Emergency Department Visits Research Articles

Home Health Care Use and Outcomes After Coronary Artery Bypass Grafting Among Medicare Beneficiaries.

Home health care (HHC) has been increasingly used to improve care transitions and avoid poor outcomes, but there is limited data on its use and efficacy following coronary artery bypass grafting. The purpose of this study was to describe HHC use and its association with outcomes among Medicare beneficiaries undergoing coronary artery bypass grafting. Retrospective analysis of 100% of Medicare fee-for-service files identified 77 331 beneficiaries undergoing coronary artery bypass grafting and discharged to home between July 2016 and December 2018. The primary exposure of HHC use was defined as the presence of paid HHC claims within 30 days of discharge. Hierarchical logistic regression identified predictors of HHC use and the percentage of variation in HHC use attributed to the hospital. Propensity-matched logistic regression compared mortality, readmissions, emergency department visits, and cardiac rehabilitation enrollment at 30 and 90 days after discharge between HHC users and nonusers. A total of 26 751 (34.6%) of beneficiaries used HHC within 30 days of discharge, which was more common among beneficiaries who were older (72.9 versus 72.5 years), male (79.4% versus 77.4%), White (90.2% versus 89.2%), and not Medicare-Medicaid dual eligible (6.7% versus 8.8%). The median hospital-level rate of HHC use was 31.0% (interquartile range, 13.7%-54.5%) and ranged from 0% to 94.2%. Nearly 30% of the interhospital variation in HHC use was attributed to the discharging hospital (intraclass correlation coefficient, 0.296 [95% CI, 0.275-0.318]). Compared with non-HHC users, those using HHC were less likely to have a readmission or emergency department visit, were more likely to enroll in cardiac rehabilitation, and had modestly higher mortality within 30 or 90 days of discharge. A third of Medicare beneficiaries undergoing coronary artery bypass grafting used HHC within 30 days of discharge, with wide interhospital variation in use and mixed associations with clinical outcomes and health care utilization.

Evaluating the accuracy of a state-of-the-art large language model for prediction of admissions from the emergency room.

Artificial intelligence (AI) and large language models (LLMs) can play a critical role in emergency room operations by augmenting decision-making about patient admission. However, there are no studies for LLMs using real-world data and scenarios, in comparison to and being informed by traditional supervised machine learning (ML) models. We evaluated the performance of GPT-4 for predicting patient admissions from emergency department (ED) visits. We compared performance to traditional ML models both naively and when informed by few-shot examples and/or numerical probabilities. We conducted a retrospective study using electronic health records across 7 NYC hospitals. We trained Bio-Clinical-BERT and XGBoost (XGB) models on unstructured and structured data, respectively, and created an ensemble model reflecting ML performance. We then assessed GPT-4 capabilities in many scenarios: through Zero-shot, Few-shot with and without retrieval-augmented generation (RAG), and with and without ML numerical probabilities. The Ensemble ML model achieved an area under the receiver operating characteristic curve (AUC) of 0.88, an area under the precision-recall curve (AUPRC) of 0.72 and an accuracy of 82.9%. The naïve GPT-4's performance (0.79 AUC, 0.48 AUPRC, and 77.5% accuracy) showed substantial improvement when given limited, relevant data to learn from (ie, RAG) and underlying ML probabilities (0.87 AUC, 0.71 AUPRC, and 83.1% accuracy). Interestingly, RAG alone boosted performance to near peak levels (0.82 AUC, 0.56 AUPRC, and 81.3% accuracy). The naïve LLM had limited performance but showed significant improvement in predicting ED admissions when supplemented with real-world examples to learn from, particularly through RAG, and/or numerical probabilities from traditional ML models. Its peak performance, although slightly lower than the pure ML model, is noteworthy given its potential for providing reasoning behind predictions. Further refinement of LLMs with real-world data is necessary for successful integration as decision-support tools in care settings.

Estimating the size and scope of the academic emergency physician workforce.

