ABSTRACT
BACKGROUND: The growth in the capabilities of telehealth have made it possible to identify individuals with a higher risk of uncontrolled diabetes and provide them with targeted support and resources to help them manage their condition. Thus, predictive modeling has emerged as a valuable tool for the advancement of diabetes management. OBJECTIVE: This study aimed to conceptualize and develop a novel machine learning (ML) approach to proactively identify participants enrolled in a remote diabetes monitoring program (RDMP) who were at risk of uncontrolled diabetes at 12 months in the program. METHODS: Registry data from the Livongo for Diabetes RDMP were used to design separate dynamic predictive ML models to predict participant outcomes at each monthly checkpoint of the participants' program journey (month-n models) from the first day of onboarding (month-0 model) up to the 11th month (month-11 model). A participant's program journey began upon onboarding into the RDMP and monitoring their own blood glucose (BG) levels through the RDMP-provided BG meter. Each participant passed through 12 predicative models through their first year enrolled in the RDMP. Four categories of participant attributes (ie, survey data, BG data, medication fills, and health signals) were used for feature construction. The models were trained using the light gradient boosting machine and underwent hyperparameter tuning. The performance of the models was evaluated using standard metrics, including precision, recall, specificity, the area under the curve, the F1-score, and accuracy. RESULTS: The ML models exhibited strong performance, accurately identifying observable at-risk participants, with recall ranging from 70% to 94% and precision from 40% to 88% across the 12-month program journey. Unobservable at-risk participants also showed promising performance, with recall ranging from 61% to 82% and precision from 42% to 61%. Overall, model performance improved as participants progressed through their program journey, demonstrating the importance of engagement data in predicting long-term clinical outcomes. CONCLUSIONS: This study explored the Livongo for Diabetes RDMP participants' temporal and static attributes, identification of diabetes management patterns and characteristics, and their relationship to predict diabetes management outcomes. Proactive targeting ML models accurately identified participants at risk of uncontrolled diabetes with a high level of precision that was generalizable through future years within the RDMP. The ability to identify participants who are at risk at various time points throughout the program journey allows for personalized interventions to improve outcomes. This approach offers significant advancements in the feasibility of large-scale implementation in remote monitoring programs and can help prevent uncontrolled glycemic levels and diabetes-related complications. Future research should include the impact of significant changes that can affect a participant's diabetes management.
ABSTRACT
The Fifth Artificial Pancreas Workshop: Enabling Fully Automation, Access, and Adoption was held at the National Institutes of Health (NIH) Campus in Bethesda, Maryland on May 1 to 2, 2023. The organizing Committee included representatives of NIH, the US Food and Drug Administration (FDA), Diabetes Technology Society, Juvenile Diabetes Research Foundation (JDRF), and the Leona M. and Harry B. Helmsley Charitable Trust. In previous years, the NIH Division of Diabetes, Endocrinology, and Metabolic Diseases along with other diabetes organizations had organized periodic workshops, and it had been seven years since the NIH hosted the Fourth Artificial Pancreas in July 2016. Since then, significant improvements in insulin delivery have occurred. Several automated insulin delivery (AID) systems are now commercially available. The workshop featured sessions on: (1) Lessons Learned from Recent Advanced Clinical Trials and Real-World Data Analysis, (2) Interoperability, Data Management, Integration of Systems, and Cybersecurity, Challenges and Regulatory Considerations, (3) Adaptation of Systems Through the Lifespan and Special Populations: Are Specific Algorithms Needed, (4) Development of Adaptive Algorithms for Insulin Only and for Multihormonal Systems or Combination with Adjuvant Therapies and Drugs: Clinical Expected Outcomes and Public Health Impact, (5) Novel Artificial Intelligence Strategies to Develop Smarter, More Automated, Personalized Diabetes Management Systems, (6) Novel Sensing Strategies, Hormone Formulations and Delivery to Optimize Close-loop Systems, (7) Special Topic: Clinical and Real-world Viability of IP-IP Systems. "Fully automated closed-loop insulin delivery using the IP route," (8) Round-table Panel: Closed-loop performance: What to Expect and What are the Best Metrics to Assess it, and (9) Round-table Discussion: What is Needed for More Adaptable, Accessible, and Usable Future Generation of Systems? How to Promote Equitable Innovation? This article summarizes the discussions of the Workshop.
