AI-Driven Nutrition Applications: A Literature Review of Adoption, Functionality, and Future Potential
Cynthia Williams, PhD, MHA, PT1, Anna K. van Niekerk, DO2, Thomas C. DeLuca, DO3, Stephen M. McMullan, MD4, Dusty Marie Narducci, MD5, George Pujalte, MD6
1University of Central Florida; 2Mayo Clinic; 3Mayo Clinic School of Graduate Medical Education, Mayo Clinic; 4Mayo Clinic School of Graduate Medical Education, Mayo Clinic College of Medicine and Science; 5DocDustry; 6Department of Family Medicine, Mayo Clinic, Jacksonville, Florida
Description: AI-driven nutrition applications are transforming dietary management by offering personalized, culturally adaptive meal planning tools. Widely adopted in upper-middle-income countries, these platforms integrate health data, image recognition, and wearable inputs to support chronic disease care. Despite challenges in data quality and equity, future models promise enhanced clinical integration and patient engagement.
Abstract: AI-enabled nutrition planning is rapidly transforming dietary management by offering personalized, accessible, and adaptive tools tailored to diverse health needs. As of 2023, over 20 million individuals globally used AI-powered meal planning applications, with 5 million daily active users. These tools are primarily adopted in upper-middle-income countries and target diet-related conditions such as diabetes, obesity, and gastrointestinal disorders. Unlike traditional nutrition interventions, which often require high literacy and manual tracking, AI applications leverage image recognition, barcode scanning, and wearable device integration to streamline dietary assessment and provide real-time, culturally adapted recommendations. In the U.S., several mobile platforms now utilize health data, including disease states, blood pressure, and goals, to generate personalized meal plans based on guidelines from reputable sources such as the USDA and the American Dietary Guidelines. Some applications calculate energy needs using the Mifflin-St. Jeor Equation and align macronutrient distribution with AMDR standards. Future models envision full integration with electronic health records, enabling dynamic feedback, motivational support, and provider summaries to be shared between visits. These tools may also identify at-risk patients across panels and connect users to local resources. While promising, challenges remain around data quality, cultural sensitivity, and regulatory oversight. This review explores current applications, their benefits, limitations, and future potential in clinical nutrition care.
Classification of Research: Information Technology
Method: Literature Review
Results: AI-powered nutrition applications are revolutionizing dietary management by providing personalized, accessible, and adaptable tools tailored to diverse health needs. Apps like FoodLogger use image recognition to analyze meals and suggest healthier alternatives, while MyFitnessPal integrates barcode scanning and fitness devices for comprehensive tracking. NutriGenie utilizes DNA data and lifestyle inputs to create personalized meal plans, while SmartEat adjusts nutrition according to food availability and individual preferences. Samsung Food creates meal plans from refrigerator photos to reduce waste and expenses. Tummy AI considers dietary restrictions, goals, local food prices, and cooking skill level. Yuka simplifies food choices by assigning health scores based on ingredient analysis and global standards. These tools enhance autonomy, reduce barriers to professional nutrition support, and improve continuity of care. They also support real-time feedback, motivational strategies, and integration with electronic health records. Despite their promise, challenges remain around data quality, cultural sensitivity, and regulatory oversight.
Conclusions: AI-enabled nutrition planning holds promise for a personalized future in dietary care. These tools can integrate with electronic health records to deliver personalized meal plans tailored to individual health data, preferences, and real-time feedback. By analyzing inputs such as calorie intake, exercise, sleep, stress, and lab results, AI applications adapt recommendations and provide educational resources and motivational support between clinical visits. They also enhance provider efficiency by summarizing patient progress and identifying individuals at risk for targeted interventions. With growing patient and provider acceptance, these tools enhance accessibility, autonomy, and continuity of care, although challenges persist regarding data quality, cultural relevance, and regulatory compliance.
Association Between Telemedicine Use and Hospitalization Among US Adults with Diabetes: Exploring the Role of Comorbidity Burden
Description: Telemedicine expanded exponentially for diabetes management during the pandemic, yet the relationship between telemedicine use and health care utilization in real-world settings remains poorly understood. Up-to-date national studies examining whether this association differs by patient comorbidity burden are needed to inform evidence-based telemedicine policy as temporary coverage flexibilities expire.
Abstract: Objective: To assess the association between telemedicine use and inpatient care utilization among US adults with diabetes, adjusting for comorbidity burden and exploring clinical complexity as an effect modifier.
Methods: We conducted a retrospective cohort study using longitudinal and office-based visit files from the 2022-2023 Medical Expenditure Panel Survey. The primary outcome was any inpatient expenditure (Year 2);primary exposure was any telemedicine visit (Year 1). Comorbidity burden was measured using MEPS priority chronic conditions, categorized as low (≤2 conditions) versus high (≥3 conditions).
We used weighted logistic regression models to assess main effects and effect modification using an interaction term (telemedicine × comorbidity burden), adjusting for sociodemographic factors (age, sex, region, race/ethnicity, insurance) and health-related variables (BMI, self-rated health).
Classification of Research: Clinician Experience
Method: Secondary Data Analysis
Results: Among 869 unweighted (27.9 million weighted) participants, 11.9% used telemedicine and 53.3% had high comorbidity burden. Mean inpatient expenditure was $9,525 for telemedicine users versus $5,023 for non-users. After adjustment, telemedicine users had 2.84 times the odds of hospitalization (p<0.05);those with ≥3 comorbidities had 2.28 times the odds (p<0.05). The interaction term approached significance (p=0.089). Exploratory stratified analyses revealed no association among low-comorbidity patients (OR=1.55, p=0.54) but strong association among high-comorbidity patients (OR=4.85, 95% CI: 2.2-10.9, p<0.001).
Conclusions: Comorbidity burden partially explained telemedicine-hospitalization association but did not fully account for them. Exploratory analyses suggest the relationship may differ by clinical complexity, with stronger associations among patients with multiple chronic conditions, though limited statistical power (interaction p=0.089) precludes definitive conclusions. The pattern is consistent with appropriate targeting of telemedicine to clinically complex patients requiring intensive management. These hypothesis-generating findings warrant replication in larger studies with detailed clinical measures. As policymakers consider permanent telemedicine coverage expansion, higher utilization among telemedicine users may reflect appropriate patient selection rather than adverse effects of virtual care.
Association of Virtual Transitional Care Intervention With Post-Discharge Emergency Department Utilization: A Two-Year System Wide Evaluation
Houston Methodist Research Institute, 1Center for Connected Care Innovation and Implementation-Research, 2 Center for Health Data Science & Analytics
Description: Virtual Urgent Care (VUC), launched in 2019, was leveraged in 2023 as an existing resource to address common post emergency department (ED) discharge concerns at Houston Methodist Hospital. Following the initial evaluation of the TCI utilization, this second-year analysis focuses on patient engagement, assessing how virtual transitional care influences continuity of care during critical ED-to-home transition.
Abstract: ED overcrowding and avoidable revisits are common, resulting in operational burdens and adverse clinical outcomes. At Houston Methodist, an urban hospital in Houston, Texas, the VUC program supports the transition of care following ED discharge through automated SMS text messaging, and subsequent virtual evaluation. Early implementation demonstrated feasibility and reduced spontaneous ED revisits, motivating evaluation of longer-term performance and sustainability. The TCI program has since been expanded across all EDs and affiliated freestanding emergency care centers (ECCs) within our health system. This retrospective observational cohort study included 156,193 patients discharged from EDs within a single U.S. hospital system between September 2023 and July 2025. Patients were categorized based on engagement with the virtual transitional care intervention as VUC-completed or VUC-noncompleted. The primary outcome was patient engagement with the automated SMS outreach and subsequent virtual visit completion;secondary outcomes included 30-day ED returns. As a retrospective observational study, outcomes may be influenced by unmeasured differences and should be interpreted in the context of potential selection bias, warranting further evaluation using prospective methodologies.
Classification of Research: Access to Care
Method: Observational
Results: Of 156,193 discharged patients, 77.29% (120,724/156,193) received an automated SMS text messaging for the virtual transitional care program, and 2.8% (3,373/120,724) accessed the scheduling link. Of these, 50.7% (1,710/3,373) requested a visit, 40.88% (699/1,710) scheduled, and 62.37% (436/699) completed the visit (completed group). Of those completing a VUC visit, 129 (29.6%) received additional care, including new prescriptions, ambulatory referrals, and ED referral when clinically indicated. Observed ED return rates within 30 days were lower among VUC-completed patients compared with VUC-noncompleted patients. Engagement and utilization patterns remained stable throughout the two-year evaluation period.
Conclusions: Engagement with a virtual transitional care intervention was associated with sustained patient participation and favorable patterns of post-discharge ED utilization over a two-year period. System-wide implementation demonstrated stability and scalability of the intervention. While limited by its observational design, these findings suggest that virtual transitional care may support continuity of care during ED-to-home transitions and inform future prospective and risk-adjusted studies.
Evaluating a Telehealth-enabled Collaborative Care Program for Behavioral Health in Rural Primary Care practices within a Large Academic Medical center
Caitlin Koob, PhD, MS, Research Associate-Faculty, Emily Johnson, PhD, Professor, Andrew Alkis, MD, Center for Telehealth Psychiatrist and Clinical Assistant Professor, Candace Sprouse-McClam, PhD, LISW-CP (S), Collaborative Care Clinical Manager, Katie Kirchoff, MSHI, Database Administrator, Jennifer Dahne, PhD, Clinical Psychologist and Professor
Medical University of South Carolina (MUSC)
Description: Psychiatric collaborative care management (CoCM) effectively addresses behavioral health in primary care;however, uptake in rural communities remains challenging. Our telehealth-enabled CoCM program seeks to improve access and quality of behavioral health services across rural South Carolina. This study aims to describe barriers and facilitators to telehealth CoCM patient enrollment.
Abstract: Purpose: To describe the barriers and facilitators to enrollment among patients referred to the CoCM program by their primary care providers, through mixed-methods evaluation.
Methods: Data were triangulated from multiple sources, including electronic health records and patient-reported experiences among those enrolled and not enrolled, via surveys (N=43) and interviews (N=5), to inform program improvements and leverage data-driven decisions. Descriptive statistics and rapid qualitative data analysis were conducted to comprehensively evaluate targeted outcomes.
Results: Over the past year, CoCM has grown to serve 44 primary care clinics statewide, compared to 18 clinics previously. Preliminary analyses suggest that CoCM is increasing reach, and more than one-third of enrollees graduated from the program (37.0%). Further, enrolled patients largely experienced improvements in their mental health since joining CoCM (62.5%) and highlighted benefits of the telehealth-enabled delivery, accessibility, flexibility, and program support. The majority of those who did not enroll reported they did not remember receiving program information from their primary care provider (58%), but a similar program would be beneficial to their health and well-being (68%).
Conclusions: Further work is needed to describe the utility of the CoCM program, understand its impact on patients’ health care access, and evaluate the sustainability of this model.
Classification of Research: Quality Improvement
Method: Implementation Science
Results: From June 1, 2023 through June 30, 2024, 303 patients were referred to CoCM and enrolled (58.4% of total referred patients), and 112 of enrolled patients (37.0%) graduated from the program. Patients participated in CoCM for a mean of 164 days. Enrollees largely experienced improvements in their mental health since joining CoCM (62.5%) and highlighted benefits of the telehealth-enabled delivery, accessibility, flexibility, and program support. The majority of those who did not enroll reported that they did not remember receiving program information from their primary care provider (58%), but a similar program would be beneficial to their health and well-being (68%).
Conclusions: Feedback from referred patients, including those enrolled and not enrolled, informs data-driven decisions to improve accessibility, patient outcomes, and long-term scalability. Last fall, CoCM transitioned from grant funding to standard billing practices, and its impact continues to grow with the program. Next steps include robust analyses of patient-reported data (i.e., sleep and mood trackers, clinical screening tools) in a patient-facing telehealth platform. Additional sustainability measures include revenue per program participant, rather than reliance on grant funding, and revenue growth rate (month-over-month) from program participants.
Expanding Digital Health Support for Pediatric Onboarding in Remote Patient Monitoring
Aaron Farber-Chen, MSN, RN, FNP-C; aaron.farber-chen@childrens.harvard.edu; Connected Care Director, Logan Harper, MS, Digital Health Product Manager, Susan Barrett, Digital Health Support Manager, Ravneet Kaur, MBA, Digital Health Data Analytics Manager
Boston Children's Hospital
Description: At Boston Children’s Hospital (BCH), this project expands the role of Digital Health Support (DHS) from troubleshooting to structured onboarding and education for pediatric patients and caregivers in Remote Patient Monitoring (RPM). By ensuring consistency, strengthening confidence, and reducing caregiver and provider burden, this model drives engagement and offers a scalable approach across diverse specialties and populations.
Abstract: RPM has grown across pediatric specialties, but programs continue to face challenges with variable onboarding, digital literacy, and sustained engagement. Without structured education and early support, families struggle to fully participate in RPM.
Building on prior work, the BCH RPM Product Team partnered with DHS to expand its role beyond troubleshooting. Beginning with the Pediatric Cardiac Nutrition (PECAN) program, DHS Navigators prepared families with guided walkthroughs of account access, device setup, and submission of the first measurement, providing consistent and supported introductions to RPM.
This collaboration reduces variation in onboarding, eases caregiver burden, and preserves provider time for direct patient care. Current engagement is approximately 50%, leaving significant room for improvement. With DHS-led onboarding, we aim to increase engagement to at least 75% and achieve ≥80% satisfaction with the onboarding process.
Although piloted in pediatrics, this model is broadly applicable. By standardizing onboarding and education, it provides a replicable framework to strengthen engagement, reduce caregiver and provider burden, and support sustainable RPM adoption across health systems.
Classification of Research: Quality Improvement
Method: Implementation Science
Results: This initiative established structured onboarding led by DHS, with Navigators guiding families through account access, device setup, and first measurement submission. Offloading onboarding tasks to DHS eases caregiver burden and preserves provider time for clinical care. Current engagement is ∼50%;with DHS involvement, we anticipate ≥75% engagement and ≥80% satisfaction. Results presented at the conference will highlight caregiver satisfaction, provider time saved, and early indicators of clinical impact, illustrating a scalable and replicable approach for RPM across programs.
Conclusions: Expanding DHS from a reactive service to a proactive partner in onboarding and education creates a scalable, replicable model for RPM. This approach strengthens caregiver confidence, eases caregiver burden, reduces variation, and preserves provider time. In pediatrics, where caregivers are central, structured onboarding ensures families are prepared to engage. This framework is equally relevant across specialties, where consistent education and support are essential for adoption. By embedding structured onboarding, this initiative improves experience, supports caregiver involvement, and provides a sustainable model for health systems implementing remote monitoring.
