for 30-day readmissions in heart failure patients. Such models can be used to identify heart failure patients with impending hospitalization, enabling care teams to target interventions at their most high-risk patients and improving overall clinical outcomes.

Healthcare utilization in older patients using personal emergency response systems: an analysis of electronic health records and medical alert data : Brief Description: A Longitudinal Retrospective Analyses of healthcare utilization rates in older patients using Personal Emergency Response Systems from 2011 to 2015.

BACKGROUND: Personal Emergency Response Systems (PERS) are traditionally used as fall alert systems for older adults, a population that contributes an overwhelming proportion of healthcare costs in the United States. Previous studies focused mainly on qualitative evaluations of PERS without a longitudinal quantitative evaluation of healthcare utilization in users. To address this gap and better understand the needs of older patients on PERS, we analyzed longitudinal healthcare utilization trends in patients using PERS through the home care management service of a large healthcare organization. METHODS: Retrospective, longitudinal analyses of healthcare and PERS utilization records of older patients over a 5-years period from 2011-2015. The primary outcome was to characterize the healthcare utilization of PERS patients. This outcome was assessed by 30-, 90-, and 180-day readmission rates, frequency of principal admitting diagnoses, and prevalence of conditions leading to potentially avoidable admissions based on Centers for Medicare and Medicaid Services classification criteria. RESULTS: The overall 30-day readmission rate was 14.2%, 90-days readmission rate was 34.4%, and 180-days readmission rate was 42.2%. While 30-day readmission rates did not increase significantly (p = 0.16) over the study period, 90-days (p = 0.03) and 180-days (p = 0.04) readmission rates did increase significantly. The top 5 most frequent principal diagnoses for inpatient admissions included congestive heart failure (5.7%), chronic obstructive pulmonary disease (4.6%), dysrhythmias (4.3%), septicemia (4.1%), and pneumonia (4.1%). Additionally, 21% of all admissions were due to conditions leading to potentially avoidable admissions in either institutional or non-institutional settings (16% in institutional settings only). CONCLUSIONS: Chronic medical conditions account for the majority of healthcare utilization in older patients using PERS. Results suggest that PERS data combined with electronic medical records data can provide useful insights that can be used to improve health outcomes in older patients.

Telemedical Education: Training Digital Natives in Telemedicine.

Telemedicine plays an important role in the delivery of medical care, and will become increasingly prominent going forward. Current medical students are among the first generation of "digital natives" who are well versed in the incorporation of technology into social interaction. These students are well positioned to apply advances in communications to patient care. Even so, providers require training to effectively leverage these opportunities. Therefore, we recommend introducing telemedicine training into medical school curricula and propose a model for incorporation.

A Remote Medication Monitoring System for Chronic Heart Failure Patients to Reduce Readmissions: A Two-Arm Randomized Pilot Study.

BACKGROUND: Heart failure (HF) is a chronic condition affecting nearly 5.7 million Americans and is a leading cause of morbidity and mortality. With an aging population, the cost associated with managing HF is expected to more than double from US $31 billion in 2012 to US $70 billion by 2030. Readmission rates for HF patients are high-25% are readmitted at 30 days and nearly 50% at 6 months. Low medication adherence contributes to poor HF management and higher readmission rates. Remote telehealth monitoring programs aimed at improved medication management and adherence may improve HF management and reduce readmissions. OBJECTIVE: The primary goal of this randomized controlled pilot study is to compare the MedSentry remote medication monitoring system versus usual care in older HF adult patients who recently completed a HF telemonitoring program. We hypothesized that remote medication monitoring would be associated with fewer unplanned hospitalizations and emergency department (ED) visits, increased medication adherence, and improved health-related quality of life (HRQoL) compared to usual care. METHODS: Participants were randomized to usual care or use of the remote medication monitoring system for 90 days. Twenty-nine participants were enrolled and the final analytic sample consisted of 25 participants. Participants completed questionnaires at enrollment and closeout to gather data on medication adherence, health status, and HRQoL. Electronic medical records were reviewed for data on baseline classification of heart function and the number of unplanned hospitalizations and ED visits during the study period. RESULTS: Use of the medication monitoring system was associated with an 80% reduction in the risk of all-cause hospitalization and a significant decrease in the number of all-cause hospitalization length of stay in the intervention arm compared to usual care. Objective device data indicated high adherence rates (95%-99%) among intervention group participants despite finding no significant difference in self-reported adherence between study arms. The intervention group had poorer heart function and HRQoL at baseline, and HRQoL declined significantly in the intervention group compared to controls. CONCLUSIONS: The MedSentry medication monitoring system is a promising technology that merits continued development and evaluation. The MedSentry medication monitoring system may be useful both as a standalone system for patients with complex medication regimens or used to complement existing HF telemonitoring interventions. We found significant reductions in risk of all-cause hospitalization and the number of all-cause length of stay in the intervention group compared to controls. Although HRQoL deteriorated significantly in the intervention group, this may have been due to the poorer HF-functioning at baseline in the intervention group compared to controls. Telehealth medication adherence technologies, such as the MedSentry medication monitoring system, are a promising method to improve patient self-management,the quality of patient care, and reduce health care utilization and expenditure for patients with HF and other chronic diseases that require complex medication regimens. TRIAL REGISTRATION: ClinicalTrials.gov NCT01814696; https://clinicaltrials.gov/ct2/show/study/NCT01814696 (Archived by WebCite® at http://www.webcitation.org/6giqAVhno).