Subject(s)
Diabetes Mellitus, Type 1 , Pancreas, Artificial , Humans , Diabetes Mellitus, Type 1/drug therapy , Insulin/therapeutic use , Blood Glucose , Artificial Intelligence , Insulin Infusion Systems , Insulin, Regular, Human/therapeutic use , Automation , Hypoglycemic Agents/therapeutic useABSTRACT
Unrecognized transmission is a major contributor to ongoing TB epidemics in high-burden, resource-constrained settings. Limitations in diagnosis, treatment, and infection control in health-care and community settings allow for continued transmission of drug-sensitive and drug-resistant TB, particularly in regions of high HIV prevalence. Health-care facilities are common sites of TB transmission. Improved implementation of infection control practices appropriate for the local setting and in combination, has been associated with reduced transmission. Community settings account for the majority of TB transmission and deserve increased focus. Strengthening and intensifying existing high-yield strategies, including household contact tracing, can reduce onward TB transmission. Recent studies documenting high transmission risk community sites and strategies for community-based intensive case finding hold promise for feasible, effective transmission reduction. Infection control in community settings has been neglected and requires urgent attention. Developing and implementing improved strategies for decreasing transmission to children, within prisons and of drug-resistant TB are needed.
Subject(s)
Tuberculosis/transmission , Antitubercular Agents/therapeutic use , Community-Acquired Infections/prevention & control , Cross Infection/prevention & control , Health Facilities/standards , Humans , South Africa , Tuberculosis/drug therapy , Tuberculosis/prevention & control , Tuberculosis, Multidrug-Resistant/prevention & controlABSTRACT
BACKGROUND: Diabetes clinicians are key facilitators of continuous glucose monitoring (CGM) provision, but data on provider behavior related to CGM use and CGM generated data are limited. METHODS: We conducted a national survey of providers caring for people with diabetes on CGM-related opinions, facilitators and barriers to prescription, and data review practices. RESULTS: Of 182 survey respondents, 73.2% worked at academic centers, 70.6% were endocrinologists, and 70.7% practiced in urban settings. Nearly 70% of providers reported CGM use in the majority of their patients with type 1 diabetes. Half of the providers reported CGM use in 10% to 50% of their patients with type 2 diabetes. All respondents believed CGM improved quality of life and could optimize diabetes control. We found no differences in reported rates of CGM use based on providers' years of experience, patient volume, practice setting, or clinic type. Most providers reviewed CGM data each visit (97.7%) and actively involved patients in the data interpretation (98.8%). Only 14.1% of clinicians reported reviewing CGM data without any prompting from patients or their family members outside of visits. Most providers (80.7%) reported their CGM data review was valued by patients although only half reported having adequate time (45.1%) or an efficient process (56.1%) to do so. CONCLUSIONS: Despite uniform support for CGM by providers, ongoing challenges related to cost, insurance coverage, and difficulties with prescription were major barriers to CGM use. Increased use of CGM in appropriate populations will necessitate improvements in data access and integration, clearly defined workflows, and decreased administrative burden to obtain CGM.