Feasibility and Early Outcomes of a Multi-Site Virtual Hospitalist Admissions Program in a Large Academic Health System
Farhan Ishaq MD MPH MBA,1,3 Ngoc-Anh Anh Nguyen, MD,1,2 Sarah Sossong MPH, Henry Ellison, MD,3 Grace Lee, BS,1 Lindsay Randle, MBA, Melissa Gomez MBA, Jannika Machnik MS, Brendan Holderread, MD, Sarah Pletcher, MD MHCDS1
1Center for Connected Care, Innovation and Implementation Research, Houston Methodist Hospital, Houston, TX, 2Department of Medicine, Houston Methodist Hospital, Houston, TX, USA, 3Houston Methodist Research Institute, Houston Methodist Hospital, Houston, TX, 4Department of Surgery, Houston Methodist Hospital, Houston, TX
Description: Houston Methodist implemented a nocturnal TeleHospitalist program to address staffing shortages and admission delays. The program proved feasible, safe, and well accepted by patients and staff. Older adults responded positively, indicating that age did not limit acceptance. Early results demonstrate improved timeliness, workflow efficiency, and continuity, supporting broader adoption of virtual admissions.
Abstract: Background: TeleHospitalist models have emerged as a solution associated with improved outcomes, enhanced care coordination improved guideline adherence, reduced length of stay, lower costs, more timely admissions, fewer adverse events, reduced 30-day readmissions, fewer ICU transfers, and better outcomes care coordination. Approximately 48.6% of hospital admissions during overnight hours night, and staffing shortages contribute to admission delays, fragmented care, and hospitalist burnout, creating a persistent mismatch between demand and available staffing.
Although telemedicine has been widely adopted in many health care settings, its use in general inpatient admissions is not well studied. We evaluated the feasibility, operational performance, and early perceptions of a virtual hospitalist admissions program implemented across a large academic health system.
Methods: In October 2024, Houston Methodist launched a nocturnal TeleHospitalist program providing centralized coverage via audio-video technology integrated into the electronic medical record. Initially staffed from 6:00 PM to 2:00 AM five nights per week, the program expanded to 6:00 PM to 6:00 AM and seven nights per week by May 2025. Operational metrics were extracted retrospectively from the Epic EHR. Patient experience was assessed through a six-item survey, and staff feedback was collected via structured surveys and qualitative input.
Classification of Research: Quality Improvement
Method: Implementation Science
Results: From October 2024 to August 2025, 1,575 TeleHospitalist encounters were completed (median patient age 65 years [IQR 49–75];56% ≥65 years). The median hospital length of stay was 38.5 hours (IQR 17.3–66.3), and the median ED arrival-to-admission time was 7.3 hours (IQR 4.7–12.8). Among 311 patient survey respondents (19.7%;median age 66.5 years [IQR 52.3–75.6];56% ≥65 years old), satisfaction scores were consistently high (median 4 [IQR 3–5] across all domains). Staff survey responses (n=29) demonstrated strong endorsement: 86% favored continuation, 79% reported comfort with the technology, and 75% rated admission processes as better than traditional workflows. Qualitative feedback emphasized improved timeliness, easier provider access, and operational efficiency, though concerns included privacy and impersonality of virtual care.
Conclusions: A system-wide virtual hospitalist admissions program was feasible, safe, and well accepted by both patients and staff. Importantly, older adults (≥65 years), who represented 56% of survey respondents, reported positive experiences with the virtual admission process, suggesting that age is not a barrier to acceptance of TeleHospitalist care. Early results indicate improvements in perceived timeliness, communication, and workflow efficiency while addressing nocturnal staffing gaps. These findings support the potential of TeleHospitalist models to strengthen continuity of care in hospital medicine. Future research should assess long-term clinical outcomes, cost-effectiveness, and strategies to optimize privacy and patient engagement.
Friendi.fi: An AI Friend and Rules-based Coach Duo Significantly Improves Sleep Quality and Quantity
Victor Wang, MS, Chantal; Kerssens, PhD, Co-Founder
Friend.fi
Description: We piloted a sleep coaching program based National Council on Aging (NCOA) guidance, delivered through Friendi.fi, a generative AI-powered friend supported by human moderation for quality and safety, and guided by rules-based coach persona. Friendi.fi significantly improved sleep quantity and quality by Day 10, sustained through Day 30.
Abstract: Sleep is a foundational health behavior that strongly influences mental and physical well-being, and the onset of cognitive impairment and dementia. Digital coaching has emerged as a scalable intervention, but many programs fail to sustain engagement and long-term impact. We tested whether an AI persona acting as an engaging friend, which converses and relates to each user more engagingly than a purpose-driven coach or typical chatbot, could drive sustained engagement and thereby measurable improvements in health behaviors and outcomes when combined with a separate, rules-based coach persona that administers guidelines and evidence-based practices without “contaminating” the engagingness and informality of the user's relationship with the friend persona. This novel arrangement is patent pending, and the AI model used to generate the friend persona's utterances is proprietary.
We recruited adult participants online and delivered a structured 18-day coaching program based on NCOA sleep guidelines. Outcomes were measured at baseline, Day 10, Day 18, and Day 30 using validated scales. Analyses included unmatched group comparisons and within-subject Wilcoxon Signed-Rank tests on completers.
Results show that the Friendi.fi program drove significant improvements in sleep quantity and quality by Day 10, sustained through Day 30. Self-rated health and mental health improvements were directionally positive.
Classification of Research: Patient Experience
Method: Survey/Qualitative
Results: (1) Sleep quantity improved significantly by Day 10 (p=0.01) and Day 18 (p=0.01), and remained elevated at Day 30 (p=0.03), even after the conclusion of coaching; (2) Sleep quality also improved significantly by Day 10 (p=0.03) and Day 18 (p=0.01), and gains were maintained at Day 30 (p=0.02); (3) SRH and SRMH improvements were not statistically significant, though directionally positive, with SRMH trending toward significance (p=0.07); and (4) Most sleep behavior improvements occurred by Day 10, while SRMH gains appeared later, consistent with the expected lag from behavior change to mental health benefit.
Conclusions: This study introduces a novel dual-persona framework for digital health: an AI “friend” persona that fosters engagement through natural, relational conversation, paired with a distinct rules-based health coach persona that delivers evidence-based guidance without compromising the authenticity of the friend relationship. Supported by human moderation for safety and quality, this architecture enables both trust and adherence in ways that traditional coaching apps and chatbots struggle to achieve. By demonstrating measurable, sustained improvements in sleep behavior with this arrangement, our findings highlight not only a scalable pathway for behavior change but also a new paradigm for designing AI-mediated health interventions.
From Backgrounds to Trust: National Study on Telemedicine Professionalism and Patient Perceptions
Bree Holtz, PhD
Michigan State University
Description: Telemedicine professionalism extends beyond medical expertise to include nonverbal cues, such as the provider's background. This abstract presents a national survey study, building on experimental findings that backgrounds influence patient perceptions of credibility. Preliminary design and early insights highlight implications for patient trust, satisfaction, and the development of telehealth training and guidelines.
Abstract: Telemedicine has become a sustained mode of care delivery, yet guidance on how providers should present themselves in virtual environments remains limited. Previous experimental work demonstrated that telemedicine backgrounds significantly influence perceptions of professionalism and credibility: providers framed by clinical or home office settings were rated as more expert, competent, and ethical than those in plain, neutral environments. To expand these findings, we developed a national survey study examining how patients from diverse backgrounds evaluate provider professionalism in telemedicine encounters. The survey assesses the influence of visual factors—such as background, provider attire, and setting—while also capturing demographic variation, prior telehealth use, and patient expectations of trust and satisfaction. Data collection has been completed (N ≈ 640), and analysis is underway. This study addresses the need for evidence-based recommendations on visual professionalism in telemedicine and will offer broader insights than prior student-focused research. Findings will help inform training programs, support professional standards, and guide providers in creating telemedicine environments that foster patient trust, reduce uncertainty, and improve overall care quality.
Classification of Research: Patient Experience
Method: Survey/Qualitative
Results: An experimental study demonstrated that telemedicine backgrounds strongly shape perceptions of credibility. Providers appearing against a plain white background were rated significantly lower in terms of expertness, competence, and ethicality compared to those in clinical or home office settings. These findings suggest that neutral or decontextualized environments diminish professionalism cues. Building on these results, a national survey study has been completed to capture patient perceptions across diverse demographics and telehealth experiences. While data analysis is ongoing, preliminary design insights confirm that patients are highly sensitive to visual professionalism cues in virtual health care encounters.
Conclusions: Visual context matters in telemedicine. Early evidence suggests that provider backgrounds serve as nonverbal cues, directly influencing perceptions of trust, competence, and ethicality. The forthcoming national survey will expand this evidence base, identifying how demographic differences, prior exposure to telehealth, and environmental factors influence patient confidence. These insights are essential for developing training modules, professional guidelines, and policy recommendations that ensure consistent, high-quality patient experiences in virtual care. By framing professionalism as both clinical expertise and communicative environment, this research underscores how visual presentation can help reduce patient uncertainty and strengthen trust in telehealth delivery.
From Intake to Visit: Evaluating Registration Modality on Telehealth Engagement for Opioid Use Disorder
Renoj Varughese, MD1, Joshua Lynch, DO, Professor of Emergency and Addiction Medicine1, Tiffany Jenzer, PhD, Assistant Research Professor,1 Kate VanRensselaer, BS, Director of WNY Opioid Addiction Treatment, BestSelf Behavioral Health, Inc., Michael Hyde, DMSc, PA, Physician Assistant,1 Kailyn Tomilin, MPH, Program Evaluator,1 Brian Clemency, DO, MBA, Vice Chair - Academic & Faculty Affairs1
1Department of Emergency Medicine, University at Buffalo
Description: This retrospective study evaluated visit completion for patients presenting via phone hotline, online form or crossover group: online request patients transitioned to the hotline. Patients from the hotline were more likely to complete their visits, compared to patients coming through the online form and the cross over group.
Abstract: Early in the COVID-19 pandemic, online clinics arose to help treat patients with opioid use disorder (OUD). An online bridge clinic was started for New York OUD patients. As the public health emergency continued, the US Drug Enforcement Agency allowed patients with OUD to be treated with buprenorphine after an audio only evaluation. In response, an OUD treatment hotline was created for Western New York (WNY). While the hotline and the online clinic are staffed by the same individuals, they had different funding models that permitted different modalities. Patients who requested an evaluation via the hotline would complete registration via a phone call;patients presenting to the online clinic would request an appointment via an online form. Online clinic patients were required to complete audio-video visits;hotline patients were permitted audio only visits. Patients in WNY who presented to the online clinic after filling out an online form were crossed over to the hotline;patients outside of WNY remained in the online clinic group. This creates a crossover group of WNY patients who registered via an online form but were subsequently permitted audio only visits. This study evaluates the likelihood of visit completion for patients presenting through the phone hotline vs traditional online clinic vs the crossover group.
Classification of Research: Access to Care
Method: Observational
Results: There were 4,209 visits requests, 65.5% in the hotline group, 29.9% in the clinic group, 4.6% in the crossover group. A chi-square test of independence was done to examine the association between the patient’s intake group and the visit completion. Patients coming through the hotline were more likely to complete their visit (89.4% completed), compared patients coming through the online clinic (78.1% completed), and the crossover group (78.2% completed), p< 0.001.
Conclusions: The hotline group had higher visit completion rates compared to the other two groups. Even though the cross over group was allowed to complete visits through audio only, the completion rate was still lower than the hotline. It may be that calling for an appointment selects for patients who are likely to complete the visit. An online form may require less engagement. Furthermore, the hotline includes an operator who often reassures and addresses barriers (a form of motivational interviewing) that the online form cannot, which may impact patient engagement. Research like this can inform future designs of patient programs. More research is needed to better understand the benefits of a hotline.
Optimizing Clinical Integration of Virtual Reality Visual Field Devices: A Comparative Study at Bascom Palmer Eye Institute
Sarah Tajran, MD, MS, Tele-Ophthalmology Research Fellow1, Karla Casanova, MD, Tele-Ophthalmology Research Fellow1, Giselle Ricur, MD, MS, MBA, FATA, Executive Director Virtual Eye Care
1Bascom Palmer Eye Institute
Description: This study evaluates integration strategies for Heru Prime and Olleyes VisuAll virtual reality (VR) visual field devices in the outpatient workflow at Bascom Palmer Eye Institute. By comparing technician-led and dedicated “VR Imager” workflows against the gold standard Humphrey Visual Field, we assessed efficiency, technician burden, and patient throughput to guide clinical implementation.
Abstract:Purpose: To evaluate the clinical integration of two VR visual field devices, Heru Prime and Olleyes VisuAll, into the outpatient workflow at Bascom Palmer Eye Institute Lennar Foundation Medical Center, focusing on test efficiency, technician burden, and patient throughput.
Methods: Patients underwent visual field testing using either Heru Prime, Olleyes VisuAll, or the standard Humphrey Visual Field (HVF). Following technician workup, a dedicated VR Imager guided patients through headset use and testing. Workflow intervals were recorded, with visual field duration defined from “VR With” to “VR Complete.” Preliminary data were analyzed using descriptive statistics and one way ANOVA with pairwise t tests to compare test durations across devices.
Results: Sixty two tests were analyzed (Heru: n=29;Olleyes: n=27;HVF: n=6). Mean test times were: Heru Prime – 14.1 ± 4.9 minutes; Olleyes – 11.2 ± 4.5 minutes; HVF – 19.7 ± 7.6 minutes. ANOVA showed differences in mean durations (p = 0.0011). Olleyes had shorter test times compared with both Heru (p = 0.023) and HVF (p = 0.039);Heru and HVF were more comparable (p = 0.13).
Conclusion: Although HVF data collection is ongoing, early findings indicate VR devices, Olleyes and Heru, may improve workflow efficiency. Continued testing will help clarify their clinical value in high volume ophthalmology settings.
Classification of Research: Quality Improvement
Method: Descriptive
Results: A total of 62 tests were analyzed (Heru Prime: n=29; Olleyes: n=27; HVF: n=6). Mean test durations were 14.1 ± 4.9 minutes for Heru Prime, 11.2 ± 4.5 minutes for Olleyes, and 19.7 ± 7.6 minutes for HVF. ANOVA demonstrated significant differences across groups (p = 0.0011). Pairwise comparisons showed Olleyes significantly faster than Heru Prime (p = 0.023) and HVF (p = 0.039), while Heru vs HVF was not significant (p = 0.13). These results indicate that VR devices, particularly Olleyes, improve efficiency compared with standard HVF, reducing technician burden and enhancing patient throughput.