Text to Move: A Randomized Controlled Trial of a Text-Messaging Program to Improve Physical Activity Behaviors in Patients With Type 2 Diabetes Mellitus.

BACKGROUND: Text messages are increasingly being used because of the low cost and the ubiquitous nature of mobile phones to engage patients in self-care behaviors. Self-care is particularly important in achieving treatment outcomes in type 2 diabetes mellitus (T2DM). OBJECTIVE: This study examined the effect of personalized text messages on physical activity, as measured by a pedometer, and clinical outcomes in a diverse population of patients with T2DM. METHODS: Text to Move (TTM) incorporates physical activity monitoring and coaching to provide automated and personalized text messages to help patients with T2DM achieve their physical activity goals. A total of 126 English- or Spanish-speaking patients with glycated hemoglobin A1c (HbA1c) >7 were enrolled in-person to participate in the study for 6 months and were randomized into either the intervention arm that received the full complement of the intervention or a control arm that received only pedometers. The primary outcome was change in physical activity. We also assessed the effect of the intervention on HbA1c, weight, and participant engagement. RESULTS: All participants (intervention: n=64; control: n=62) were included in the analyses. The intervention group had significantly higher monthly step counts in the third (risk ratio [RR] 4.89, 95% CI 1.20 to 19.92, P=.03) and fourth (RR 6.88, 95% CI 1.21 to 39.00, P=.03) months of the study compared to the control group. However, over the 6-month follow-up period, monthly step counts did not differ statistically by group (intervention group: 9092 steps; control group: 3722 steps; RR 2.44, 95% CI 0.68 to 8.74, P=.17). HbA1c decreased by 0.07% (95% CI -0.47 to 0.34, P=.75) in the TTM group compared to the control group. Within groups, HbA1c decreased significantly from baseline in the TTM group by -0.43% (95% CI -0.75 to -0.12, P=.01), but nonsignificantly in the control group by -0.21% (95% CI -0.49 to 0.06, P=.13). Similar changes were observed for other secondary outcomes. CONCLUSION: Personalized text messaging can be used to improve outcomes in patients with T2DM by employing optimal patient engagement measures.

Personalized Telehealth in the Future: A Global Research Agenda.

As telehealth plays an even greater role in global health care delivery, it will be increasingly important to develop a strong evidence base of successful, innovative telehealth solutions that can lead to scalable and sustainable telehealth programs. This paper has two aims: (1) to describe the challenges of promoting telehealth implementation to advance adoption and (2) to present a global research agenda for personalized telehealth within chronic disease management. Using evidence from the United States and the European Union, this paper provides a global overview of the current state of telehealth services and benefits, presents fundamental principles that must be addressed to advance the status quo, and provides a framework for current and future research initiatives within telehealth for personalized care, treatment, and prevention. A broad, multinational research agenda can provide a uniform framework for identifying and rapidly replicating best practices, while concurrently fostering global collaboration in the development and rigorous testing of new and emerging telehealth technologies. In this paper, the members of the Transatlantic Telehealth Research Network offer a 12-point research agenda for future telehealth applications within chronic disease management.