Academic emergency medicine (EM) is foundational to the EM specialty through the development of new knowledge and clinical training of resident physicians. Despite recent increased attention to the future of the EM workforce, no evaluations have specifically characterized the U.S. academic EM workforce. We sought to estimate the national proportion of emergency physicians (EPs) identified as academic and the proportion of emergency department (ED) visits that take place at academic sites. We performed a cross-sectional analysis of EPs and EDs using data from the American Hospital Association, the Centers for Medicare & Medicaid Services, and Doximity's Residency Navigator. EPs were identified as "academic" if they were affiliated with at least one facility determined to be academic, defined as EDs officially designated by the Accreditation Council for Graduate Medical Education (ACGME) as clinical training sites at accredited EM residency programs. Our primary outcomes were to estimate the national proportion of EPs identified as academic and the proportion of ED visits performed at academic sites. Our analytic sample included 26,937 EPs practicing clinically across 4920 EDs and providing care during 130,471,386 ED visits. Among EPs, 11,720 (43.5%) were identified as academic, and among EDs, 635 (12.9%) were identified as academic sites, including 585 adult/general sites, 45 pediatric-specific sites, and 10 sites affiliated with the Department of Veterans Affairs. In 2021, academic EDs provided care for 42,794,106 ED visits or 32.8% of all ED visits nationally. Approximately four in 10 EPs practice in at least one clinical training site affiliated with an ACGME-accredited EM residency program, and approximately one in three ED visits nationally occur in these academic EDs. We encourage further work using alternative definitions of an academic EPs and EDs, along with longitudinal research to identify trends in the workforce's composition.

A protocol to determine the acceptability and feasibility of a pilot intervention emergency department virtual observation unit fall prevention program

BackgroundAs a third of all community dwellers aged 65+ fall each year, falls are common reasons for older adults to present to an Emergency Department (ED). Although EDs should assess patients’ multifactorial fall risks to prevent future fall-related injuries, this frequently does not occur. We describe our protocol to determine the feasibility, acceptability, and safety of a pilot ED Virtual Observation Unit (VOU) Falls program.MethodsTo ensure standardized conduct and reporting, the Standard Protocol Items for Intervention Trials (SPIRIT) guidelines will be used. The VOU is a program where patients are sent home from the ED but are part of a virtual observation unit in that they can call on-call ED physicians while they are being treated for conditions such as cellulitis, congestive heart failure, or pneumonia. A paramedic conducts daily visits with the patient and facilitates a telemedicine consult with an ED physician. VOU nursing staff conduct daily assessments of patients via telemedicine. The ED VOU Falls program is one of the VOU pathways and is a multi-component fall prevention program for fall patients who present after an ED visit. The paramedic conducts a home safety evaluation, a Timed Up and Go Test (TUG). During the VOU visit, the ED physician conducts a telemedicine visit, while the paramedic is visiting the home, to review patients' fall-risk-increasing drugs and their TUG test. We will determine feasibility by calculating rates of patient enrollment refusal, and adherence to fall-risk prevention recommendations using information from 3-month follow-up telephone calls, as well as qualitative interviews with the paramedics. We will determine the acceptability of the ED VOU Falls program based on patient and provider surveys using a Likert scale. We will ask VOU nursing staff to report any safety issues encountered while the patient is in the ED VOU Falls program (e.g., tripping hazards). We will use the chi-square test or Fisher’s exact test for categorical variables, Student’s t-test for continuous variables, and Mann-Whitney for nonparametric data. We will review interview transcripts and generate codes. Codes will then be extracted and organized into concepts to generate an overall theme following grounded theory methods. This is a pilot study; hence, results cannot be extrapolated. However, a definite trial would be the next step in the future to determine if such a program could be implemented as part of fall prevention interventions.DiscussionThis study will provide insights into the feasibility and acceptability of a novel ED VOU Falls program with the aim of ultimately decreasing falls. In the future, such a program could be implemented as part of fall prevention interventions.

Community-Acquired Pneumonia Diagnosis Following Emergency Department Visits for Respiratory Illness.