Conclusions: This study demonstrates the feasibility and efficiency of integrating VR visual field devices into the outpatient workflow at a tertiary eye institute. Both Heru Prime and Olleyes reduced test duration compared with the standard HVF, with Olleyes providing the shortest and most consistent times. These improvements suggest that VR-based testing can decrease technician burden and enhance patient throughput, supporting adoption in high-volume clinics. While sample size for HVF was limited and preliminary, findings highlight VR platforms as practical, patient-friendly alternatives that may streamline clinical operations and improve access to visual field testing.
Redesigning Access: A Virtual Approach to Specialty Care Delivery
Dunc Williams, PhD, Emily, Warr, MSN, Administrator, Center for Telehealth, Caitlin, Koob, PhD, Research Associate Faculty
Medical University of South Carolina
Description: This study uses multiple data sources, including electronic medical record data, to evaluate the impact of a virtual specialty program and a comparison group of in-person clinic visits. By examining key performance indicators, we explore how to reduce barriers to specialty care, enhance multi-level outcomes, and improve efficiency.
Abstract: Patients often face significant barriers to accessing in-person specialty care, leading to long delays and poor health outcomes. Waitlists for specialty appointments can extend up to six months. Health care providers in in-person settings frequently experience high workloads, limited flexibility, and resource constraints that contribute to dissatisfaction, emphasizing the need for alternative care models to support provider well-being.
To address these challenges, we developed and implemented a virtual specialty service line aimed at improving access and reducing wait times, while considering patient and provider preferences for delivery mode. This innovative model offers fully virtual care across high-demand specialties and primary care, while coordinating in-person services such as labs, imaging, and pharmacy within the patient’s local community. Key outcomes examined in this study include provider and patient satisfaction, provider relative value units, Current Procedural Terminology codes, average appointment wait times, and percentage of new and returning visits. We categorize each visit as either having an in-person clinic provider or a virtual specialty provider, and compare outcomes across these four provider categories. The goal of this study is to evaluate the impact of implementing a virtual specialty program on physician productivity, provider satisfaction, and patient access to specialty care.
Classification of Research: Access to Care
Method: Descriptive
Results: Occurring between January and June 2025, our sample includes 65,920 visits, 6,588 virtual, and 59,332 in-person. Blue Cross Blue Shield was the most common payer for virtual (40%), while Medicare covered 49% of in-person patients. Neurology was the most common in-person specialty (26%), and Endocrinology led virtual visits (28%). CPT Code 99214 (established patient moderate complexity) was the most used across both types (37% of virtual visits and 35% of in-person). 22% of virtual visits used CPT code 99204 (new patient), and the 2nd most common (14%) for in-person was 99215 (established patient high complexity).
Conclusions: Offering a virtual-only specialty service can significantly enhance access for patients, especially in underserved or remote areas. Our data shows high utilization of virtual visits for specialties, suggesting strong demand. While specific procedures require in-person care, virtual specialty services may expand access to new patient populations, improve continuity, reduce wait times, and optimize resource allocation. Strategic implementation could lead to more effective health care delivery and better patient outcomes. Ongoing analysis will examine physician productivity, visit outcomes (e.g., cancellation, no-show rates), patient satisfaction, and provider satisfaction across virtual and in-person groups.
Telehealth Utilization and Effectiveness within the U.S. Health Care System
Whitney Garney, PhD, MPH, Kala Reindel, PhD, Research Scientist, Sanjeela Gandhari, MPH, Graduate Assistant, Research, Amyia Harris, M.Ed. Graduate Assistant – Research, Tyler Watson, M.Ed., Graduate Assistant - Research, Rachana Talekar, MPH, Program Manager, Carly McCord, PhD, Executive Director
Texas A&M Telehealth Institute
Description: This literature review documents the extent to which different medical disciplines in the United States (U.S.) health care system utilize telehealth. Results describe the extent to which each discipline engages in telehealth, what health conditions are frequently treated, and telehealth effectiveness compared to treatment as usual.
Abstract: To promote the use of telehealth in the U.S., it is important to understand how different medical disciplines utilize telehealth and how effective these visit types are at meeting clinical and patient needs. This scoping study identified systematic literature reviews (SLR) reporting on telehealth effectiveness across all medical disciplines in the U.S. Guided by PRISMA, we used MEDLINE to search for SLRs reporting on the use and effectiveness of telehealth in the US health care system. Studies were included if 1) the sample primarily based in the U.S. and 2) if they assessed clinical effectiveness of a telehealth intervention with a control. 5,677 papers were identified with 42 duplicates; 5,635 papers were screened with 3,969 excluded due to topical irrelevance. Two researchers did a full text review of 1,666 papers, and 1,606 were excluded because they didn’t meet inclusion criteria. 60 SLRs were included in the analysis. Data was extracted using a Covidence form. Fields captured medical specialty, modality, remote patient monitoring used (if applicable), study population information, geography, medical condition treated, clinical outcome(s) measured, and outcome effects.
Classification of Research: Clinical Effectiveness
Method: Literature Review
Results: The majority (40%) of SLRs were in Psychology/Psychiatry, followed by Primary Care (17%), and Pharmacy (10%). The primary health conditions assessed were depression/anxiety, hba1c, blood pressure, and glycemic control. All SLRs within Endocrinology, Cardiology, Pulmonology, and Critical Care reported positive outcomes. Telehealth was least effective in Obstetrics with only 25% of SLRs reporting positive outcomes, then Pharmacy and Pediatrics with 50% of SLRs reporting positive outcomes. 87.5% Psychology/Psychiatry and 80% of Primary Care SLRs reported positive outcomes. 15 (25%) SLRs utilized remote patient monitoring. The primary devices were at-home blood pressure cuffs (33%), glucometers (27%), and fitness trackers (20%).
Conclusions: Based on this review of SLRs, health care systems should consider telehealth for Endocrinology, specifically to treat/manage diabetes, as well as Cardiology for both medication adherence and cardiovascular disease management. This review confirms the effectiveness of telehealth in the Psychology/Psychiatry discipline. There is potential to improve telehealth utilization in Primary Care, which frequently uses telehealth to treat obesity concerns. Providers may choose to avoid telehealth for Obstetrics based on mixed findings of effectiveness. Lastly, there is great potential for telehealth utilization in Pulmonology and Cardiology, but additional research should be conducted to increase evidence.
Telehealth-First Obesity Care Across Multi-Site Clinics: Sustained Clinical Outcomes and New Revenue Pathways
Mohammed Rami Bailony, MD, CEO and Co-Founder, Patricia Dizon, Pharmacist, Executive Assistant
Enara Health
Description: We implemented a telehealth-first obesity and cardiometabolic program across diverse clinics. Patients achieved 15–17% sustained weight loss with improvements in HbA1c, blood pressure, and LDL. Insurance reimbursed >80% of claims, generating $311 per patient per month and $3M in annualized revenue for one cardiology practice, proving both clinical and financial sustainability.
Abstract: Background: Telemedicine has been widely adopted for acute care, but its role in chronic disease management is less defined. Obesity, a key driver of cardiometabolic disease, requires longitudinal, multidisciplinary treatment that can be optimized through virtual delivery.
Methods: We conducted an implementation study of a telehealth-first obesity platform embedded within multiple clinical settings, including cardiology and primary care practices. The intervention included physician-directed care, diet and exercise therapy, behavioral health support, optional anti-obesity pharmacotherapy, and connected home scales. Encounters were primarily delivered via telehealth, supported by a digital care coordination platform. Outcomes included sustained weight loss, cardiometabolic improvements, reimbursement rates, and practice-level financial performance.
Results: Across >2,400 patients, average sustained weight loss was 15–17% at 18 months, with significant reductions in HbA1c (-1.1%), systolic blood pressure (-8 mmHg), and LDL (-14 mg/dL). More than 70% maintained ≥10% weight loss long term. In a cardiology practice subanalysis (n=792), 83% of claims were reimbursed, with an average $311 collected per patient per month, translating into >$3M in new annualized revenue. Clinicians reported reduced burden due to optimized digital workflows.
Conclusion: Telehealth-first delivery of obesity care produces surgery-level outcomes, measurable cardiometabolic improvements, and a financially sustainable, reimbursable service line—demonstrating the potential of digital-native models to transform chronic disease management.
Classification of Research: Clinical Outcomes
Method: Implementation Science
Results: Patients achieved sustained 15–17% weight loss at 18 months, with significant improvements in HbA1c (-1.1%), systolic blood pressure (-8 mmHg), and LDL cholesterol (-14 mg/dL). Over 70% maintained ≥10% weight loss, and >40% achieved ≥15% loss. In a cardiology subanalysis (n=792), 83% of claims were reimbursed, generating $311 per patient per month and >$3M annualized practice revenue. Clinicians reported reduced workload through telehealth-enabled workflows.
Conclusions: Telehealth-first obesity care can deliver sustained clinical outcomes while creating new, profitable service lines for clinics. By embedding multidisciplinary care within telehealth infrastructure, practices achieved both improved patient health and reliable reimbursement. This model reduces clinician burden and proves scalable across diverse settings, offering a replicable framework for digital-native chronic disease management.
Telemedicine Service in India is Poised for a New Transformation in Terms of Structured and Cost-effective Business/Revenue Model
Satyamurthy Lakkavalli, Ravi Amble, CEO
Suquino Teleheath LLP (India)
Description: Telemedicine service in India is poised for a new transformation. With the business analytics study and applications, an optimized and viable revenue model was evolved for application for telehealth care service, judiously amortizing the costs connectivity, software cloud, operation & maintenance and call center support respectively, are the key cosiderations.
Abstract: Two decades of telemedicine movement in India faced a roller coaster journey with the challenges of technology evolution, demonstration, and standardization by Government of India and executed through Indian Space Research Organization and Department of Information Technology.
The fast developments and later retardation in Telehealth service are the hallmarks of implementation process in any country because of inherent mind set/orthodoxy among stake holders, compounded by technology obsolescence, ambiguous Business models and uncertain regulatory frame work.
Growth of several telemedicine centers all over India with major specialty hospitals providing tele-consultation and treatment to various rural remote district hospitals/primary care centers in many States, showed the potential thrust of telehealth during the first decade of telemedicine implementation in India. Examples are:
National Tele-Oncology, Tele-ophthalmology, Pan-Africa Telemedicine networks by the Government of India.
State wide telemedicine networks connected to Specialty Hospitals/Medical Institutions.
Successful implementation of COVID telehealth care under a PPP model, a major gain from telemedicine.
Business analytics study and applications ushered a new direction. An optimized and viable revenue model was evolved for future application for telehealth care service by diligently amortizing the costs covering connectivity, software cloud, operation & maintenance, and call center support respectively are highlighted.
Classification of Research: Access to Care
Method: Observational
Results: An optimized and viable Business/Revenue model has been evolved as highlighted below.
Government run hospitals to provide free or subsidized health care through existing government authorized insurance agencies under cash less program.
Government health care providers can contract corporate/private hospitals to provide telehealth care and charge per consultation basis,
Private/corporate hospitals with their established revenue model can provide telehealth care to affordable beneficiaries through private/corporate insurance agencies as applicable.
Smaller private hospitals/clinics to have a tailored software as service revenue model wherein a part of cost per consultation can be made applicable to both technology and clinical service providers.
Conclusions: An optimum revenue model has been evolved based on the study of business analytics as applicable for Indian environment for future application for telehealth care services, diligently amortizing each of cost covering connectivity, software cloud, operation, maintenance and call center support respectively. This effort has been showing good results and being applied progressively.
As health care is the primary responsibility of any government administration for societal benefits, many countries have understood the need of providing financial support with special emphasis for telehealth care. Indian approach of trade off in evolving suitable business model can usher better promise in future.
Transforming Women’s Health Through Tech-Enabled Care: Evidence from InovCares’ Texas Collaboration
Jason Wallace,1 J’Nae Taylor, MS2, Maternal Care Navigator, Chelsea Cooper, BS,2 Research Assistant, Lovemore Chirombo, BS,2 Chief Technology Officer, Pelumi Adedayo, MD,2 Chief Medical Officer, Henry Imperial, MD,1 Chief Medical Officer, Taibat Eribo MD, Medical Director, HHM Health; Kyrah K. Brown, PhD, Associate Professor, University of Texas at Arlington
1New Horizon Health Center, 2InovCares
Description: This study evaluates InovCares, a tech-enabled maternal health platform implemented in Texas Federally Qualified Health Centers (FQHC). It improved clinical outcomes and patient satisfaction through digital care coordination, social support services, and bilingual engagement. High app usage and transportation assistance highlight its scalable, equity-driven model for advancing maternal health in underserved communities.
Abstract: Texas continues to face significant maternal health disparities. This paper evaluates the implementation of InovCares, a mobile-first maternal health platform deployed across two FQHCs in partnership with a managed care organization. Over a 12-month period, InovCares combined remote biometric monitoring, digital care coordination, and services addressing social determinants of health such as transportation and housing. Among 605 enrolled pregnant individuals, two-thirds continued pediatric care post-delivery. Clinical outcomes showed reduced emergency department visits, lower preterm birth rates (7%), and fewer cesarean deliveries (22%) compared to state benchmarks. Patient satisfaction data revealed consistently high ratings for staff friendliness, app engagement, and transportation support. Over 90% of respondents used the app weekly or more, and nearly all expressed appreciation for bilingual interfaces and culturally responsive care. Key lessons included the importance of integrating care navigation into existing workflows, leveraging bilingual digital tools, and forming strategic payer partnerships to support reimbursement for non-clinical services. InovCares demonstrates a scalable, equity-driven model that aligns with value-based care priorities and offers a replicable framework for improving maternal health outcomes in underserved communities.
Classification of Research: Clinical Effectiveness
Method- Other: Intervention Study
Results: The InovCares pilot in Texas showed encouraging signs of improved maternal health outcomes and patient experience. Among 605 participants, emergency visits declined, and rates of preterm birth (7%) and cesarean delivery (22%) were lower than state averages. Two-thirds continued pediatric care post-delivery. Survey responses reflected generally high satisfaction, with frequent app use and appreciation for bilingual support and transportation assistance. Lessons learned emphasized the value of integrating care navigation, fostering payer collaboration, and using culturally responsive digital tools. While further evaluation is needed, these results suggest that InovCares offers a promising framework for advancing maternal health equity in underserved settings.