Heart failure remote monitoring: evidence from the retrospective evaluation of a real-world remote monitoring program.

BACKGROUND: Given the magnitude of increasing heart failure mortality, multidisciplinary approaches, in the form of disease management programs and other integrative models of care, are recommended to optimize treatment outcomes. Remote monitoring, either as structured telephone support or telemonitoring or a combination of both, is fast becoming an integral part of many disease management programs. However, studies reporting on the evaluation of real-world heart failure remote monitoring programs are scarce. OBJECTIVE: This study aims to evaluate the effect of a heart failure telemonitoring program, Connected Cardiac Care Program (CCCP), on hospitalization and mortality in a retrospective database review of medical records of patients with heart failure receiving care at the Massachusetts General Hospital. METHODS: Patients enrolled in the CCCP heart failure monitoring program at the Massachusetts General Hospital were matched 1:1 with usual care patients. Control patients received care from similar clinical settings as CCCP patients and were identified from a large clinical data registry. The primary endpoint was all-cause mortality and hospitalizations assessed during the 4-month program duration. Secondary outcomes included hospitalization and mortality rates (obtained by following up on patients over an additional 8 months after program completion for a total duration of 1 year), risk for multiple hospitalizations and length of stay. The Cox proportional hazard model, stratified on the matched pairs, was used to assess primary outcomes. RESULTS: A total of 348 patients were included in the time-to-event analyses. The baseline rates of hospitalizations prior to program enrollment did not differ significantly by group. Compared with controls, hospitalization rates decreased within the first 30 days of program enrollment: hazard ratio (HR)=0.52, 95% CI 0.31-0.86, P=.01). The differential effect on hospitalization rates remained consistent until the end of the 4-month program (HR=0.74, 95% CI 0.54-1.02, P=.06). The program was also associated with lower mortality rates at the end of the 4-month program: relative risk (RR)=0.33, 95% 0.11-0.97, P=.04). Additional 8-months follow-up following program completion did not show residual beneficial effects of the CCCP program on mortality (HR=0.64, 95% 0.34-1.21, P=.17) or hospitalizations (HR=1.12, 95% 0.90-1.41, P=.31). CONCLUSIONS: CCCP was associated with significantly lower hospitalization rates up to 90 days and significantly lower mortality rates over 120 days of the program. However, these effects did not persist beyond the 120-day program duration.

"Friending" teens: systematic review of social media in adolescent and young adult health care.

BACKGROUND: Social media has emerged as a potentially powerful medium for communication with adolescents and young adults around their health choices. OBJECTIVE: The goal of this systematic review is to identify research on the use of social media for interacting with adolescents and young adults in order to achieve positive health outcomes. METHODS: A MEDLINE/PubMed electronic database search was performed between January 1, 2002 and October 1, 2013, using terms to identify peer-reviewed research in which social media and other Web 2.0 technologies were an important feature. We used a systematic approach to retrieve papers and extract relevant data. RESULTS: We identified 288 studies involving social media, of which 87 met criteria for inclusion; 75 studies were purely observational and 12 were interventional. The ways in which social media was leveraged by these studies included (1) observing adolescent and young adult behavior (n=77), (2) providing health information (n=13), (3) engaging the adolescent and young adult community (n=17), and (4) recruiting research participants (n=23). Common health topics addressed included high-risk sexual behaviors (n=23), alcohol, tobacco, and other drug use (n=19), Internet safety (n=8), mental health issues (n=18), medical conditions (n=11), or other specified issues (n=12). Several studies used more than one social media platform and addressed more than one health-related topic. CONCLUSIONS: Social media technologies offer an exciting new means for engaging and communicating with adolescents and young adults; it has been successfully used to engage this age group, identify behaviors, and provide appropriate intervention and education. Nevertheless, the majority of studies to date have been preliminary and limited in their methodologies, and mostly center around evaluating how adolescents and young adults use social media and the resulting implications on their health. Although these explorations are essential, further exploration and development of these strategies into building effective interventions is necessary.

The effect of technology-based interventions on pain, depression, and quality of life in patients with cancer: a systematic review of randomized controlled trials.