Community-acquired pneumonia (CAP) is often considered for children presenting to the emergency department (ED) with respiratory symptoms. It is unclear how often children are diagnosed with CAP following an ED visit for respiratory illness. We performed a retrospective case-control study to evaluate 7-day CAP diagnosis among children 3 months to 18 years discharged from the ED with respiratory illness from 2011 to 2021 and who receive care at 4 hospital-affiliated primary care clinics. Logistic regression was performed to assess for predictors of 7-day CAP diagnosis. Seventy-four (0.7%, 95% confidence interval [CI] = 0.6%, 0.9%) of 10 329 children were diagnosed with CAP within 7 days, and fever at the index visit was associated with increased odds of diagnosis (odds ratio [OR] = 3.32, 95% CI = 1.75-6.28). Community-acquired pneumonia diagnosis after discharge from the ED with respiratory illness is rare, even among children who are febrile at time of initial evaluation.

Enhancing Patient Flow in Emergency Departments: A Machine Learning and Simulation-Based Resource Scheduling Approach

The efficient scheduling of resources within emergency departments (EDs) is crucial to minimizing patient length of stay (LoS) times and maximizing the utilization of limited resources. Reducing patient wait times can enhance the operation of emergency departments and improve patient satisfaction and the quality of medical care. This study develops a simulation model using Discrete Event Simulation (DES) methodology, examining six resource scheduling policies that consider different combinations of general and senior physicians. By leveraging six scheduling policies and machine learning techniques, this model dynamically identifies the most effective scheduling policy, based on a comprehensive dataset of ED visits in South Korea. The ED simulation achieves an accuracy rate of 90% and demonstrates that our proposed integrated machine learning approach reduces average length of stay (LoS) to approximately 322.91 min, compared to 327.10 min under traditional methods. This study underscores the potential of integrating DES and machine learning to enhance resource management in EDs.

Black-White Disparities in Asthma Hospitalizations and ED Visits Among Medicaid-Enrolled Children.

Asthma is a common, potentially serious childhood chronic condition that disproportionately afflicts Black children. Hospitalizations and emergency department (ED) visits for asthma can often be prevented. Nearly half of children with asthma are covered by Medicaid, which should facilitate access to care to manage and treat symptoms. We provide new evidence on racial disparities in asthma hospitalizations and ED visits among Medicaid-enrolled children. We used comprehensive Medicaid claims data from the Transformed Medicaid Statistical Information System. Our study population included 279 985 Medicaid-enrolled children with diagnosed asthma. We identified asthma hospitalizations and ED visits occurring in 2019. We estimated differences in the odds of asthma hospitalizations and ED visits for non-Hispanic Black versus non-Hispanic white children, adjusting for sex, age, Medicaid eligibility group, Medicaid plan type, state, and rurality. In 2019, among Black children with asthma, 1.2% had an asthma hospitalization and 8.0% had an asthma ED visit compared with 0.5% and 3.4% of white children with a hospitalization and ED visit, respectively. After adjusting for other characteristics, the rates for Black children were more than twice the rates for white children (hospitalization adjusted odds ratio 2.45, 95% confidence interval 2.23-2.69; ED adjusted odds ratio 2.42; 95% confidence interval 2.33-2.51). There are stark racial disparities in asthma hospitalizations and ED visits among Medicaid-enrolled children with asthma. To diminish these disparities, it will be important to implement solutions that address poor quality care, discriminatory treatment in health care settings, and the structural factors that disproportionately expose Black children to asthma triggers and access barriers.

The effect of the COVID-19 pandemic on pediatric emergency department utilization in three regions in Switzerland.

The COVID-19 pandemic was associated with a decrease in emergency department (ED) visits. However, contradictory, and sparse data regarding children could not yet answer the question, how pediatric ED utilization evolved throughout the pandemic. Our objectives were to investigate the impact of the pandemic in three language regions of Switzerland by analyzing trends over time, describe regional differences, and address implications for future healthcare. We conducted a retrospective, longitudinal cohort study at three Swiss tertiary pediatric EDs (March 1st, 2018-February 28th, 2022), analyzing the numbers of ED visits (including patients` age, triage categories, and urgent vs. non-urgent cases). The impact of COVID-19 related non-pharmaceutical interventions (NPIs) on pediatric ED utilization was assessed by interrupted time series (ITS) modelling. Based on 304'438 ED visits, we found a drop of nearly 50% at the onset of NPIs, followed by a gradual recovery. This primarily affected children 0-4years, and both non-urgent and urgent cases. However, the decline in urgent visits appeared to be more pronounced in two centers compared to a third, where also hospitalization rates did not decrease significantly during the pandemic. A subgroup analysis showed a significant decrease in respiratory and gastrointestinal diseases, and an increase in the proportion of trauma patients during the pandemic. The COVID-19 pandemic had substantial effects on number and reasons for pediatric ED visits, particularly among children 0-4years. Despite equal regulatory conditions, the utilization dynamics varied markedly between the three regions, highlighting the multifactorial modification of pediatric ED utilization during the pandemic. Furthermore, future policy decisions should take regional differences into account.