Subject(s)
Diabetes Mellitus, Type 1 , Diabetes Mellitus, Type 2 , Humans , Diabetes Mellitus, Type 2/drug therapy , Blood Glucose , Blood Glucose Self-Monitoring , Quality of Life , Diabetes Mellitus, Type 1/drug therapyABSTRACT
Diabetes Technology Society hosted its annual Diabetes Technology Meeting from November 3 to November 5, 2022. Meeting topics included (1) the measurement of glucose, insulin, and ketones; (2) virtual diabetes care; (3) metrics for managing diabetes and predicting outcomes; (4) integration of continuous glucose monitor data into the electronic health record; (5) regulation of diabetes technology; (6) digital health to nudge behavior; (7) estimating carbohydrates; (8) fully automated insulin delivery systems; (9) hypoglycemia; (10) novel insulins; (11) insulin delivery; (12) on-body sensors; (13) continuous glucose monitoring; (14) diabetic foot ulcers; (15) the environmental impact of diabetes technology; and (16) spinal cord stimulation for painful diabetic neuropathy. A live demonstration of a device that can allow for the recycling of used insulin pens was also presented.
Subject(s)
Diabetes Mellitus, Type 1 , Humans , Diabetes Mellitus, Type 1/drug therapy , Blood Glucose , Blood Glucose Self-Monitoring , Insulin/therapeutic use , Insulin Infusion Systems , Technology , Hypoglycemic Agents/therapeutic useSubject(s)
Capillary Leak Syndrome/diagnosis , Diagnostic Errors , Hereditary Angioedema Type III/diagnosis , Immunoglobulins, Intravenous/therapeutic use , Abdominal Pain/etiology , Acute Kidney Injury/etiology , Capillary Leak Syndrome/therapy , Complement C1 Inhibitor Protein/analysis , Delayed Diagnosis , Dyspnea/etiology , Humans , Hypotension/etiology , Male , Middle Aged , Monoclonal Gammopathy of Undetermined Significance/complications , Monoclonal Gammopathy of Undetermined Significance/diagnosis , Pericardial Effusion/etiology , RecurrenceABSTRACT
With the first commercially available smart insulin pens, the predominant insulin delivery device for millions of people living with diabetes is now coming into the digital age. Smart insulin pens (SIPs) have the potential to reshape a connected diabetes care ecosystem for patients, providers, and health systems. Existing SIPs are enhanced with real-time wireless connectivity, digital dose capture, and integration with personalized dosing decision support. Automatic dose capture can promote effective retrospective review of insulin dose data, particularly when paired with glucose data. Patients, providers, and diabetes care teams will be able to make increasingly data-driven decisions and recommendations, in real time, during scheduled visits, and in a more continuous, asynchronous care model. As SIPs continue to progress along the path of digital transformation, we can expect additional benefits: iteratively improving software, machine learning, and advanced decision support. Both these technological advances, and future care delivery models with asynchronous interactions, will depend on easy, open, and continuous data exchange between the growing number of diabetes devices. SIPs have a key role in modernizing diabetes care for a large population of people living with diabetes.
Subject(s)
Diabetes Mellitus , Ecosystem , Diabetes Mellitus/drug therapy , Humans , Insulin , Insulin Infusion Systems , Machine LearningABSTRACT
In this study by Alva et al, accuracy of a second-generation factory calibrated continuous glucose monitoring system is evaluated. Compared to the first-generation FreeStyle Libre 14-day system (FSL), accuracy was improved throughout the 14-day wear period, including improved accuracy in hypoglycemia for adults and youth. The addition of optional real-time alerts for hypoglycemia and hyperglycemia as well as an integrated continuous glucose monitor (iCGM) designation by the FDA may further enable users to benefit from using CGM in real time, including in future automated insulin delivery systems. As CGM accuracy, affordability, and accessibility improve, we anticipate increased uptake of CGM by people on intensive insulin therapy, and also potential benefits and expansion into a broader patient population. There are growing opportunities to leverage cloud-connected CGM devices in the increasingly virtual, continuous telehealth-driven diabetes care model, which will require more focus on development and use of data interoperability standards.