BACKGROUND: The burden of cancer is increasing; projections over the next 2 decades suggest that the annual cases of cancer will rise from 14 million in 2012 to 22 million. However, cancer patients in the 21st century are living longer due to the availability of novel therapeutic regimens, which has prompted a growing focus on maintaining patients' health-related quality of life. Telehealth is increasingly being used to connect with patients outside of traditional clinical settings, and early work has shown its importance in improving quality of life and other clinical outcomes in cancer care. OBJECTIVE: The aim of this study was to systematically assess the literature for the effect of supportive telehealth interventions on pain, depression, and quality of life in cancer patients via a systematic review of clinical trials. METHODS: We searched PubMed, EMBASE, Google Scholar, CINAHL, and PsycINFO in July 2013 and updated the literature search again in January 2015 for prospective randomized trials evaluating the effect of telehealth interventions in cancer care with pain, depression, and quality of life as main outcomes. Two of the authors independently reviewed and extracted data from eligible randomized controlled trials, based on pre-determined selection criteria. Methodological quality of studies was assessed by the Cochrane Collaboration risk of bias tool. RESULTS: Of the 4929 articles retrieved from databases and relevant bibliographies, a total of 20 RCTs were included in the final review. The studies were largely heterogeneous in the type and duration of the intervention as well as in outcome assessments. A majority of the studies were telephone-based interventions that remotely connected patients with their health care provider or health coach. The intervention times ranged from 1 week to 12 months. In general, most of the studies had low risk of bias across the domains of the Cochrane Collaboration risk of bias tool, but most of the studies had insufficient information about the allocation concealment domain. Two of the three studies focused on pain control reported significant effects of the intervention; four of the nine studies focus on depression reported significant effects, while only the studies that were focused on quality of life reported significant effects. CONCLUSIONS: This systematic review demonstrates the potential of telehealth interventions in improving outcomes in cancer care. However, more high-quality large-sized trials are needed to demonstrate cogent evidence of its effectiveness.

Representation of health conditions on Facebook: content analysis and evaluation of user engagement.

BACKGROUND: A sizable majority of adult Internet users report looking for health information online. Social networking sites (SNS) like Facebook represent a common place to seek information, but very little is known about the representation and use of health content on SNS. OBJECTIVE: Our goal in this study was to understand the role of SNS in health information seeking. More specifically, we aimed to describe how health conditions are represented on Facebook Pages and how users interact with these different conditions. METHODS: We used Google Insights to identify the 20 most searched for health conditions on Google and then searched each of the resulting terms on Facebook. We compiled a list of the first 50 Facebook "Pages" results for each health condition. After filtering results to identify pages relevant to our research, we categorized pages into one of seven categories based on the page's primary purpose. We then measured user engagement by evaluating the number of "Likes" for different conditions and types of pages. RESULTS: The search returned 50 pages for 18 of the health conditions, but only 48 pages were found for "anemia" and 5 pages were found for "flu symptoms", yielding a total of 953 pages. A large number of pages (29.4%, 280/953) were irrelevant to the health condition searched. Of the 673 relevant pages, 151 were not in English or originated outside the United States, leaving 522 pages to be coded for content. The most common type of page was marketing/promotion (32.2%, 168/522) followed by information/awareness (20.7%, 108/522), Wikipedia-type pages (15.5%, 81/522), patient support (9.4%, 49/522), and general support (3.6%, 19/522). Health conditions varied greatly by the primary page type. All health conditions had some marketing/promotion pages and this made up 76% (29/38) of pages on acquired immunodeficiency syndrome (AIDS). The largest percentage of general support pages were cancer (19%, 6/32) and stomach (16%, 4/25). For patient support, stroke (67%, 4/6), lupus (33%, 10/30), breast cancer (19%, 6/31), arthritis (16%, 6/36), and diabetes (16%, 6/37) ranked the highest. Six health conditions were not represented by any type of support pages (ie, human papillomavirus, diarrhea, flu symptoms, pneumonia, spine, human immunodeficiency virus). Marketing/promotion pages accounted for 46.73% (10,371,169/22,191,633) of all Likes, followed by support pages (40.66%, 9,023,234/22,191,633). Cancer and breast cancer accounted for 86.90% (19,284,066/22,191,633) of all page Likes. CONCLUSIONS: This research represents the first attempts to comprehensively describe publicly available health content and user engagement with health conditions on Facebook pages. Public health interventions using Facebook will need to be designed to ensure relevant information is easy to find and with an understanding that stigma associated with some health conditions may limit the users' engagement with Facebook pages. This line of research merits further investigation as Facebook and other SNS continue to evolve over the coming years.