Clinical and healthcare use outcomes after cessation of long term opioid treatment due to prescriber workforce exit: quasi-experimental difference-in-differences study

ABSTRACTObjectiveTo examine the association between prescriber workforce exit, long term opioid treatment discontinuation, and clinical outcomes.DesignQuasi-experimental difference-in-differences studySetting20% sample of US Medicare beneficiaries, 2011-18.ParticipantsPeople receiving long term opioid treatment whose...

Multimorbidity prevalence and health outcome prediction: assessing the impact of lookback periods, disease count, and definition criteria in health administrative data at the population-based level

BackgroundHealth administrative databases play a crucial role in population-level multimorbidity surveillance. Determining the appropriate retrospective or lookback period (LP) for observing prevalent and newly diagnosed diseases in administrative data presents challenge in estimating multimorbidity prevalence and predicting health outcome. The aim of this population-based study was to assess the impact of LP on multimorbidity prevalence and health outcomes prediction across three multimorbidity definitions, three lists of diseases used for multimorbidity assessment, and six health outcomes.MethodsWe conducted a population-based study including all individuals ages > 65 years on April 1st, 2019, in Québec, Canada. We considered three lists of diseases labeled according to the number of chronic conditions it considered: (1) L60 included 60 chronic conditions from the International Classification of Diseases (ICD); (2) L20 included a core of 20 chronic conditions; and (3) L31 included 31 chronic conditions from the Charlson and Elixhauser indices. For each list, we: (1) measured multimorbidity prevalence for three multimorbidity definitions (at least two [MM2+], three [MM3+] or four (MM4+) chronic conditions); and (2) evaluated capacity (c-statistic) to predict 1-year outcomes (mortality, hospitalisation, polypharmacy, and general practitioner, specialist, or emergency department visits) using LPs ranging from 1 to 20 years.ResultsIncrease in multimorbidity prevalence decelerated after 5–10 years (e.g., MM2+, L31: LP = 1y: 14%, LP = 10y: 58%, LP = 20y: 69%). Within the 5–10 years LP range, predictive performance was better for L20 than L60 (e.g., LP = 7y, mortality, MM3+: L20 [0.798;95%CI:0.797-0.800] vs. L60 [0.779; 95%CI:0.777–0.781]) and typically better for MM3 + and MM4 + definitions (e.g., LP = 7y, mortality, L60: MM4+ [0.788;95%CI:0.786–0.790] vs. MM2+ [0.768;95%CI:0.766–0.770]).ConclusionsIn our databases, ten years of data was required for stable estimation of multimorbidity prevalence. Within that range, the L20 and multimorbidity definitions MM3 + or MM4 + reached maximal predictive performance.

Longitudinal comparisons in health, participation, life satisfaction, and depressive symptoms from pre-to post Covid 19 pandemic among people with spinal cord injuries.

To identify changes in health, healthcare utilization, participation, and life satisfaction and depressive symptoms from pre-pandemic to post-pandemic among ambulatory and nonambulatory participants with spinal cord injury (SCI). Longitudinal study with the first measurement taken within three months prior to pandemic restrictions and two follow-ups at approximately one year intervals. Medical University in the Southeastern USA. 219 adult participants with SCI, including ambulatory (n= 155) and nonambulatory (n= 64). Self-report assessment of health outcomes from the Behavioral Risk Factor Surveillance System; health service utilization including physician visits, emergency department visits and hospitalizations; items from the Craig Handicap Assessment Reporting Technique; three life satisfaction scales from the Life Situation Questionnaire, and the brief version of the Patient Health Questionnaire. Not applicable. After using a Z score correction for nonnormality, none of the time effects or interaction effects of time by ambulatory status were significant. Six comparisons between ambulatory and nonambulatory were statistically significant. Ambulatory participants reported significantly 3 more days in poor physical health (p = .02) and 2 more days feeling worried, tense, or anxious in the last 30 days (p = .03). They visited emergency department on 0.3 fewer occasions (p = .02), while reporting leaving the house 1 more day every week (p = .02), 2 hours more sitting tolerance (p<.01), and 1 higher score of vocational satisfaction (p = .03). The absence of statistically significant changes from pre- to post-pandemic and the absence of time by ambulatory status interactions suggest stability of outcomes, even in the presence of pandemic challenges.