Conclusions: The implementation of InovCares across two Texas FQHCs yielded promising results in addressing maternal health disparities. The platform’s integration of digital tools and social support services contributed to improved clinical outcomes, including reduced emergency visits and lower rates of preterm birth and cesarean delivery. Patient feedback highlighted strong satisfaction with bilingual support, transportation assistance, and app usability. Continued pediatric care post-delivery and frequent app engagement suggest sustained impact. Lessons learned underscore the importance of workflow integration, culturally responsive design, and payer collaboration. Overall, InovCares offers a replicable, equity-focused model for enhancing maternal health in underserved settings through tech-enabled care.
Virtual Multidisciplinary Gastrointestinal Care (MGC) for Adults with Gastrointestinal (GI) Needs: a Retrospective Cohort Study
Grace Wang, PhD, MPH,1 Sanskriti Varma, MD, Faculty Gastroenterologist, Division of Gastroenterology, Department of Medicine, Massachusetts General Hospital, Harvard Medical School, Boston, USA, Sameer K. Berry, MD, MBA,1 CoFounder and Chief Medical Officer, Director of Health care Economics Research and Clinical Assistant Professor of Medicine, New York University Grossman School of Medicine
1Oshi Health
Description: Highly prevalent gastrointestinal (GI) disorders significantly decrease quality of life and are a considerable cost to the US health system. While multidisciplinary GI care (MGC) has proven efficacious in clinical trials, its effectiveness and ability to be scaled outside of controlled settings remain uncertain. The study describes the delivery of virtual MGC at scale, and its objective is to demonstrate its effectiveness and describe patient characteristics, outcomes, and the relationship between engagement and symptom improvement.
Abstract: We conducted a retrospective cohort study of patients who received virtual MGC between April 2021 and August 2025. Inclusion criteria for this analysis were: provided intake data;completed onboarding appointment with a GI advanced practice provider, received team-based care with access to GI providers, registered dieticians, and behavioral health providers depending on patient need;and reported on symptom improvement and control after receiving virtual MGC. We collected data on patient demographics, engagement (e.g., time to visits, number and duration of visits, number and duration of calls greater than 5 minutes, number of patient-initiated messages), satisfaction, and symptom-related patient-reported outcomes. We used descriptive statistics, paired t-tests, and logistic regression models to analyze the data.
Classification of Research: Clinical Effectiveness
Method: Observational
Results: The study’s 9,405 patients had a mean age of 43.45 (±12.34) years and were predominantly female (65.92%). The median time to first scheduled visit was 6 days (interquartile range [IQR]: 3-9). The median number of visits and calls to GI providers was 2 (IQR 1-3), while registered dietitians and behavioral health providers saw a median of 2 (IQR 1-3) and 1 (IQR 0-2) visits, respectively. On average, patients initiated 9.32 (±19.09) chat messages with their virtual MGC team. We found that 93.55% reported symptom control, and 91.06% of patients reported symptom improvement. Patient satisfaction with virtual MGC was high, with a mean score of 4.85 (± 0.50) out of 5. In adjusted regression models, increased engagement with the care team was significantly associated with symptom improvement. Patients with four or more GI engagements had 3.57 times higher odds of symptom improvement compared to those with only one (95% Confidence Interval [CI] 2.29, 5.57). Patients with three or more behavioral health engagements had 2.09 times higher odds of improvement (95% CI: 1.37, 3.18), while those with three or more dietitian engagements had 11.34 times higher odds (95% CI: 7.45, 17.27) compared to those with no engagements with these providers.
Conclusions: This study is the first to demonstrate the delivery of virtual MGC at scale to a large, diverse patient population across the U.S. Our findings show that this model is not only feasible but also effective, as patients experienced significant symptom improvement and reported high satisfaction with their care. The observed association between increased patient engagement and improved outcomes highlights the value of the multidisciplinary approach, reinforcing findings from controlled clinical trials. The ability of virtual MGC to achieve these results outside of a clinical trial setting suggests that it is a viable solution to access challenges in GI care.
Poster Presentations
A Retrospective, Claims-Based Evaluation of Total Cost of Care in a Virtual Women’s Health Program
Description: A retrospective claims analysis of 600 women in a virtual women's health program demonstrated a $3,429 reduction in total costs per member over six months compared to matched controls. The integrated care model reduced specialist visits, hospitalizations, and emergency visits while improving outcomes. This offers a scalable approach for health systems addressing fragmented women's care.
Abstract: Women experience a disproportionate burden of chronic, multisystem, and life stage–dependent health conditions, yet care delivery remains fragmented, episodic, and narrowly focused on reproductive events. This misalignment contributes to delayed diagnosis, escalation to high-cost procedural care, avoidable utilization, and persistent barriers to accessing coordinated, condition-specific treatment. As health care systems face rising costs and workforce strain, scalable models that comprehensively address women’s health while improving outcomes and controlling total cost of care are urgently needed.
We conducted a retrospective, propensity score–matched cohort study evaluating a comprehensive virtual women’s health care and navigation program deployed within a commercial payer population. Deidentified claims data from May 2023 through May 2025 were analyzed. Women receiving care through the virtual program were matched to controls receiving usual care based on state, age, baseline utilization and cost, medication counts, behavioral health utilization, and high-cost condition burden. Outcomes were assessed using a difference-in-differences framework comparing health care utilization and spending during the six months before and after program initiation. Prespecified subgroup analyses examined surgical and non-surgical complex gynecologic populations.
A total of 600 women receiving virtual comprehensive care were matched to 2,941 controls. Program participation was associated with a $3,429 reduction in total allowed costs per member over six months (P=0.02), driven primarily by lower facility, professional, and gynecologic spending. Participants experienced fewer specialist visits, inpatient admissions, outpatient encounters, and emergency department visits, alongside a relative increase in primary care utilization. Subgroup analyses demonstrated savings of $2,197 per member in the surgical complex gynecologic cohort (P=0.02), driven by fibroids and heavy menstrual bleeding, and $3,635 per member in the non-surgical cohort (P<0.001), driven by menopause. High patient engagement and satisfaction further supported the effectiveness of coordinated, virtual care delivery.
These findings demonstrate that integrated virtual women’s health care can improve clinical outcomes while reducing health care utilization and total cost of care in real-world commercial populations. The model is scalable and may be implemented by other organizations seeking sustainable strategies to improve access, advance health equity, and support value-based care.
Classification of Research: Cost Analyses
Method: Cost Analysis
Results: Among 600 program participants matched to 2,941 controls, enrollment in comprehensive virtual women’s health care was associated with a $3,429 reduction in total allowed costs per member over six months (P=0.02), driven by decreased facility, professional, and gynecologic spending. Participants had fewer specialist visits, inpatient admissions, outpatient encounters, and emergency department visits, with increased primary care utilization. Subgroup analyses showed savings of $2,197 per member in the surgical complex gynecologic cohort (P=0.02) and $3,635 per member in the non-surgical cohort (P<0.001). High engagement and satisfaction accompanied these outcomes.
Conclusions: A comprehensive virtual women’s health care and navigation program was associated with meaningful reductions in total cost of care and health care utilization while improving patient engagement and outcomes. By addressing women’s health needs earlier and more holistically, this scalable telehealth model reduces reliance on high-cost episodic care and supports value-based delivery. These findings suggest integrated virtual women’s health programs represent an effective approach to improving access, advancing health equity, and achieving sustainable cost savings across commercial populations.
AI-powered Markerless Motion Capture in Rehabilitation: Enhancing Patient Experience and Engagement
Shayen Bhatia, Luke Riggan, BS, PhD
Prime Health Care
Description: This study evaluates the impact of a rehabilitation-specific AI-powered markerless motion capture system on patient experience and satisfaction across two hospital-based rehabilitation clinics.
Abstract: Digital health technologies are rapidly transforming rehabilitation, offering opportunities to enhance outcomes, improve patient engagement, and increase accessibility. Among these, markerless motion capture represents a major advancement, enabling clinicians to analyze movement without sensors or cumbersome setups. While technical accuracy has been established, less is known about how patients experience such technologies. Patient-reported outcomes are increasingly recognized as critical clinical trial endpoints because they capture dimensions of care that matter most to patients. Affordable, scalable solutions that enhance engagement and objectively measure progress are urgently needed to improve patient experience, adherence, and outcomes across diverse populations.
An AI-powered, markerless motion-capture platform was implemented in two hospital-based rehabilitation clinics. The system uses a single consumer-grade camera and proprietary neural networks trained in rehabilitation-specific movements to detect joints, quantify motion quality, and provide real-time visual and auditory feedback. This pilot evaluated the platform’s impact on patient experience and usability as part of standard therapy sessions. Seventy-nine patients participated across 150 encounters, representing a range of neurological and musculoskeletal conditions. Patients completed 214 post-session surveys evaluating motivation, confidence, usability, and perceived support using a 0–4 Likert scale. Therapists provided structured feedback regarding ease of integration, patient engagement, and clinical workflow impact.
Classification of Research: Patient Experience
Method: Survey/Qualitative
Results: Feedback was highly positive. Most patients reported benefits across multiple domains: 77% improved movement control, 82% greater confidence, 88% higher motivation, and 75% functional gains in strength, balance, or range of motion. Usability was rated positively by 95%, with 98% citing strong staff encouragement. Among 16 therapist surveys (∼9 providers), all agreed patients valued and easily used the technology, though only 40% perceived reduced anxiety compared with 72% of patients. Therapists highlighted its value for progress tracking and engagement, especially for gait assessment and motivation in acute rehabilitation settings.
Conclusions: AI-powered markerless motion capture effectively transforms rehabilitation into an engaging, data-driven, and motivating process. By providing objective feedback and progress visualization through simple, scalable technology, it directly addresses long-standing challenges of disengagement and inequitable access. The results demonstrate that digital health innovations can meaningfully improve patient experience, therapist satisfaction, and functional recovery, supporting a new paradigm of accessible, AI-enhanced rehabilitation. This AI-powered innovation demonstrates how digital technology can extend equitable access, reimagine rehabilitation, and improve outcomes globally. Larger studies incorporating both patient and provider perspectives are warranted to confirm these results.
Association Between Telemedicine Use and Hospitalization Among US Adults with Diabetes: Exploring the Role of Comorbidity Burden
Description: Telemedicine expanded exponentially for diabetes management during the pandemic, yet the relationship between telemedicine use and health care utilization in real-world settings remains poorly understood. Up-to-date national studies examining whether this association differs by patient comorbidity burden are needed to inform evidence-based telemedicine policy as temporary coverage flexibilities expire.
Abstract: Objective: To assess the association between telemedicine use and inpatient care utilization among U.S. adults with diabetes, adjusting for comorbidity burden and exploring clinical complexity as an effect modifier.
Methods: We conducted a retrospective cohort study using longitudinal and office-based visit files from the 2022-2023 Medical Expenditure Panel Survey. The primary outcome was any inpatient expenditure (Year 2);primary exposure was any telemedicine visit (Year 1). Comorbidity burden was measured using MEPS priority chronic conditions, categorized as low (≤2 conditions) versus high (≥3 conditions).
We used weighted logistic regression models to assess main effects and effect modification using an interaction term (telemedicine × comorbidity burden), adjusting for sociodemographic factors (age, sex, region, race/ethnicity, insurance) and health-related variables (BMI, self-rated health).
Classification of Research: Clinician Experience
Method: Secondary Data Analysis
Results: Among 869 unweighted (27.9 million weighted) participants, 11.9% used telemedicine and 53.3% had high comorbidity burden. Mean inpatient expenditure was $9,525 for telemedicine users versus $5,023 for non-users. After adjustment, telemedicine users had 2.84 times the odds of hospitalization (p<0.05);those with ≥3 comorbidities had 2.28 times the odds (p<0.05). The interaction term approached significance (p=0.089). Exploratory stratified analyses revealed no association among low-comorbidity patients (OR=1.55, p=0.54) but strong association among high-comorbidity patients (OR=4.85, 95% CI: 2.2-10.9, p<0.001).
Conclusions: Comorbidity burden partially explained telemedicine-hospitalization association but did not fully account for them. Exploratory analyses suggest the relationship may differ by clinical complexity, with stronger associations among patients with multiple chronic conditions, though limited statistical power (interaction p=0.089) precludes definitive conclusions. The pattern is consistent with appropriate targeting of telemedicine to clinically complex patients requiring intensive management. These hypothesis-generating findings warrant replication in larger studies with detailed clinical measures. As policymakers consider permanent telemedicine coverage expansion, higher utilization among telemedicine users may reflect appropriate patient selection rather than adverse effects of virtual care.
Equity-Centered AI Tool to Mitigate Digital Bias and Support Behavioral Change in Diabetes Care
Cynthia Williams, PhD, MHA, PT
University of Central Florida
Description: This proposed mixed-methods study aims to develop a conceptual equity-centered digital tool to mitigate AI-driven bias in diabetes care. By integrating inclusive data, community input, and culturally responsive design, the tool supports behavioral and dietary changes while ensuring fair, personalized outcomes for vulnerable populations often excluded from digital health innovations.
Abstract: Artificial intelligence tools are increasingly used to support diabetes management. However, these innovations often perpetuate systemic inequities due to digital bias, which stems from non-representative data, algorithmic exclusion, and culturally unresponsive design. Vulnerable populations, including racial and ethnic minorities, low-income individuals, and those experiencing food insecurity, are disproportionately affected. These same populations face significant barriers to adopting the behavioral and dietary changes essential for effective diabetes control. This study is grounded in a theoretical framework that proposes discussing how to develop an equity-centered tool to reduce bias and encourage behavior change when health determinants may make this challenging. Utilizing the Health Equity Implementation Framework is crucial in guiding the ethical and inclusive application of artificial intelligence in health care. It ensures that AI tools are developed and implemented with attention to structural inequities, community context, and social determinants of health, helping prevent algorithmic bias and promote equitable health outcomes. The tool will be grounded in equity-centered design and informed by the lived experiences of marginalized communities.
Classification of Research: Access to Care
Method- Other: Mixed Method
Results: The EquiCare Diabetes app is a conceptual smartphone tool designed to reduce digital bias and promote equitable diabetes care. It integrates inclusive data collection, culturally responsive behavioral and dietary guidance, and real-time fairness checks to ensure recommendations are unbiased across racial, ethnic, and socioeconomic groups. Grounded in the Health Equity Implementation Framework, the app addresses structural barriers and social determinants of health by offering multilingual support, offline access, and community-driven content. It empowers users to make sustainable lifestyle changes while ensuring artificial intelligence systems are transparent, representative, and accountable—ultimately advancing equity in digital health outcomes for vulnerable populations.
Conclusions: Design equity-focused smartphone tool designed to reduce digital bias and promote equitable diabetes care through inclusive data practices, culturally responsive guidance, and fairness-driven artificial intelligence. Grounded in the Health Equity Implementation Framework, it addresses structural barriers and social determinants of health while empowering users to make sustainable changes in their behavioral and dietary habits. Embedding equity into digital health design is essential for dismantling disparities and ensuring equitable access to care. Future research should evaluate the app’s effectiveness across diverse populations and explore its scalability for other chronic conditions and health systems.