Subject(s)
Diabetes Mellitus, Type 1 , Hypoglycemia , Adolescent , Adult , Algorithms , Blood Glucose , Blood Glucose Self-Monitoring/instrumentation , Child , Diabetes Mellitus, Type 1/drug therapy , HumansABSTRACT
Importance: The internal medicine (IM) chief residency is a position of leadership and honor common in IM programs, but the goals, responsibilities, and experiences of those who undertake it can be highly variable. Studies assessing the experience and impressions of the chief resident (CR) position from the viewpoint of the IM CRs are lacking. Objective: To describe the structure, responsibilities, and perceptions of the IM CR role across IM residency programs. Design, Setting, and Participants: A cross-sectional, simple descriptive electronic survey for current CRs was administered between April and June 2018 across US IM residency programs accredited by US Accreditation Council for Graduate Medical Education. A 2-step nonrandom sampling approach was used: first, snowball sampling was performed using the authors' professional networks, and second, the survey was sent to the Association of Program Directors in Internal Medicine (APDIM) CR listserv. Data analysis was performed from June 2020 to August 2020. Exposures: Participation as a CR for an IM residency program in the 2017 to 2018 academic year. Main Outcomes and Measures: Descriptive CR personal and program characteristics and CR perceptions of administrative, clinical, and leadership experiences. Results: Among 169 unique responses, 77 participants (46%) were female and 89 (53%) were White. The response rate was 57% (70 of 122 surveys) in the snowball sample and 12% (99 of 842 surveys) in the APDIM listserv. The 2 sampled groups were combined for analysis. Most respondents (125 CRs [74%]) were from academic or university-based programs. Common across CR responses was responsibility for administrative tasks, clinical work, and educational efforts. Most CRs (111 of 157 respondents [71%]) reported being the primary schedulers for the residency program. Clinical work differed widely across respondents. Only 70 of 156 respondents (45%) reported having an academic title associated with the CR role. CRs reported inconsistent evaluation throughout the year, with high percentages reporting never receiving feedback on teaching (34 respondents [23%]), clinical abilities (67 respondents [45%]), or leadership abilities (60 respondents [40%]). Most CRs (107 respondents [69%]) agreed that they find work as a CR fulfilling and 117 (74%) agreed they would do chief residency again. Conclusions and Relevance: Despite its ubiquity in training programs across the US, the IM CR experience is very different across programs. Recommendations are provided to consider for improvement of the CR experience, including structured feedback opportunities, maximizing educational and clinical experiences, and standardizing policies.
Subject(s)
Internal Medicine , Internship and Residency , Cross-Sectional Studies , Education, Medical, Graduate , Female , Humans , Internal Medicine/education , LeadershipABSTRACT
PURPOSE OF REVIEW: The role of telehealth in the care of people with type 1 diabetes (T1D) has expanded dramatically during the coronavirus pandemic, and is expected to remain a major care delivery modality going forward. This review explores the landscape of recent evidence for telehealth in T1D care. RECENT FINDINGS: Telemedicine for routine T1D care has shown equivalence to standard in-person care, with respect to glycemic control, while also increasing access, convenience, and satisfaction. Telehealth use promotes increased engagement of adolescents with T1D. Telehealth platforms have successfully been used in the care of microvascular complications and to support mental health related to diabetes. Machine learning and advanced decision support will increasingly be used to augment T1D care, as recent evidence suggests increasing capabilities to improve glycemic control. A spectrum of digital connected care services are emerging to support people with diabetes with daily management of diabetes. Finally, policy and systems are required that promote data interoperability, telemedicine provision, and reimbursement to support the ongoing growth of telehealth in T1D. SUMMARY: A developing field of evidence supports use of telehealth in T1D. As this care modality scales, it has the potential to increase access to high-quality diabetes care for many people with T1D.