The Impact of Palliative Care Consultation on Aggressive Medical Interventions in End-of-life Among Patients with Metastatic Breast Cancer: Insights from the U.S. National Patient Sample.

Advancement in treatment has led to prolonged survival and a rising number of women living with metastatic breast cancer (MBC) in the United States. Due to its high symptom burden, it is recommended that palliative care be integrated into the standard care to help improve quality of life. However, little is known about the use of palliative care among MBC patients in the nation. To determine utilization of palliative care consult (PCC) after metastasis and the influence of PCC on healthcare utilization in the end of life among women living with MBC in the US. This retrospective cohort study examined a national electronic health record database to quantify the PCC use after metastasis diagnosis until death and the associations of PCC with Emergency Department (ED), Intensive Care Unit (ICU), and chemotherapies in the end-of-life women (age ≥ 18 years) living with MBC. From a cohort of 2615 deceased MBC patients, 37% received PCC in the last 6 months of life. Patients who had received PCC in the end-of-life were more likely to be hospitalized, admitted to ED and ICU, and receive chemotherapies in the last 60 days before death. However, patients who had received end-of-life PCC had less hospital and ED visits and received less chemotherapies after PCC initiated. While PCC can reduce end-of-life aggressive interventions, it was underutilized among patients with MBC in the end-of-life. A myriad of clinical and patient factors may still challenge timely consultation. We urge for future endeavors in developing strategies to remove barriers in the implementation, especially earlier in the disease course, to assure timely PC treatments and reduce discomfort amid aggressive interventions for MBC.

The association between population health management tools and clinician burnout in the United States VA primary care patient-centered medical home

BackgroundTechnological burden and medical complexity are significant drivers of clinician burnout. Electronic health record(EHR)-based population health management tools can be used to identify high-risk patient populations and implement prophylactic health practices. Their impact on clinician burnout, however, is not well understood. Our objective was to assess the relationship between ratings of EHR-based population health management tools and clinician burnout.MethodsWe conducted cross-sectional analyses of 2018 national Veterans Health Administration(VA) primary care personnel survey, administered as an online survey to all VA primary care personnel (n = 4257, response rate = 17.7%), using bivariate and multivariate logistic regressions. Our analytical sample included providers (medical doctors, nurse practitioners, physicians’ assistants) and nurses (registered nurses, licensed practical nurses). The outcomes included two items measuring high burnout. Primary predictors included importance ratings of 10 population health management tools (eg. VA risk prediction algorithm, recent hospitalizations and emergency department visits, etc.).ResultsHigh ratings of 9 tools were associated with lower odds of high burnout, independent of covariates including VA tenure, team role, gender, ethnicity, staffing, and training. For example, clinicians who rated the risk prediction algorithm as important were less likely to report high burnout levels than those who did not use or did not know about the tool (OR 0.73; CI 0.61-0.87), and they were less likely to report frequent burnout (once per week or more) (OR 0.71; CI 0.60-0.84).ConclusionsBurned-out clinicians may not consider the EHR-based tools important and may not be using them to perform care management. Tools that create additional technological burden may need adaptation to become more accessible, more intuitive, and less burdensome to use. Finding ways to improve the use of tools that streamline the work of population health management and/or result in less workload due to patients with poorly managed chronic conditions may alleviate burnout. More research is needed to understand the causal directional of the association between burnout and ratings of population health management tools.