Evaluating a Telehealth-enabled Collaborative Care Program for Behavioral Health in Rural Primary Care Practices within a Large Academic Medical Center
Caitlin Koob, PhD, MS, Research Associate-Faculty, Emily Johnson, PhD, Professor, MUSC, Andrew Alkis, MD, Center for Telehealth Psychiatrist and Clinical Assistant Professor, Candace Sprouse-McClam, PhD, LISW-CP (S), LCSW, Collaborative Care Clinical Manager, Katie Kirchoff, MSHI, Database Administrator, Jennifer Dahne, PhD, Clinical Psychologist and Professor
Medical University of South Carolina (MUSC)
Description: Psychiatric collaborative care management (CoCM) effectively addresses behavioral health in primary care;however, uptake in rural communities remains challenging. Our telehealth-enabled CoCM program seeks to improve access and quality of behavioral health services across rural South Carolina. This study aims to describe barriers and facilitators to telehealth CoCM patient enrollment.
Abstract: Purpose: To describe the barriers and facilitators to enrollment among patients referred to the CoCM program by their primary care providers, through mixed-methods evaluation.
Methods: Data were triangulated from multiple sources, including electronic health records and patient-reported experiences among those enrolled and not enrolled, via surveys (N=43) and interviews (N=5), to inform program improvements and leverage data-driven decisions. Descriptive statistics and rapid qualitative data analysis were conducted to comprehensively evaluate targeted outcomes.
Results: Over the past year, CoCM has grown to serve 44 primary care clinics statewide, compared to 18 clinics previously. Preliminary analyses suggest that CoCM is increasing reach, and more than one-third of enrollees graduated from the program (37.0%). Further, enrolled patients largely experienced improvements in their mental health since joining CoCM (62.5%) and highlighted benefits of the telehealth-enabled delivery, accessibility, flexibility, and program support. The majority of those who did not enroll reported they did not remember receiving program information from their primary care provider (58%), but a similar program would be beneficial to their health and well-being (68%).
Conclusions: Further work is needed to describe the utility of the CoCM program, understand its impact on patients’ health care access, and evaluate the sustainability of this model.
Classification of Research: Quality Improvement
Method: Implementation Science
Results: From June 1, 2023 through June 30, 2024, 303 patients were referred to CoCM and enrolled (58.4% of total referred patients), and 112 of enrolled patients (37.0%) graduated from the program. Patients participated in CoCM for a mean of 164 days. Enrollees largely experienced improvements in their mental health since joining CoCM (62.5%) and highlighted benefits of the telehealth-enabled delivery, accessibility, flexibility, and program support. The majority of those who did not enroll reported that they did not remember receiving program information from their primary care provider (58%), but a similar program would be beneficial to their health and well-being (68%).
Conclusions: Feedback from referred patients, including those enrolled and not enrolled, informs data-driven decisions to improve accessibility, patient outcomes, and long-term scalability. Last fall, CoCM transitioned from grant funding to standard billing practices, and its impact continues to grow with the program. Next steps include robust analyses of patient-reported data (i.e., sleep and mood trackers, clinical screening tools) in a patient-facing telehealth platform. Additional sustainability measures include revenue per program participant, rather than reliance on grant funding, and revenue growth rate (month-over-month) from program participants.
Evaluating the Impact of a Virtual Nursing Model on Providers within a Large Academic Medical center: Incorporating Unit-level Indicators and Nurse-reported Outcomes
Dunc Williams, PhD, Caitlin, Koob, PhD, Research Associate Faculty, Jillian, Harvey, PhD, Professor, Katie Kirchoff1, MSHI Data & Performance Manager, Peter Gardella1, MBA, RN Director of Operations/Nursing, Emily H. Warr1, MSN, Administrator, Dee W. Ford, MD, MSCR, Professor of Medicine, Division Director – Pulmonary, Critical Care, Allergy and Sleep Medicine, Program Director – National Telehealth Center of Excellence
Medical University of South Carolina, 1Center for Telehealth
Description: A Virtual Nursing Model (VRN) was implemented across an academic medical center as an innovative care delivery model, leveraging teams of remote nurses to offset time-consuming tasks. We surveyed bedside and virtual nurses (N=132) to assess their perceived value of the VRN model. Overall, nurses provided favorable perceptions of VRN.
Abstract: Systemic gaps in U.S. health care delivery have created workforce issues that impact patient outcomes and costs. As a result, there is growing concern for an estimated deficit of 3.2 million health care workers by 2026 nationwide;in one state, 39% of surveyed nurses reported intentions of leaving their job within one-year.
To address these concerns, a VRN was developed as an innovative care delivery model, leveraging teams of remote, virtual nurses to support bedside nurses in providing care via technology. This study evaluates bedside and virtual nurses’ perceived value and attitudes towards a implemented VRN model, following its pilot implementation.
From May 22, 2025 through September 9, 25, we surveyed 692 nurses (650 bedside nurses, 42 virtual only contract nurses) from MUSC’s 41 VRN units. Of the 692 nurses surveyed, 132 completed the survey, 10 completed most of the survey but did not submit it (all 10 bedside), totaling 142 surveys eligible for evaluation.
Overall, the majority of nurses’ reported benefits of a team-based approach to care (63%) and time saved, particularly related to admission documentation (61%) and more time to spend on rounding (54%). Additionally, 29% of nurses reported improved satisfaction and 27% felt the patient experience improved since VRN implementation.
Classification of Research: Clinician Experience
Method: Survey/Qualitative
Results: When asked how VRN impacted workload, 87% of bedside nurses said it slightly or significantly decreased workload. Nurses ranked various uses of virtual nursing as a 4 or 5 on a 1-5 scale (with 5 being the most impactful), including admissions (92%), discharges (36%), quality documentation/surveillance (31%), patient education (25%), and care planning (20%). Overall, nurses reported being very satisfied with the virtual nursing program 61% of the time. Comments from nurses in two free-text questions on the survey were generally favorable, supporting quantified results presented in this study.
Conclusions: This study is situated within a larger, multi-level evaluation of the VRN model’s impact on patient, provider, and systemic outcomes. Favorable perceptions of the VRN model from bedside and virtual nurses’ indicate feasibility of continued VRN expansion across innovative use-cases, including high-touch departments such as emergency departments and intensive care units.
Impact of EHR Order Changes on Telemetry Utilization: A Retrospective Review
Aline Kalaydjian, MD1,2, Research Data Scientist, Lindsay Randle, MBA, Director of Virtual Medicine; Virtual Medicine Department, Amaris Fuentes, PharmD, Manager IT Clinical Informatics, System Quality and Patient Safety, Farhan Ishaq1,2, MD, MPH, MBA, Clinical Research Fellow, Jada Malveaux1,2, MA, Clinical Research Scientist, Sarah Pletcher, MD, MHCDS1,2, Vice President of Virtual Medicine, Department of Surgery, Ngoc-Anh Anh Nguyen, MD, Director of Virtual Urgent Care1,2
Department of Medicine, Houston Methodist Hospital, Houston TX, USA, 1Center for Connected Care Innovation and Implementation-Research, 2Houston Methodist Research Institution
Description: This retrospective review assessed the impact of electronic health record (HER) order changes on telemetry use in non-ICU acute care units. Telemetry episodes briefly declined after implementation, while monitoring duration decreased overall, suggesting EHR interventions is a potential tool to increase adherence to American Heart Association (AHA) telemetry guidelines.
Abstract: Background: Inpatient cardiac telemetry is widely used in acute care settings and plays a critical role in arrhythmia detection. However, telemetry is frequently overused, contributing to unnecessary monitoring, increased operational burden, and inefficient resource utilization. EHR redesign offers a scalable approach to align telemetry use with evidence-based guidelines. To address this problem, Houston Methodist, an urban hospital in Houston, TX, implemented an EHR order change in February 2025 to automatically discontinue telemetry orders after a three-day maximum, consistent with AHA recommendations. In April 2025, a subsequent EHR modification removed telemetry from multiple default order sets to further reduce inappropriate initiation.
Methods: We conducted a retrospective review of 19,045 inpatients admitted to 31 acute care non-ICU/IMU units at Houston Methodist Hospital from 08/01/2024, to 10/31/2025. In 02/2025, EHR system order changes were implemented to auto-discontinue telemetry orders at the three-day maximum, and in 04/2025, a subsequent order change to remove telemetry from several non-indicated order sets was implemented, aligning with AHA recommendations. To accurately measure telemetry use, telemetry monitoring episodes defined as continuous monitoring periods after an initial order, and telemetry monitoring periods reflecting monitoring duration, were used. Clinical appropriateness of telemetry was not assessed.
Classification of Research: Quality Improvement
Method: Observational
Results: From August 1, 2024, to January 31, 2025, a total of 7,959 monitoring episodes were observed, with a mean (SD) of 1,326.5 (63.9) monitoring episodes per month. After February 2025 auto-discontinuation of telemetry orders and April 2025 removal of telemetry from non-indicated order sets, a total of 9,626 monitoring episodes were observed between May 1, 2025, and January 31, 2025, with a monthly average of 1,604.3 (29.5), presenting a 20.9% change. Total monthly telemetry monitoring duration decreased from an average of 6,000.7 (296.9) days per month prior to February 2025 EHR order change, to an average of 5,086.4 (231.7) days per month after 04/2025 EHR order change, reflecting a 15.2% decrease.
Conclusions: EHR-based telemetry order changes, including automatic discontinuation at day 3 in adherence with AHA guidelines and removal from non-indicated order sets, were associated with a transient reduction in total monthly telemetry monitoring duration but did not produce sustained long-term decreases in discrete telemetry monitoring episodes in acute care settings. These findings suggest that EHR-based interventions are a scalable and potential tool to improve the appropriateness of telemetry ordering. Future work will explore the clinical appropriateness of telemetry orders relative to guideline indications and evaluate complementary strategies, including targeted clinician education and outreach, to further optimize telemetry ordering practices.
Implementation of System-Wide Continuous Physiologic Monitoring and Overall Inpatient Outcomes
Jannika Machnik, MS, Farhan Ishaq, MD, MPH, MBA, Clinical Research Fellow, Sarah Sossong, BA, MPH, Farzan Sasangohar, PhD, CPAHA, Research Director – CCCIIR, Sarah, Pletcher, MD, MHCDS, VP Virtual Medicine, Ngoc-Anh Nguyen, MD, Medical Director - CCCIIR and Vice Chair of Research
Houston Methodist
Description: Houston Methodist implemented continuous physiologic monitoring across adult medical–surgical units, integrating wearable sensors with centralized clinical oversight. In a large pre/post analysis of over 320,000 admissions, mortality declined among general-ward patients. ICU transfer rates were unchanged, and LOS had a moderate increase.
Abstract: Houston Methodist, an eight-hospital health system, implemented system-wide continuous physiologic monitoring across all adult medical-surgical units to address surveillance gaps associated with intermittent vital sign assessments. Wearable sensors streamed continuous heart rate, respiratory rate, and skin temperature data into a centralized Virtual Operations Center, enabling real-time oversight and standardized escalation workflows.
We evaluated this redesign using a retrospective cohort of 322,531 adult inpatient admissions, comparing aligned 12-month pre- and post-implementation periods. Outcomes included inpatient mortality, level-of-care escalation, length of stay, and Community Emergency Response Team activations.
Despite a slightly higher comorbidity burden in post-implementation, inpatient mortality declined from 0.32% to 0.23% (RR 0.70;95% CI: 0.61-0.80;p<0.001). The largest reduction occurred among patients who remained on general medical-surgical units, where mortality decreased from 0.29% to 0.18% (RR 0.61;95% CI 0.52-0.71;p<0.001). Mortality among patients requiring transfer to ICU/IMU did not differ significantly. Length of stay increased modestly by 0.19 days (p<0.001), ICU transfer rates were unchanged, and CERT activations rose slightly (1.15 ± 0.46 to 1.18 ± 0.51), consistent with earlier recognition of deterioration.
These findings demonstrate that continuous physiologic monitoring can be implemented at scale across a multihospital system and is associated with improved early recognition of patient deterioration and reduced mortality on general wards.
Classification of Research: Quality Improvement
Method: Observational
Results: Across 322,531 adult admissions, baseline demographics were similar between periods, although the post-implementation cohort had a slightly higher burden of chronic disease. Inpatient mortality declined from 0.32% pre-implementation to 0.23% post-implementation (RR 0.70;95% CI: 0.61-0.80;p<0.001). Mortality reductions were concentrated among patients managed exclusively on medical-surgical units, decreasing from 0.29% to 0.18% (RR 0.61;95% CI 0.52-0.71;p<0.001). Mortality among patients transferred to ICU or IMU did not differ. ICU and IMU transfer rates remained stable. Median hospital length of stay increased modestly by 0.19 days. CERT and Code Blue activations increased slightly, consistent with earlier detection of deterioration.
Conclusions: This system-wide evaluation demonstrates that continuous physiologic monitoring, when integrated with centralized oversight and standardized escalation workflows, is associated with meaningful improvements for inpatient safety. Mortality reductions were observed despite higher baseline comorbidity and without increases in ICU utilization, supporting the value of earlier recognition before critical illness develops. The findings show that the greatest impact occurs on medical-surgical units, where surveillance gaps are most pronounced. Although operational and adoption challenges remain, these results suggest that continuous monitoring represents a scalable, practical approach to strengthening inpatient surveillance, improving reliability of clinical response, and reducing preventable harm in modern hospital care.
Implementing Weekly Safety Huddles to Drive Patient Care and Safety in Telemedicine
Kristen DeDent, BA, LSSBB, Director, Quality and Safety, Patti Griffin, MHA, MBA, RN, FACHE, Vice President, Quality & Safety, Vanisa Patel, MPH, CPHQ, Director, Quality and Accreditation
Access TeleCare
Description: Implement a patient safety event reporting system and weekly patient safety huddles in a fast-paced telemedicine environment spanning multiple service lines and client partners across the United States. Our primary goal is improving patient care and safety, but first we need baseline data and a reporting system to assist in identifying biggest opportunities for continuous improvement.
Abstract: We’ve grown significantly the past few years, including the addition of new companies and processes to Access TeleCare. That growth has brought multiple submission inputs to our patient safety event reporting process. Data collection has shown us that given our high encounter volume, high churn, and anecdotally frustrated clients and patients, we aren't capturing all of the low-risk, near miss events. A project initially focused on capture rate quickly expanded to include:
Simplifying the reporting process and improve security of safety event information
Implementing weekly safety huddles
Improving response time to client
Increasing the number of reported safety events
Reducing days to respond to a client when an event is reported
Improving this process is important because it's a Joint Commission requirement and a company goal is to delight and deliver our clients to reduce churn. We’ve used Lean Six Sigma to optimize this process.