Subject(s)
Diabetes Mellitus, Type 1/therapy , Telemedicine , COVID-19 , Delivery of Health Care , Humans , Mental Health , Telemedicine/methodsABSTRACT
BACKGROUND: During the COVID-19 pandemic, telemedicine use rapidly and dramatically increased for management of diabetes mellitus. It is unknown whether access to telemedicine care has been equitable during this time. This study aimed to identify patient-level factors associated with adoption of telemedicine for subspecialty diabetes care during the pandemic. METHODS: We conducted an explanatory sequential mixed-methods study using data from a single academic medical center. We used multivariate logistic regression to explore associations between telemedicine use and demographic factors for patients receiving subspecialty diabetes care between March 19 and June 30, 2020. We then surveyed a sample of patients who received in-person care to understand why these patients did not use telemedicine. RESULTS: Among 1292 patients who received subspecialty diabetes care during the study period, those over age 65 were less likely to use telemedicine (OR: 0.34, 95% CI: 0.22-0.52, P < .001), as were patients with a primary language other than English (OR: 0.53, 95% CI: 0.31-0.91, P = .02), and patients with public insurance (OR: 0.64, 95% CI: 0.49-0.84, P = .001). Perceived quality of care and technological barriers were the most common reasons cited for choosing in-person care during the pandemic. CONCLUSIONS: Our findings suggest that, amidst the COVID-19 pandemic, there have been disparities in telemedicine use by age, language, and insurance for patients with diabetes mellitus. We anticipate telemedicine will continue to be an important care modality for chronic conditions in the years ahead. Significant work must therefore be done to ensure that telemedicine services do not introduce or widen population health disparities.
Subject(s)
COVID-19/prevention & control , Communicable Disease Control , Diabetes Mellitus/therapy , Healthcare Disparities , Telemedicine , Adolescent , Adult , Aged , Aged, 80 and over , COVID-19/epidemiology , California/epidemiology , Child , Child, Preschool , Communicable Disease Control/methods , Delivery of Health Care/methods , Delivery of Health Care/organization & administration , Delivery of Health Care/standards , Delivery of Health Care/statistics & numerical data , Diabetes Mellitus/epidemiology , Diabetes Mellitus, Type 1/epidemiology , Diabetes Mellitus, Type 1/therapy , Diabetes Mellitus, Type 2/epidemiology , Diabetes Mellitus, Type 2/therapy , Endocrinology/methods , Endocrinology/organization & administration , Female , Health Services Accessibility/organization & administration , Health Services Accessibility/standards , Health Services Accessibility/statistics & numerical data , Healthcare Disparities/statistics & numerical data , Humans , Infant , Male , Middle Aged , Pandemics , Primary Health Care/organization & administration , Primary Health Care/standards , Primary Health Care/statistics & numerical data , Quarantine , SARS-CoV-2 , Telemedicine/organization & administration , Telemedicine/statistics & numerical data , Young AdultABSTRACT
The COVID-19 pandemic has had a profound impact on the nation's health care system, including on graduate medical education (GME) training programs. Traditionally, residency and fellowship training program applications involve in-person interviews conducted on-site, with only a minority of programs offering interviews remotely via a virtual platform. However, in light of the COVID-19 pandemic, it is anticipated that most interviews will be conducted virtually for the 2021 application cycle and possibly beyond. Therefore, GME training programs need to prepare for the transition to virtual interviews using evidence-based practices. At the University of California, San Francisco, a multidisciplinary task force was convened to review existing literature about virtual interviews and determine best practices. This article summarizes these findings, first discussing the advantages and disadvantages of the virtual interview format and then providing evidence-based best practices for GME training programs. Specifically, the authors make the following recommendations: develop a detailed plan for the interview process, consider using standardized interview questions, recognize and respond to potential biases that may be amplified with the virtual interview format, prepare your own trainees for virtual interviews, develop electronic materials and virtual social events to approximate the interview day, and collect data about virtual interviews at your own institution. With adequate preparation, the virtual interview experience can be high yield, positive, and equitable for both applicants and GME training programs.