Emergency Department Risk Model: Timely Identification of Patients for Outpatient Care Coordination

Emergency Department Risk Model: Timely Identification of Patients for Outpatient Care Coordination

Healthcare utilisation by diabetic patients in Denmark: the role of primary care in reducing emergency visits

Improving the management of diabetic patients is receiving increasing attention in the health policy agenda due to increasing prevalence in the population and raising pressure on healthcare resources. This paper examines the determinants of healthcare services utilisation in patients with type-2 diabetes, investigating the potential substitution effect of general practice visits on the utilisation of emergency department visits. By using rich longitudinal data from Denmark and a bivariate econometric model, our analysis highlights primary care services that are more effective in preventing emergency department visits and socioeconomic groups of patients with a weak substitution response. Our results suggest that empowering primary care services, such as preventive assessment visits, may contribute to reducing emergency department visits significantly. Moreover, special attention should be devoted to vulnerable groups, such as patients from low socioeconomic background and older patients, who may find more difficult achieving a large substitution response.

Pilot Process Evaluation of the Supporting Older Adults at Risk Model: A RE-AIM Approach.

Despite evidence supporting transitional care models, hospitals report challenges implementing and sustaining them. The Discharge to Assess (D2A) Model is an innovative solution to this problem but required translation from a national health system context to an U.S.-based context. We translated the central tenets of the D2A model to establish the Supporting Older Adults at Risk (SOAR) Model, which unfolds in three phases: Prepare, Transition, and Support. The purpose of this project was to conduct a process evaluation of the SOAR Model in practice using the RE-AIM Framework (Reach, Effectiveness, Adoption, Implementation, and Maintenance). Forty patients completed all SOAR Model components for a Reach of 21%. Patients averaged 80 years of age, 53% were female, and 64% Black/AA. SOAR significantly improved discharge before noon, time to first home visit, and use of the in-house pharmacy. SOAR also improved length of hospital stay, emergency department visits, and readmissions. Twenty-one of the 26 Implementation measures unfolded with 75% or greater fidelity. Sixteen of the 24 Adoption measures unfolded with 75% or greater fidelity. COVID-19 limited Maintenance. Given the model unfolds across settings over time, requiring adoption from interprofessional team members, patients, and families, future work should focus on improving reach and adoption.

Metformin use associated with lower rate of hospitalization for influenza in individuals with diabetes.

To investigate if patients with diabetes taking metformin have better outcomes versus those not taking metformin following an emergency room visit for influenza. Using electronic medical records, we performed a retrospective chart review of all adult patients with a diagnosis of diabetes seen in any Duke University Medical Center-affiliated emergency department for influenza over a 6-year period. We documented patient characteristics and comorbidities, and compared outcomes for patients taking metformin versus patients not taking metformin using both univariable and multivariable analyses. Our primary outcome was hospital admission rate. Secondary outcomes were in-hospital length of stay and in-hospital death. Our cohort included 1023 adult patients with diabetes, of whom 59.9% were female. The mean age was 62.9 years, 58.4% were African American, 36.1% were White, and 81.9% were obese or overweight. Of these patients, 347 (34%) were taking metformin. Patients with diabetes taking metformin were less likely to be hospitalized following an emergency department visit for influenza than patients with diabetes not taking metformin (56.8% vs. 70.1%; p < 0.001). Of those patients admitted, there was no statistically significant difference in length of stay or death. In patients with diabetes, metformin use is associated with lower rate of hospitalization following an emergency department visit for influenza.

The feasibility, acceptability and appropriateness of screening for frailty in Australians aged 75 years and over attending Australian general practice

Background Globally, frailty is associated with a high prevalence of avoidable hospital admissions and emergency department visits, with substantial associated healthcare and personal costs. International guidelines recommend incorporation of frailty identification and care planning into routine primary care workflow to support patients who may be identified as pre-frail/frail. Our study aimed to: (1) determine the feasibility, acceptability, appropriateness and determinants of implementing a validated FRAIL Scale screening Tool into general practices in two disparate Australian regions (Sydney North and Brisbane South); and (2) map the resources and referral options required to support frailty management and potential reversal. Methods Using the FRAIL Scale Tool, practices screened eligible patients (aged ≥75 years) for risk of frailty and referred to associated management options. The percentage of patients identified as frail/pre-frail, and management options and referrals made by practice staff for those identified as frail/pre-frail were recorded. Semi-structured qualitative interviews were conducted with practice staff to understand the feasibility, acceptability, appropriateness and determinants of implementing the Tool. Results The Tool was implemented by 19 general practices in two Primary Health Networks and 1071 consenting patients were assessed. Overall, 80% of patients (n = 860) met the criterion for frailty: 33% of patients (n = 352) were frail, and 47% were pre-frail (n = 508). They were predominantly then referred for exercise prescription, medication reviews and geriatric assessment. The Tool was acceptable to staff and patients and compatible with practice workflows. Conclusions This study demonstrates that frailty is identified frequently in Australians aged ≥75 years who visit their general practice. It’s identification, linked with management support to reverse or reduce frailty risk, can be readily incorporated into the Medicare-funded annual 75+ Health Assessment.