A Digital Therapeutic Application (ePAL) to Manage Pain in Patients With Advanced Cancer: A Randomized Controlled Trial.

BACKGROUND: Patients with advanced cancer often experience immense cancer pain that negatively impacts their quality of life. Interventions to address cancer-related pain are limited. METHODS: We conducted a randomized trial of a digital therapeutic app (ePAL) for patients with advanced cancer receiving care in a specialty palliative care clinic at a tertiary care hospital. Patients were randomized to ePAL or usual care. ePAL included 1) active pain monitoring; 2) artificial intelligence algorithm to triage patient symptoms; and 3) patient education to address barriers to pain management. Participants were instructed to use ePAL over eight weeks. Patient-reported pain symptoms were assessed at baseline, Week-4, and Week-8 (primary endpoint) using the Brief Pain Inventory. Secondary outcomes include pain-related hospitalizations by Week-8. RESULTS: We enrolled 112 patients who were randomly assigned to ePAL (N = 56) or usual care (N = 56). Patients utilized ePAL on average 2.1 times per week to report pain symptoms, and 47.6% reported their pain at least once per week over eight weeks. Patients randomized to ePAL reported lower pain scores at Week-4 (mean: 3.16 vs. 4.28, P = 0.010) and week-8 (mean:2.99 vs. 4.05, P = 0.017), compared to those receiving usual care. Participants randomized to ePAL were less likely to experience a pain-related hospitalization compared to those in the usual care group (7.1% vs. 23.2% P = 0.018) CONCLUSIONS: ePAL was associated with lower patient-reported pain and fewer pain-related hospitalizations compared to usual care in patients with advanced cancer. This study demonstrates the promise of digital therapeutics for improving patients' symptoms while reducing burdensome hospitalizations.

Partners HealthCare Center for Connected Health.

This article reviews the history, current status, and future plans of the Partners HealthCare Center for Connected Health (the Center). Established in 1995 by Harvard Medical School teaching hospitals, the Center develops strategies to move healthcare from the hospital and doctor's office into the day-to-day lives of patients. It leverages information technology to help manage chronic conditions, maintain health and wellness, and improve adherence to prescribed regimen, patient engagement, and clinical outcomes. Since inception, it has served over 30,000 patients. The Center's core functions include videoconference-based real-time virtual visits, home vital sign monitoring, store-and-forward online consultations, social media, mobile technology, and other novel methods of providing care and enabling health and wellness remotely and independently of traditional time and geographic constraints. It offers a wide range of services, programs, and research activities. The Center comprises over 40 professionals with various technical and professional skills. Internally within Partners HealthCare, the role of the Center is to collaborate, guide, advise, and support the experimentation with and the deployment and growth of connected health technologies, programs, and services. Annually, the Center engages in a deliberative planning process to guide its annual research and operational agenda. The Center enjoys a diversified revenue stream. Funding sources include institutional operating budget/research funds from Partners HealthCare, public and private competitive grants and contracts, philanthropic contributions, ad hoc funding arrangements, and longer-term contractual arrangements with third parties.

Evaluating a web-based self-management program for employees with hypertension and prehypertension: a randomized clinical trial.