Classification of Research – Other: Patient Safety
Method: Survey/Qualitative
Results: Early results (5-months) show:
Compliance in meeting 30-day response time to client: Improved from 68% to 89%
Other metrics underway including number of reported patient safety events.
Conclusions: Early results are very promising. We implemented a new internal tracking process (May) and weekly patient safety huddles (May). In August, we streamlined multiple reporting input processes to single, secure eForm. We continue to refine the process, collaborate with key stakeholders, keeping patient safety and our client partners at the heart of what we do.
Innovating Connected Care with Ambient AI: Early Outcomes from Houston Methodist’s Emergency Department
Jada A. Malveaux, MA,1,2 Farhan Ishaq, MD, MPH, MBA,1,2 Sarah Pletcher MD, MHCDS,1-3 Ngoc-Anh Anh Nguyen MD1,2,4
1Center for Connected Care Innovation and Implementation-Research, Houston Methodist Hospital; 2Houston Methodist Research Institution, Houston Methodist Hospital; 3Department of Surgery, Houston Methodist Hospital; and 4Department of Medicine, Houston Methodist Hospital
Description: This study evaluates the integration of ambient artificial intelligence (AI) documentation across emergency, inpatient, and outpatient settings. Survey and usage data show strong clinician satisfaction, >24% reduction in documentation time, and improved work-life balance. Findings highlight the potential of ambient AI to streamline workflows, enhance patient focus, and improve operational efficiency system-wide.
Abstract: Background: Within the Emergency Department (ED), providers can face significant challenges including high-acuity cases, rapid assessments, detailed documentation, and demanding patient-to-clinician ratios. Traditional scribes may offer only partial relief, with limitations that persist despite training and workflow adjustments. Clinicians have perceived documentation as a primary barrier to time at the bedside and overall work-life balance, highlighting the need for an innovative solution. The advent of technology and artificial intelligence has brought solutions to remedy this.
Ambient AI (Ambient Care, San Francisco, CA) is a novel documentation platform that passively captures and interprets clinician–patient interactions, generating real-time medical notes directly within the EHR, without the need for dictation or manual input.
Methods: Ambient AI documentation was integrated into the Houston Methodist Emergency Department (ED) in November 2024 via Epic EHR and expanded in a phased roll out. Provider experience was evaluated using a structured electronic survey conducted in May 2025. Survey domains included satisfaction, documentation quality, patient focus, work-life balance, and perceived impact on patient experience. Adoption and usage metrics, including time spent in documentation and number of active users, were obtained from Epic and Ambient Care dashboards. Descriptive statistics were used to summarize survey responses and utilization patterns.
Classification of Research: Quality Improvement
Method: Survey/Qualitative
Results: In a May 2025 survey of ED clinicians (N=24), 100% reported overall satisfaction with Ambience, with an average rating of 4.9/5 and 100% stating they would recommend the platform to colleagues. Nearly all (96%) agreed Ambience improved documentation quality, 92% reported improved work-life balance, and 100% indicated enhanced ability to focus on patients and positive impact on patient experience. 94% reported improved quality of clinical documentation and out of 56% users with scribes, 73% stated they would no longer need one with this Ambient AI. Usage metrics demonstrated a >24% reduction in documentation time, with rapid adoption across ED and hospital settings. By October 2025, 289 providers across inpatient, emergency, and outpatient specialties were actively using the platform.
Conclusions: Our findings demonstrate that ambient AI documentation achieves high clinician acceptance, rapid adoption, and seamless integration within a high-acuity emergency setting. Providers consistently report reduced documentation burden, improved work-life balance, and greater ability to focus on patients. Operational metrics align with these perceptions, showing more than a 24% reduction in documentation time and improved patient throughput. By streamlining workflows, reducing administrative burden, and enhancing documentation accuracy, ambient AI enables providers to devote more attention to patient care and decision-making while supporting both clinician well-being and system-level efficiency. Further evaluation of Ambient AI is warranted in diverse settings.
Intelligent Management Enhances Quality and Efficiency of EMTCT
Dr. Yulin Yang, Professor, Hailiang Cao, Director
Shanghai Center for Women and Children's Health
Description: Multi-source data of Elimination of Mother-To-Child Transmission (EMTCT) poses challenges to evaluation efforts. Based on “intelligent management” concept, we has established data management models such as “positive intelligent trigger-three level verification by laboratory technician/obstetrician/program commissioner-rapid encrypted direct reporting”, “one-stop” collaborative management of screening and intervention for infected pregnant women.
Abstract: The assessment of EMTCT requires accurate and reliable data. Currently, the data source is fragmented, involving multiple departments for separate management, resulting data dispersion, difficulties in sharing, lack of effective data verification and mechanisms in the management of HIV/syphilis/HBsAg(+) pregnant women and their infants. Therefore, taking the EMTCT assessment as an opportunity, we have tapped into the valuable experience of key institutions and strengthen closed-loop management.The hospitals have optimized and upgraded their multi-party data intelligent collaboration platform, integrated mother-to-child blocking services, standardized follow-up diagnosis and treatment, providing reference for the optimization of the “HIV/AIDS, syphilis, and HBsAg(+)”module of Shanghai maternal and child health information system.
Based on the three-level quality mechanism of “real-time reporting-regular verification-bidirectional feedback”, we automatically screens positive reports through preset logical rules in the information system, generates a list of suspected cases, and pushes it to program commissioners to proactively discover positive cases. Additionally, the laboratory department collaborates to verify syphilis, provides an abnormal result inquiry mechanism, and conducts centralized monitoring and proactive screening of syphilis positive results across the hospital;meanwhile, the pharmacy department verifies cases of HBsAg(+) pregnant women through the distribution of hepatitis B immune globulin and obstetric injection records.
Classification of Research: Regulatory & Policy Research
Method: Implementation Science
Results: Led by the Obstetrics Safety Office, we establish a “one-stop”collaborative management mechanism, achieving efficient screening and intervention for pregnant women infected with HIV, syphilis and HBsAg(+). Detection data is pushed in real time, and pregnant women with positive syphilis test results are automatically triggered to receive text message and phone notifications within 24 hours, ensuring timely and early treatment. Pregnant women with HBsAg(+) will get with automatic reminders for follow-up appointments. Consent forms for treatment and records of treatment follow-ups are managed electronically. Finally, the process from positive detection to case reporting is completed within just 48 hours.
Conclusions: EMTCT data management system has formed a tripartite cooperation community. The original data is transformed into a targeted early warning list, ensuring the elimination indicators are scientific and precise. Empowering services through intelligent management to enhance efficiency, breaking down departmental barriers, and facilitating infected pregnant women and their children to receive effective and timely intervention, thereby maximizing the prevention of mother-to-child transmission.Integrating medical and prevention efforts, moving from mere data accumulation to intelligent early warning, introducing rule-based data interaction and business process re-engineering.
Launching Pediatric eConsults at MUSC Health: Early Experience
Description: The use of eConsults in pediatrics lags adult medicine, despite documented benefits of increased access, decreased time to FTF visits, and high provider satisfaction. Effective implementation requires stakeholder engagement, technologic capacity and ongoing monitoring. This poster describes the implementation of a pediatric electronic consult (eConsult) service at the Medical University of South Carolina, evaluating implementation strategies, utilization, provider satisfaction, and feasibility to improve access to pediatric subspecialty expertise while reducing unnecessary face-to-face (FTF) referrals.
Abstract: Pediatric subspecialty access is often limited by referral delays and reliance on informal consultations. eConsults provide asynchronous, documented communication between primary care providers and specialists, yet pediatric implementation data remain limited. MUSC Health implemented a pediatric eConsult service within the EMR to support provider-to-provider consultation. Implementation strategies included standardized ordering workflows, defined response expectations, and provider education. Early implementation outcomes were assessed from February through December 2025, including utilization, specialty participation, ordering patterns, and turnaround time. Thirty-nine pediatric eConsults were completed across multiple subspecialties during the evaluation period. eConsults addressed focused clinical questions appropriate for asynchronous review. Turnaround times consistently met the program goal of specialist response within two business days, demonstrating timely engagement and workflow feasibility. Specialists provided documented recommendations or guidance regarding the need for in-person referral. Early implementation of a pediatric eConsult service at MUSC Health was feasible and achieved timely specialist response within established targets. Initial utilization across subspecialties suggests successful integration into existing clinical workflows. Ongoing evaluation will focus on utilization trends, specialty expansion, and additional implementation outcomes to inform future service refinement.
Classification of Research: Quality Improvement
Method: Implementation Science
Results: From February to December 2025, a multidisciplinary engagement team recruited 6 subspecialities to participate in eConsults, simultaneously promoting the service with primary care providers within the institution. Over this time, providers from 14 unique sites ordered 39 eConsults. Median turnaround time met the program goal of response within 2 business days. Both ordering and reading providers show high satisfaction on initial surveys. Specialists provided either management recommendations or guidance for in-person referral, reducing unnecessary visits, and formalizing documentation of clinical advice. The service demonstrated operational feasibility and integration into existing EMR workflows with minimal disruption to provider practice.
Conclusions: Implementation of a pediatric eConsult service at MUSC was feasible and achieved timely subspecialty responses within the targeted two-business-day window. Early utilization across multiple subspecialties demonstrates initial adoption and successful integration into Epic workflows. These findings support continued monitoring of utilization trends, specialty participation, and implementation outcomes to guide future service refinement and evaluation.
Multidimensional Determinants of AI-Based Digital Health care Technology Adoption in South Korea: An Integrated TAM and ABM Measurement Framework
Byunghee Choi, PhD
Korea Institute of Oriental Medicine
Description: Using national official statistics (N=7,000), this study tests an integrated Technology Acceptance Model (TAM)–Andersen’s Behavioral Model (ABM) framework for AI-based health care adoption in South Korea. Partial least squares structural equation modeling shows digital capability, health context, and prior use significantly shape intention alongside core TAM pathways, informing measurement and user-segmented telehealth strategies.
Abstract: Adoption of AI-based health care remains uneven, indicating that technology perceptions and health care-use context jointly shape acceptance. This study tests an integrated TAM-ABM framework to explain intention to use AI-based health care in South Korea. We analyzed the 2024 Digital Divide Survey conducted by the National Information Society Agency (N=7,000). Measures included perceived ease of use (3 items), perceived usefulness (2), attitude toward digital technology (1), health satisfaction as a need factor (1), smart device capability as an enabling factor (7), and prior AI-based health care use frequency during the past month as a behavioral factor. Partial least squares structural equation modeling was estimated using SmartPLS 4 with 5,000-bootstrap inference, controlling for sex, age, disability status, living arrangement, and household income. Model fit was adequate (SRMR=0.028), and intention was explained at R2=0.192. Significant paths included enabling factor→perceived ease of use (β=0.450, p<.001); perceived ease of use→perceived usefulness (β=0.547, p<.001) and attitude (β=0.260, p<.001); perceived usefulness→attitude (β=0.505, p<.001) and intention (β=0.113, p<.001); attitude→intention (β=0.101, p<.001); behavioral factor→intention (β=0.274, p<.001); and need factor→perceived usefulness and intention (β=0.130, p<.001). These results indicate heterogeneous adoption pathways combining digital capability, prior experience, and health context with core TAM mechanisms, supporting a measurement framework to inform telehealth policy, user segmentation, and service design.
Classification of Research: Measurement Frameworks & Tools
Method: Survey/Qualitative
Results: Using national survey data (N=7,000), the PLS-SEM model demonstrated adequate fit (SRMR=0.028). Intention to adopt AI-based health care was explained at R2=0.192. Smart device capability (enabling factor) showed a strong positive association with perceived ease of use (β=0.450, p<.001). Perceived ease of use significantly influenced perceived usefulness (β=0.547, p<.001) and attitude (β=0.260, p<.001). Perceived usefulness was positively associated with attitude (β=0.505, p<.001) and intention (β=0.113, p<.001). Attitude (β=0.101, p<.001), prior AI health care use (β=0.274, p<.001), and health satisfaction as a need factor (β=0.130, p<.001) were also directly associated with intention.
Conclusions: This study demonstrates that adoption of AI-based health care is shaped by both technology perceptions and health care-use context. Beyond core TAM mechanisms, digital capability and prior use experience substantially strengthen intention, highlighting the importance of readiness and experiential exposure. Notably, higher health satisfaction was also associated with intention, suggesting that AI-based health care is valued not only for illness-driven needs but also for preventive and health-maintenance purposes. These findings imply that telehealth policies should move beyond deficit-oriented targeting and adopt user-segmented strategies that combine usability-focused design, digital capability support, and low-barrier trial opportunities to enhance broader and more sustainable adoption.
Operationalizing Predictive Models into Clinical Practice
Julie Simonson, PharmD, Saira Haque, PhD, MHSA, FAMIA, ACHIP, Sr. Director, Advanced Medical Solutions, Jay Pauly, PharmD, BCPS, CPh, Sr. Director Payer Account Medical Lead
Pfizer Inc
Description: Predictive models have a number of uses such as informing selection of in-person versus virtual care. The evidence is characterized by validation, pilot, and feasibility studies and understanding considerations for deployment is imperative so that their benefits can be realized. We sought to explicate practical considerations for model operationalization.
Abstract: We focused on practical considerations beyond the research setting to inform utilization in clinical practice and insights on service selection. We conducted a targeted literature scan to understand the state of evidence on operationalizing models in primary care settings. This was used to identify areas of interest and inform development of an interview guide for individual interviews. Participants were providers and staff at primary care practices across the United States and were recruited based on their knowledge of and experience with predictive models. Themes from the literature scan and individual interviews were used to refine the topic list for a subsequent group interview with the same participants. All interviews were documented by a notetaker, analyzed by an interdisciplinary team, and themes were extracted and refined during weekly meetings with the study team. Overall, while perspectives on predictive models were optimistic, their application into practice and utility for service selection was strongly influenced by clinical, financial and operational considerations. Participants were familiar with predictive models yet had some concerns and identified practical considerations regarding incorporation into practice. Decision-making regarding model utility, potential impact on the population, cost/benefit analyses, implementation considerations and approaches, and operational implications were themes highlighted by participants.
Classification of Research – Other: Information Technology and Measurement Frameworks/Tools
Method: Survey/Qualitative
Results: Decision-making: Understanding the evidence base. Potential patient benefits (eg, service mode selection, identification, management) and translating model output into clear, actionable next steps. Were the strongest factors for model selection and implementation.