Telehealth Expansion and Medicare Beneficiaries’ Care Quality and Access

Understanding the association of telehealth use with health care outcomes is fundamental to determining whether telehealth waivers implemented during the COVID-19 public health emergency should be made permanent. The current literature has yielded inconclusive findings owing to its focus on select states, practices, or health care systems. To estimate the association of telehealth use with outcomes for all Medicare fee-for-service (FFS) beneficiaries by comparing hospital service areas (HSAs) with different levels of telehealth use. This US population-based, retrospective cohort study was conducted from July 2022 to April 2023. Participants included Medicare claims of beneficiaries attributed to HSAs with FFS enrollment in Parts A and B. Low, medium, or high tercile of telehealth use created by ranking HSAs according to the number of telehealth visits per 1000 beneficiaries. The primary outcomes were quality (ambulatory care-sensitive [ACS] hospitalizations and emergency department [ED] visits per 1000 FFS beneficiaries), access to care (clinician encounters per FFS beneficiary), and cost (total cost of care for Part A and/or B services per FFS Medicare beneficiary) determined with a difference-in-difference analysis. In this cohort study of claims from approximately 30 million Medicare beneficiaries (mean [SD] age in 2019, 71.04 [1.67] years; mean [SD] percentage female in 2019, 53.83% [2.14%]) within 3436 HSAs, between the second half of 2019 and the second half of 2021, mean ACS hospitalizations and ED visits declined sharply, mean clinician encounters per beneficiary declined slightly, and mean total cost of care per beneficiary per semester increased slightly. Compared with the low group, the high group had more ACS hospitalizations (1.63 additional hospitalizations per 1000 beneficiaries; 95% CI, 1.03-2.22 hospitalizations), more clinician encounters (0.30 additional encounters per beneficiary per semester; 95% CI, 0.23-0.38 encounters), and higher total cost of care ($164.99 higher cost per beneficiary per semester; 95% CI, $101.03-$228.96). There was no statistically significant difference in ACS ED visits between the low and high groups. In this cohort study of Medicare beneficiaries across all 3436 HSAs, high levels of telehealth use were associated with more clinician encounters, more ACS hospitalizations, and higher total health care costs. COVID-19 cases were still high during the period of study, which suggests that these findings partially reflect a higher capacity for providing health services in HSAs with higher telehealth intensity than other HSAs.

Applying natural language processing to identify emergency department and observation encounters for worsening heart failure.

Worsening heart failure (WHF) events occurring in non-inpatient settings are becoming increasingly recognized, with implications for prognostication. We evaluate the performance of a natural language processing (NLP)-based approach compared with traditional diagnostic coding for non-inpatient clinical encounters and left ventricular ejection fraction (LVEF). We compared characteristics for encounters that did vs. did not meet WHF criteria, stratified by care setting [i.e. emergency department (ED) and observation stay]. Overall, 8407 (22%) encounters met NLP-based criteria for WHF (3909 ED visits and 4498 observation stays). The use of an NLP-derived definition adjudicated 3983 (12%) of non-primary HF diagnoses as meeting consensus definitions for WHF. The most common diagnosis indicated in these encounters was dyspnoea. Results were primarily driven by observation stays, in which 2205 (23%) encounters with a secondary HF diagnosis met the WHF definition by NLP. The use of standard claims-based adjudication for primary diagnosis in the non-inpatient setting may lead to misclassification of WHF events in the ED and overestimate observation stays. Primary diagnoses alone may underestimate the burden of WHF in non-hospitalized settings.