BACKGROUND: Web-based self-management programs offer a novel approach for self-insured employers seeking to improve and maintain employee health. METHODS: We conducted a 6-month prospective, cluster-randomized controlled trial designed to evaluate whether worksite access to an automated, web-based, self-management program resulted in better blood pressure control. The trial was conducted at 6 EMC Corporation worksites in Massachusetts, each of which had at least 600 employees; 404 EMC employees with pre-hypertension or hypertension participated. Participants at 3 worksites received a home blood pressure cuff that uploaded readings to a Web site where they could view trends and read automated rules-based messages. Participants at 3 worksites received access to an onsite blood pressure cuff. Primary outcome measure was change in systolic blood pressure. Secondary outcome measures were change in diastolic blood pressure, proportion of participants achieving significant changes in systolic and diastolic blood pressure, and subject satisfaction. RESULTS: Although the mean change in systolic blood pressure was not significantly different between intervention and control groups (-1.69 vs. -0.86 mm HG, respectively, P = .49) the change in diastolic blood pressure between groups was significant. (-1.08 vs. = 1.47 mm HG, respectively, P < .001). Significantly more intervention participants experienced a >10-mm Hg decrease in systolic blood pressure or >5-mm Hg decrease in diastolic blood pressure compared to controls (22% vs 17%, P = .02 and 29% vs 16%, P = .03, respectively). Intervention participants were twice as likely to report starting a new medication (P = .02) and more likely to report improved communication with their doctor (P = .02). CONCLUSIONS: Participation in an automated online self-management program resulted in improved blood pressure among employees with prehypertension or hypertension.

Long-running telemedicine networks delivering humanitarian services: experience, performance and scientific output.

OBJECTIVE: To summarize the experience, performance and scientific output of long-running telemedicine networks delivering humanitarian services. METHODS: Nine long-running networks--those operating for five years or more--were identified and seven provided detailed information about their activities, including performance and scientific output. Information was extracted from peer-reviewed papers describing the networks' study design, effectiveness, quality, economics, provision of access to care and sustainability. The strength of the evidence was scored as none, poor, average or good. FINDINGS: The seven networks had been operating for a median of 11 years (range: 5-15). All networks provided clinical tele-consultations for humanitarian purposes using store-and-forward methods and five were also involved in some form of education. The smallest network had 15 experts and the largest had more than 500. The clinical caseload was 50 to 500 cases a year. A total of 59 papers had been published by the networks, and 44 were listed in Medline. Based on study design, the strength of the evidence was generally poor by conventional standards (e.g. 29 papers described non-controlled clinical series). Over half of the papers provided evidence of sustainability and improved access to care. Uncertain funding was a common risk factor. CONCLUSION: Improved collaboration between networks could help attenuate the lack of resources reported by some networks and improve sustainability. Although the evidence base is weak, the networks appear to offer sustainable and clinically useful services. These findings may interest decision-makers in developing countries considering starting, supporting or joining similar telemedicine networks.

Linking electronic health record-extracted psychosocial data in real-time to risk of readmission for heart failure.

BACKGROUND: Knowledge of psychosocial characteristics that helps to identify patients at increased risk for readmission for heart failure (HF) may facilitate timely and targeted care. OBJECTIVE: We hypothesized that certain psychosocial characteristics extracted from the electronic health record (EHR) would be associated with an increased risk for hospital readmission within the next 30 days. METHODS: We identified 15 psychosocial predictors of readmission. Eleven of these were extracted from the EHR (six from structured data sources and five from unstructured clinical notes). We then analyzed their association with the likelihood of hospital readmission within the next 30 days among 729 patients admitted for HF. Finally, we developed a multivariable predictive model to recognize individuals at high risk for readmission. RESULTS: We found five characteristics-dementia, depression, adherence, declining/refusal of services, and missed clinical appointments-that were associated with an increased risk for hospital readmission: the first four features were captured from unstructured clinical notes, while the last item was captured from a structured data source. CONCLUSIONS: Unstructured clinical notes contain important knowledge on the relationship between psychosocial risk factors and an increased risk of readmission for HF that would otherwise have been missed if only structured data were considered. Gathering this EHR-based knowledge can be automated, thus enabling timely and targeted care.

Association of salivary lysozyme and C-reactive protein with metabolic syndrome.

INTRODUCTION: Salivary lysozyme (SLZ) is a proteolytic enzyme secreted by oral leucocytes and contains a domain that has an affinity to advanced glycation end products (AGE). Thus, we hypothesized that SLZ would be associated with metabolic syndrome (metS), a pro-inflammatory state. METHODS: Utilizing cross-sectional data from 250 coronary artery disease (CAD) and 250 non-CAD patients, the association of SLZ with metS was tested by logistic regression analyses controlling for age, sex, smoking, total cholesterol and C-reactive protein (CRP) levels. The analyses were stratified by CAD status to control for the possible effects of CAD. RESULTS: MetS was found in 122 persons. The adjusted odds ratio (OR) for metS associated with the highest quartile of SLZ was 1.95 with 95% confidence interval (CI) 1.20-3.12, p-value=0.007, compared with the lower three quartiles combined. Among the 40 subjects with metS but without CAD, the OR was 1.63 (CI: 0.64-4.15, p=0.31), whereas in the CAD group, SLZ was significantly associated with metS [OR=1.96 (1.09-3.52), p=0.02]. In both subgroups, CRP was not significantly associated with metS. CONCLUSION: SLZ was significantly associated with metS (OR=1.95) independent of CRP level. Future longitudinal research is warranted.