Implementation: Participants favored piloting before large scale implementation to confirm needed resources, troubleshoot issues, cultivate provider buy-in, and support utilization. Ongoing education on the underlying need and evidence for the model, associated technology, and workflow changes were highlighted.
Organizational: Organizational commitment, capacity to identify and address operational considerations, establishing organizational mechanisms for technical and virtual care platform support and recommended next steps per model output were emphasized.
Conclusions: Our work points to the need to focus on a practical approach to operationalizing predictive analytic models into practice. Without incorporating best practices and principles into implementation and deployment and allocate resources for implementing, maintaining and using the model (e.g., technical costs, workforce, training), models will not be consistently used and their potential benefits (e.g., time-savings, clinical benefit, service mode selection) will be unrealized by patients, providers and health care organizations. The findings indicate that aligning along decision-making, implementation, and organizational factors may support adoption of predictive models into clinical practice.
Operations of a Virtual Nursing Unit: Understanding Activity Patterns, Time Spent, and Identifying Opportunities for Strategic Prioritization Following Rapid Expansion
Caitlin Koob, PhD, MS, Research Associate-Faculty, Kit N. Simpson, DrPH, Distinguished University Professor, Jillian Harvey, PhD, MPH, Professor, Dunc Williams, PhD, MHA, Associate Professor, Katie Kirchoff, MSHI, Database Administrator, Peter Gardella, MBA, Director of Telehealth Operations and Nursing, Emily Warr, MSN, Administrator of the Center for Telehealth, Dee Ford, MD, Professor of Medicine & Division Director for MUSC’s Division of Pulmonary, Critical Care, Allergy, and Sleep Medicine, PI of MUSC's Center of Excellence for Telehealth
Medical University of South Carolina (MUSC)
Description: A Virtual Nursing program (VRN) was rapidly implemented across 36 units in 2024. VRN supported 23,516 visits, primarily including admissions, quality-of-care, and discharge tasks. This study aims to understand patterns of VRN activities, including the distribution of time spent (in minutes), and consider opportunities for prioritization within a VRN queue.
Abstract: Purpose: To understand patterns of VRN activities, including time spent (in minutes), and consider opportunities for prioritization within a VRN queue.
Methods: Data were triangulated from multiple sources, including program tracking data and electronic health records. Using descriptive statistics, we identified the number of services combined in visits and estimated the mean number of minutes spent performing each type of task.
Results: From January 1, 2025-June 30, 2025, VRN provided support for 23,516 visits, including 58,793 specific tasks. VRNs primarily supported admissions (65%), quality-of-care (18%), and discharges (17%), and often provided multiple types of care within a visit. Overall, the amount of VRN support provided, assuming standard 2,080 work hours per FT/year and 80% productivity, would require 10.2 full-time nurses.
Across service types, visits lasted a mean of 36.8 minutes (SD=27.0) with longer visits time for admissions alone (M=34.6, SD=26.9), followed by discharges (M=25.5, SD=12.3). Quality-of-care visits were the shortest (M=9.3, SD=6.5). Lastly, 61.8% of discharges occurred in <3 hours with VRN support.
Conclusions: VRNs effectively offset time-consuming tasks for bedside nurses to provide direct patient care. This work provides a baseline understanding of VRN support allocation within one health system;however, further research is needed to understand its impact on multi-level outcomes.
Classification of Research: Quality Improvement
Method: Implementation Science
Results: Overall, VRN supported 60.2% of hospital admissions (n=23,516), providing most support for admissions (65%), followed by quality-of-care (18%) and discharge tasks (17%). Mean visit length was 36.8 minutes (SD=27.0) with longer visits time for admissions alone (M=34.6, SD=26.9), or 44.7 minutes when combined with care planning or education. Quality-of-care visits lasted a mean of 9.3 minutes (SD=6.5), or 23.8 minutes with education. Discharge visits lasted a mean of 25.5 minutes (SD=12.3), or 23.0 minutes (SD=11.4) with education. Lastly, 61.8% of discharges occurred in <3 hours from time of order with VRN support, indicating an opportunity to impact throughput metrics.
Conclusions: VRN programs were largely developed for operational and workforce support and, while these programs continue to evolve nationally, there is a growing need for metrics to manage VRN teams and increase health care efficiency in inpatient care. Our results are promising of the system-level impact of VRN support on a myriad of quality, workforce, and financial metrics. Ongoing efforts may consider prioritizing the VRN queue to measure impact on specific outcomes (i.e., timeliness of discharge), examine the impact of VRN support on nursing workload, and the cost-effectiveness of VRN in improving nursing-sensitive quality measures and throughput.
Telehealth Use in Behavioral Health Among Commercially Insured Patients (2022)
Dunc Williams, PhD, Caitlin Koob, PhD, Research Associate Faculty, Kit Simpson, DrPH, Professor
Medical University of South Carolina
Description: Using 2022 MarketScan data, we evaluated telehealth use among commercially insured individuals for which the primary or secondary diagnosis code was anxiety, depression or whether both were present. We quantified and identified utilization patterns across visit types, ages, and sex.
Abstract: Telehealth use in the U.S. has proliferated in recent years both in terms of volumes and innovative modalities. One prominent area of telehealth delivery is behavioral health. Using 2022 MarketScan data, we evaluated telehealth delivery among commercially insured individuals for whom the primary or secondary diagnosis code was anxiety, depression or whether both were present. We quantified utilization patters across visit types, ages, and sex for individuals age <64.
For 2022, we identified 15.9mn telehealth commercially insured visits in the MarketScan dataset. Of those, 6.6mn visits had a primary ICD-10 diagnosis code indicating a behavioral health condition based on the AHRQ Clinical Classifications software (CCSR) for ICD-10-CM diagnoses, v2025. We limited our sample to the 5.8mn (87.8%) who had a primary or secondary diagnosis code of anxiety, depression or whether both were present.
Classification of Research: Access to Care
Method: Secondary Data Analysis
Results: A total of 5.8mn visits were identified for 236,395 unique patients, 29.1% were male and 70.9% female. Primary diagnoses were 30.6% anxiety, 50.6% depression, and 18.8% had a mix of depression and anxiety as primary and secondary diagnoses across visits. The mean [median] number of telehealth visits per person based on diagnosis were anxiety 9.0 [4], depression 8.0 [4] and both diagnoses present 39.7 [26], indicating a much heavier use of telehealth visits for patients with both diagnoses. Having both diagnoses was most common in the youngest age group and decreased monotonically with increasing age.
Conclusions: There is a large volume of telehealth visits for anxiety and depression. Among commercially insured individuals in the 2022 sample, the largest group of patients was female between age 18 and 44 with a primary diagnosis of anxiety (14%). Patients with anxiety or depression had a mean 8.5 visits [median 4]. However, those with both diagnoses had a mean of 40 visits [median 26], with the highest visit number observed for the youngest age group with both anxiety and depression.
Through Two Lenses: Exploring Dyadic Perceptions of Telebehavioral Health
Rachana Talekar, MPH, Program Manager, Nathanael Saavedra, BS Public Health, Student Intern, Amyia Harris, M.Ed., Graduate Assistant - Research, Israel Arevalo, PhD, Assistant Research Scientist, Kala Reindel, PhD, Research Scientist, Cuevas Arney, PhD, Research Scientist, Kristen Garcia, MPH, Assistant Research Director
Texas A&M Telehealth Institute
Description: This study reviews existing literature on patient-caregiver experience with technology and telebehavioral services, synthesizing studies examining how telehealth is perceived, used, and accepted.
The findings highlight differing telehealth experiences, acceptability, and needs across clinical contexts that underscore the need to improve telehealth implementation strategies for adolescents and their caregiver dyads.
Abstract: Despite the rise in telebehavioral services, limited studies assess the experiences of adolescents and their caregiver (dyads) with telebehavioral care. Dyads experience unique care continuity, adherence to telehealth, and improvements in parental engagement or coping skills among adolescents. Although adolescents are often more digitally fluent than adults; it is unclear if technological comfort translates into meaningful therapeutic engagement among dyads.
This scoping review maps peer-reviewed literature by examining adolescent-caregiver dyadic research on telehealth experiences, technology use, acceptance and care processes in behavioral care services.
Following scoping review guidelines, a comprehensive search conducted across PubMed, Web of Science and MEDLINE identified peer-reviewed studies published between 2000-2025. Search strategies used terms related to adolescents, caregivers, dyads and telehealth, and behavioral health. Eligible studies included patient-caregiver dyads focusing on child or adolescent engaging in synchronous telehealth services. Of 301 screened articles, six met the criteria assessing adolescent or patient-caregiver dyad experience, usability of telehealth showcasing the general acceptance yet some divergence in dyad reported technology use, access and therapeutic alignment. The outcomes included usability, satisfaction, quality of care, patient experience, and self-reported perceptions of telehealth and mental health measures. Qualitative, quantitative designs were included, and audio-only or family therapy interventions were excluded.
Classification of Research: Patient Experience
Method: Scoping Review
Results: A total of 301 articles were screened, and six studies were selected meeting the inclusion criteria. These studies most frequently assessed satisfaction, perceived usability, engagement, attitude towards technology, and telehealth as outcomes. The studies reported generally higher levels of acceptance of telehealth among dyads and identified barriers reported by caregivers and clinicians related to access, therapeutic alliance, and usability.
Conclusions: This study exposes a gap in the literature examining adolescent-caregiver dyadic experience with telehealth technology in addition to their care experiences. The existing literature on patient-caregiver dyads showcases limited research conducted focusing primarily on cross-sectional study designs and caregiver perceptions of the adolescents’ experiences. This highlights the need for more dyad-centric research that links technology experiences to clinical outcomes and continued care engagement. Without a structured understanding of these experiences on a holistic level, telebehavioral health risk factors, such as disengagement, low adherence, or poor outcomes may occur, even if access and availability are achieved.
Transforming the Arrival Experience: Leveraging Virtual Providers in Triage to Improve Emergency Department Patient Experience
Marc Bartman, MD, FACEP, Jeanhyong Park, MD, FACEP, Director of Innovation, Division of EM Telehealth, Assistant Professor
Medical University of South Carolina
Description: To describe how the Virtual Provider in Triage (VPIT) model leverages telehealth to transform the ED arrival process, improve early engagement, and enhance patient experience.
Abstract: The arrival experience in the emergency department is a critical yet often overlooked component of the overall patient journey. Upon arrival, patients frequently encounter prolonged waits and limited communication, leaving them uncertain about what to expect. This experience can lead to anxiety, frustration, and disengagement. Unlike scheduled visits with a known provider, ED visits lack an established provider-patient relationship, further diminishing trust and satisfaction. These early negative impressions are difficult to overcome, even if subsequent care is high quality.
Methods: VPIT integrates a remote emergency medicine provider into the triage process to engage patients shortly after arrival. This provider initiates a tailored workup, offers anticipatory guidance, and begins building rapport with the patient, converting passive wait time into active care. To measure impact, we implemented a real-time five-question survey. Using a Net Promoter Score (NPS) format, patients were asked:
How welcome and comfortable they felt after speaking with the virtual provider
How well the provider listened and addressed their concerns
How clearly the next steps in care were explained and how included they felt in decisions
How the virtual interaction helped put them at ease while waiting
How likely they were to recommend VPIT service
Classification of Research: Patient Experience
Method: Survey/Qualitative
Results: Real-time survey responses from 1,007 patients indicate high levels of patient satisfaction along with consistently strong NPS scores across all five dimensions. By transforming the lobby experience from passive waiting to active engagement, VPIT has significantly enhanced the quality and consistency of the ED arrival process.
Net Promoter Scores (NPS) for each question were:
78
82
83
83
80
Average NPS across all categories: 81.2
n=1,007
Our results exceed both industry averages and top-tier benchmarks, indicating a significantly enhanced arrival experience.
Conclusions: VPIT transforms the ED lobby from a passive waiting space to a proactive, high-impact point of care. By integrating virtual providers into the earliest moments of the ED visit, this model fosters trust, promotes patient understanding, and improves both emotional and clinical readiness. As EDs continue to seek scalable, patient-centered innovations, VPIT offers a replicable and effective strategy to transform the front end of emergency care delivery.
Transforming Women’s Health Through Tech-Enabled Care: Evidence from InovCares’ Texas Collaboration
Mohamed Kamara1, J’Nae Taylor, MS,1 Maternal Care Navigator, Chelsea Cooper, BS,1 Research Assistant, Lovemore Chirombo, BS,1 Chief Technology Officer, Pelumi Adedayo, MD,1 Chief Medical Officer, Henry Imperial, MD, Chief Medical Officer, New Horizon Health Center, Taibat Eribo MD, Medical Director, HHM Health, Kyrah K. Brown, PhD, Associate Professor, University of Texas at Arlington
1Inovcares
Description: This study evaluates InovCares, a tech-enabled maternal health platform implemented in Texas Federally Qualified Health Centers (FQHC). It improved clinical outcomes and patient satisfaction through digital care coordination, social support services, and bilingual engagement. High app usage and transportation assistance highlight its scalable, equity-driven model for advancing maternal health in underserved communities.
Abstract: Texas continues to face significant maternal health disparities. This paper evaluates the implementation of InovCares, a mobile-first maternal health platform deployed across two FQHCs in partnership with a managed care organization. Over a 12-month period, InovCares combined remote biometric monitoring, digital care coordination, and services addressing social determinants of health such as transportation and housing. Among 605 enrolled pregnant individuals, two-thirds continued pediatric care post-delivery. Clinical outcomes showed reduced emergency department visits, lower preterm birth rates (7%), and fewer cesarean deliveries (22%) compared to state benchmarks. Patient satisfaction data revealed consistently high ratings for staff friendliness, app engagement, and transportation support. Over 90% of respondents used the app weekly or more, and nearly all expressed appreciation for bilingual interfaces and culturally responsive care. Key lessons included the importance of integrating care navigation into existing workflows, leveraging bilingual digital tools, and forming strategic payer partnerships to support reimbursement for non-clinical services. InovCares demonstrates a scalable, equity-driven model that aligns with value-based care priorities and offers a replicable framework for improving maternal health outcomes in underserved communities.
Classification of Research: Clinical Effectiveness
Method: Intervention Study
Results: The InovCares pilot in Texas showed encouraging signs of improved maternal health outcomes and patient experience. Among 605 participants, emergency visits declined, and rates of preterm birth (7%) and cesarean delivery (22%) were lower than state averages. Two-thirds continued pediatric care post-delivery. Survey responses reflected generally high satisfaction, with frequent app use and appreciation for bilingual support and transportation assistance. Lessons learned emphasized the value of integrating care navigation, fostering payer collaboration, and using culturally responsive digital tools. While further evaluation is needed, these results suggest that InovCares offers a promising framework for advancing maternal health equity in underserved settings.