Neural Network-Based Algorithm for Adjusting Activity Targets to Sustain Exercise Engagement Among People Using Activity Trackers: Retrospective Observation and Algorithm Development Study.

BACKGROUND: It is well established that lack of physical activity is detrimental to the overall health of an individual. Modern-day activity trackers enable individuals to monitor their daily activities to meet and maintain targets. This is expected to promote activity encouraging behavior, but the benefits of activity trackers attenuate over time due to waning adherence. One of the key approaches to improving adherence to goals is to motivate individuals to improve on their historic performance metrics. OBJECTIVE: The aim of this work was to build a machine learning model to predict an achievable weekly activity target by considering (1) patterns in the user's activity tracker data in the previous week and (2) behavior and environment characteristics. By setting realistic goals, ones that are neither too easy nor too difficult to achieve, activity tracker users can be encouraged to continue to meet these goals, and at the same time, to find utility in their activity tracker. METHODS: We built a neural network model that prescribes a weekly activity target for an individual that can be realistically achieved. The inputs to the model were user-specific personal, social, and environmental factors, daily step count from the previous 7 days, and an entropy measure that characterized the pattern of daily step count. Data for training and evaluating the machine learning model were collected over a duration of 9 weeks. RESULTS: Of 30 individuals who were enrolled, data from 20 participants were used. The model predicted target daily count with a mean absolute error of 1545 (95% CI 1383-1706) steps for an 8-week period. CONCLUSIONS: Artificial intelligence applied to physical activity data combined with behavioral data can be used to set personalized goals in accordance with the individual's level of activity and thereby improve adherence to a fitness tracker; this could be used to increase engagement with activity trackers. A follow-up prospective study is ongoing to determine the performance of the engagement algorithm.

A Reinforcement Learning-Based Method for Management of Type 1 Diabetes: Exploratory Study.

BACKGROUND: Type 1 diabetes mellitus (T1DM) is characterized by chronic insulin deficiency and consequent hyperglycemia. Patients with T1DM require long-term exogenous insulin therapy to regulate blood glucose levels and prevent the long-term complications of the disease. Currently, there are no effective algorithms that consider the unique characteristics of T1DM patients to automatically recommend personalized insulin dosage levels. OBJECTIVE: The objective of this study was to develop and validate a general reinforcement learning (RL) framework for the personalized treatment of T1DM using clinical data. METHODS: This research presents a model-free data-driven RL algorithm, namely Q-learning, that recommends insulin doses to regulate the blood glucose level of a T1DM patient, considering his or her state defined by glycated hemoglobin (HbA1c) levels, body mass index, engagement in physical activity, and alcohol usage. In this approach, the RL agent identifies the different states of the patient by exploring the patient's responses when he or she is subjected to varying insulin doses. On the basis of the result of a treatment action at time step t, the RL agent receives a numeric reward, positive or negative. The reward is calculated as a function of the difference between the actual blood glucose level achieved in response to the insulin dose and the targeted HbA1c level. The RL agent was trained on 10 years of clinical data of patients treated at the Mass General Hospital. RESULTS: A total of 87 patients were included in the training set. The mean age of these patients was 53 years, 59% (51/87) were male, 86% (75/87) were white, and 47% (41/87) were married. The performance of the RL agent was evaluated on 60 test cases. RL agent-recommended insulin dosage interval includes the actual dose prescribed by the physician in 53 out of 60 cases (53/60, 88%). CONCLUSIONS: This exploratory study demonstrates that an RL algorithm can be used to recommend personalized insulin doses to achieve adequate glycemic control in patients with T1DM. However, further investigation in a larger sample of patients is needed to confirm these findings.