Conclusions: The implementation of InovCares across two Texas FQHCs yielded promising results in addressing maternal health disparities. The platform’s integration of digital tools and social support services contributed to improved clinical outcomes, including reduced emergency visits and lower rates of preterm birth and cesarean delivery. Patient feedback highlighted strong satisfaction with bilingual support, transportation assistance, and app usability. Continued pediatric care post-delivery and frequent app engagement suggest sustained impact. Lessons learned underscore the importance of workflow integration, culturally responsive design, and payer collaboration. Overall, InovCares offers a replicable, equity-focused model for enhancing maternal health in underserved settings through tech-enabled care.
Trends in Time to First Contact After CERT Activation Following Implementation of a Virtual CERT Program: A Retrospective Review
Aline Kalaydjian, MD1,2, Carl Daniel, MSN, FNP, Nurse Practitioner Service, Department of Medicine, Farhan Ishaq, MD, MPH, MBA,1,2 Clinical Research Fellow, Jada A. Malveaux, MA,1,2 Clinical Research Specialist, Sarah Homer; DNP, ACNP, Director, Nursing, Department of Medicine, Laura Griffin, MSN, ACNP, Lead Nurse Practitioner, Department of Medicine, Sarah, Pletcher, MD, MHCDS,1,2,3 Chair, Department of Surgery, Ngoc-Anh Anh Nguyen, MD,1,2,3, Vice Chair
Houston Methodist Hospital, 1Center for Connected Care Innovation and Implementation-Research, 2Houston Methodist Research Institution, 3Center for Connected Care Implementation & Research, Houston Methodist Research Institute
Description: This retrospective review assessed trends in time to first contact after a CERT activation following implementation of a virtual CERT program. Mean time to first contact decreased after implementation, suggesting that virtual CERT is an appropriate intervention to reduce operational delays.
Abstract: Background: Early signs of clinical deterioration, if not promptly recognized and addressed, may lead to cardiopulmonary compromise and death. At Houston Methodist, an urban hospital in Houston, TX, a Clinical Emergency Response Team (CERT) provides rapid assessment and intervention for evolving medical emergencies, preventing further decline and supporting escalation of care. Despite established response effectiveness, operational delays in CERT arrival from time of activation to first contact were identified. To address this issue, Houston Methodist extended CERT framework by implementing a virtual CERT (vCERT) program beginning August 4, 2025. vCERT leverages existing in-room audiovisual technology to enable immediate remote response prior to in-person CERT response.
Methods: We conducted a retrospective review of 3,365 CERT activations initiated across 40 acute care non-ICU/IMU units at Houston Methodist Hospital between October 1, 2024, and November 30, 2025. vCERT program leverages in-room audiovisual technology to enable immediate remote response. Implementation was on a rolling basis in August and September 2025. In August 2025, vCERT was introduced during day shifts and on weekdays, and in September 2025, program was expanded to include night shifts and weekends, ensuring full-scale hospital-wide implementation. Time-to-first contact was defined as time from CERT activation to first contact with a CERT team member, either in-person or virtual.
Classification of Research: Quality Improvement
Method: Observational
Results: From October 1, 2024, to November 30, 2025, a total of 3,365 CERT activations across acute care units occurred. During the pre-implementation period ranging from October 1, 2024, to August 3, 2025, 2,439 CERT activations were recorded. Following the implementation of the vCERT program on August 4, 2025, 926 CERT activations occurred through November 30, 2025. Mean time to first contact decreased from 3.60 ± 1.10 minutes in the pre-implementation period, to 0.34 ± 0.06 minutes post-implementation, reflecting a 90.56% relative reduction.
Conclusions: Integration of a virtual CERT program, which utilizes existing audiovisual technology to enable immediate remote response, to an existing CERT framework was associated with a 90.56% reduction in time to first contact following a CERT activation. These findings suggest that the virtual CERT program is an appropriate intervention to reduce operational delays. Future work will explore whether the implementation of a vCERT program is associated with improvements in patient outcomes, including reduced ICU transfers, total length of stay, and in-hospital mortality.
Virtual Competencies and the Pedagogical Advances in Health Care
Marie Martinez, MS, MEd, CCC-SLP, BCBA, LBA, MS, MEd, CCC-SLP, BCBA, LBA, Tayler Elizondo, MS, Clinical Instructor in the Human Performance and Sports Studies, Ashley Ben-Jacob, MEd, DHH Education Specialist, PhD Student: Department of Rehabilitation & Health Sciences, Lucie Pell nd, PhD, BScPT, MSc, MHPE, Professor in Physical Therapy
Idaho State University
Description: This review examined telehealth competencies among medical and allied health professionals, with a focus on how these competencies, such as knowledge, skills, attitudes, and judgment, are defined, taught, and assessed within educational and clinical practice settings. We anticipate providing a framework for integrating virtual care in health care training curricula.
Abstract: Despite the evidence of telehealth capabilities, the readiness of graduates from programs of health professions to deliver virtual care remains limited, revealing a persistent gap between technological advancement and professional preparation [Shachak et al., 2024]. Addressing this gap ensures graduates possess the technical proficiency and behavioral readiness required to deliver equitable care in digitally enabled health systems. To meet this educational need, clearly defined competencies are essential to guide curriculum development and assessment for telehealth practice. Our scoping review was conducted according to the PRISMA Extension Guidelines for Scoping Reviews [Tricco et al., 2018]. The search was constructed iteratively in OneSearch, with each expansion evaluated using the precision-optimization method outlined by Alharbi and Stevenson [2020]. To examine how competencies were identified and evaluated, data extraction was guided by the Population–Concept–Context framework [Pollock et al., 2023] and focused on study aims, design, and methods used to define and assess competencies. Competencies were organized into three domains: technical, clinical, and behavioral [Citron et al., 2023; Marsillo et al., 2024; Shachak et al., 2024]. These domains will be synthesized to demonstrate consistencies among competencies to address the educational gap in health professions.
Classification of Research: Quality Improvement
Method- Other: Scoping Review
Results: This scoping review aims to identify telehealth-related competencies in health professions curricula and examine instructional strategies and assessment methods used to teach and evaluate these competencies. These aims are within the specific context of virtual health care. It is anticipated that health care providers in training do not receive specific training in delivering their specialty in a virtual setting, and these competencies are typically developed through on-the-job training. It is expected that themes will emerge, such as virtual competencies being taught live in synchronous learning environments, without additional competencies specifically related to delivering virtual care.
Conclusions: The scoping review is anticipated to provide insights into the importance of integrating virtual care in health care training curricula. This could include recommendations for specialized training received by health care providers in their training coursework. While the actual findings will be determined in April 2026, the results are projected to contribute to the limited research on evidence of virtual health care training practices.
Virtual Urgent Care Services in a Rural Health System: Capturing Patient Satisfaction and Expanding Access
Rachel Dumont, MS, OTR/L, Research Program Manager, Center for Telehealth, Tori Miller, BA, Manager of Telehealth Services, Center for Telehealth, Lauren Winter, MBA, Telehealth Analyst Team Lead, Center for Telehealth, Jennifer Fritz, RN, MSN, AVP of Hospital Services, Patient Services and the Center for Telehealth, Ross Ellison, MD, MBA, Director of Inpatient Medical Management, Inpatient Medical Management
Geisinger
Description: This virtual urgent care project was launched to aid in alleviating the demands on emergency rooms, urgent care, and primary care clinics, and to advance support for patients in a rural setting. A survey of virtual urgent care patients indicates high satisfaction and an alternative option to in-person services.
Abstract: Introduction: In rural communities, emergency departments are often a primary source of health care access. Implementing virtual urgent care services may contribute to reducing these in-person visits while also supporting rural residents’ health care needs. Following the launch of virtual urgent care services, we sought to learn about patients’ experiences.
Methods: Conducted a voluntary survey of patients’ experiences following their completed on-demand virtual urgent care visits between October 2024 to July 2025. Demographic data, virtual wait room duration, call duration, and survey data were collected and analyzed.
Results: 6,301 patients/families participated in the voluntary survey. Respondents indicated high satisfaction with their virtual visit and a high likelihood of recommending virtual urgent care services. Patients reported they would have otherwise sought treatment at in-person urgent care, primary care clinic, treat at home, do nothing/not seek other care, or go to an emergency room.
Conclusion: These findings suggest virtual urgent care options in rural communities can reduce clinic and emergency department visits, potentially lowering transportation costs and saving time for patients. High levels of patient satisfaction suggest that the virtual care model effectively supports patients’ health care needs.
Classification of Research: Patient Experience
Method: Survey/Qualitative
Results: 8,189 virtual urgent care visits were completed (70% female patients, median age 35 [range 2 years to 93 years]). 6,301 patients/families completed the voluntary survey (77% survey response rate). 96% of respondents indicated high satisfaction with their virtual visit. For alternative options to care, respondents reported they would have sought treatment at in-person urgent care (4,069 respondents, 70%), primary care doctor (781 respondents, 13%), treat at home (537 respondents, 9%), do nothing/not seek other care (252 respondents, 4%), or go to the emergency room (188 respondents, 3%). 97% of respondents reported a high likelihood of recommending this virtual service.
Conclusions: These findings suggest most patients would have sought treatment in-person options, but that the virtual urgent care option aided in saving 4,069 in-person urgent care visits, 781 primary care visits, and 188 emergency room visits. Virtual options may also reduce travel time, wait times, and scheduling challenges associated with in-person visits. The data also highlights high patient satisfaction and likelihood of recommending these virtual urgent care services to others, indicating that this delivery format is meeting patients’ health care needs. Providing virtual urgent care service options may be an effective option for rural patients seeking timely non-emergency care.
Telehealth Utilization Patterns Among Patients with Chronic Medical Conditions: A Population-Level Analysis of the 2024 NHIS Data
Georgia State University, Robinson College of Business
Description: This study examines demographic, socioeconomic, and geographic predictors of telehealth utilization among U.S. adults managing hypertension and diabetes, using the 2024 NHIS dataset. It investigates whether telehealth effectively bridges physical access barriers or remains accessible only to the socioeconomically advantaged.
Abstract: Telehealth has emerged as a critical modality for healthcare delivery, yet disparities in access persist among vulnerable populations. The rapid integration of digital health technologies has transformed chronic disease management, particularly for patients with hypertension and diabetes who require consistent provider access to prevent complications. Despite its potential, the ‘Digital Divide’ remains a significant concern. While technology offers a workaround for geographic distance, it may inadvertently create new barriers for those with limited technological literacy or lack of high-speed internet. Social determinants of health (SDOH), including insurance status and level of urbanization, significantly influence a patient's ability to engage with virtual care platforms. As the healthcare system continues to evolve post-2020s, understanding current patterns of telehealth utilization is critical for long-term policy planning. Current literature indicates that while urban centers often have better infrastructure, rural populations may have a higher clinical need for remote services due to a scarcity of local specialists. It remains unclear which specific factors—geographic, financial, or clinical—exert the strongest independent influence on telehealth adoption among those with chronic illnesses. This study addresses that gap using a nationally representative sample of over 12,000 adults.
Classification of Research: Access to Care
Method: Secondary Data Analysis
Results: The overall prevalence of telehealth use was 27.2%. Binary logistic regression identified insurance status (OR: 0.60; p < .001) and a diagnosis of Diabetes (OR: 1.52; p < .001) as significant predictors. Patients with transportation barriers were 1.71 times more likely to use telehealth (p < .001), identifying virtual care as an effective safety net for those with physical mobility constraints. Younger adults aged 18–34 showed significantly higher odds of adoption (OR: 1.54). Notably, home internet access was not a statistically significant independent predictor when controlling for socioeconomic factors (p = 0.241).
Conclusion: Telehealth serves as a vital bridge for patients facing physical access barriers, demonstrating particular utility for those with transportation challenges. However, the digital divide is increasingly defined by insurance coverage and socioeconomic status rather than hardware availability alone. Financial barriers remain the most significant gatekeepers to equitable access to virtual care. To achieve health equity, policy interventions should prioritize expanding insurance coverage for virtual services and targeting rural populations, where utilization rates remain significantly lower than in metropolitan areas. The 'democratization' of healthcare through technology is currently stalled by traditional reimbursement and policy structures.
Development and Evaluation of a Telehealth Preparedness Assessment
Julie Faieta, PhD, MOT, OTR/L; Tim Bober, MD; Steven Handler, MD, PhD, CMD
University of Pittsburgh, Veterans Administration Pittsburgh Healthcare System, Pittsburgh, PA
Description: To evaluate the sensitivity, specificity, and predictive values of the VVCMatch, a telehealth preparedness assessment developed for veterans, in a cancer population, and to co-design a targeted telehealth education program addressing barriers identified through the assessment.
Abstract: Following the COVID-19 pandemic, approximately 12% of outpatient cancer care visits are now delivered via telehealth. The Veterans Administration (VA) has invested substantially in the Veterans Video Connect (VVC) platform to expand tele-oncology access, yet veterans with cancer face significant physical and cognitive barriers to VVC use, including speech and hearing impairments and chemotherapy- or hormonal therapy-related changes in memory, processing speed, and mental flexibility. No validated tool currently exists to systematically identify a veteran's preparedness for VVC use relative to their individual functional profile, contributing to poor adoption and costly abandonment of telehealth technology.
The VVCMatch was developed through an iterative, human-centered design process. A multidisciplinary team conducted a structured task analysis of VVC engagement across tablet, smartphone, and laptop platforms, decomposing each interaction into discrete physical and cognitive task requirements. These requirements informed an item bank evaluating physical, cognitive, and digital literacy competencies, supplemented by the Montreal Cognitive Assessment.
In the current study, a target N = 30 veteran participants with prostate, colorectal, or lung cancer will complete the VVCMatch User Assessment, followed by a live VVC call evaluated by trained research personnel using a standardized Task Analysis Checklist. VVCMatch scores will be analyzed for sensitivity, specificity, and positive and negative predictive values against this reference standard. At present 11 Veterans have completed the study. Concurrently, our research team is working to develop a multimodal telehealth education program including video modules, accessible curricula, and infographics targeting identified barriers.
Method: Observational
Classification of Research: Patient Experience
Results: This study will establish preliminary psychometric evidence for the VVCMatch as a clinically viable screening tool for tele-oncology readiness, and produce an education program tailored to the cognitive and physical profiles of veterans undergoing cancer treatment.
Conclusion: The VVCMatch addresses a critical gap in tele-oncology implementation by offering a systematic, patient-centered method for evaluating VVC readiness. This work has direct implications for improving care access, reducing telehealth abandonment, and advancing equitable digital health delivery for veterans with cancer.
