Patient and clinician perspectives of a remote monitoring program for COVID-19 and lessons for future programs.

COVID Watch is a remote patient monitoring program implemented during the pandemic to support home dwelling patients with COVID-19. The program conferred a large survival advantage. We conducted semi-structured interviews of 85 patients and clinicians using COVID Watch to understand how to design such programs even better. Patients and clinicians found COVID Watch to be comforting and beneficial, but both groups desired more clarity about the purpose and timing of enrollment and alternatives to text-messages to adapt to patients' preferences as these may have limited engagement and enrollment among marginalized patient populations. Because inclusiveness and equity are important elements of programmatic success, future programs will need flexible and multi-channel human-to-human communication pathways for complex clinical interactions or for patients who do not desire tech-first approaches.

Comparative Effectiveness of an Automated Text Messaging Service for Monitoring COVID-19 at Home.

BACKGROUND: Although most patients with SARS-CoV-2 infection can be safely managed at home, the need for hospitalization can arise suddenly. OBJECTIVE: To determine whether enrollment in an automated remote monitoring service for community-dwelling adults with COVID-19 at home ("COVID Watch") was associated with improved mortality. DESIGN: Retrospective cohort analysis. SETTING: Mid-Atlantic academic health system in the United States. PARTICIPANTS: Outpatients who tested positive for SARS-CoV-2 between 23 March and 30 November 2020. INTERVENTION: The COVID Watch service consists of twice-daily, automated text message check-ins with an option to report worsening symptoms at any time. All escalations were managed 24 hours a day, 7 days a week by dedicated telemedicine clinicians. MEASUREMENTS: Thirty- and 60-day outcomes of patients enrolled in COVID Watch were compared with those of patients who were eligible to enroll but received usual care. The primary outcome was death at 30 days. Secondary outcomes included emergency department (ED) visits and hospitalizations. Treatment effects were estimated with propensity score-weighted risk adjustment models. RESULTS: A total of 3488 patients enrolled in COVID Watch and 4377 usual care control participants were compared with propensity score weighted models. At 30 days, COVID Watch patients had an odds ratio for death of 0.32 (95% CI, 0.12 to 0.72), with 1.8 fewer deaths per 1000 patients (CI, 0.5 to 3.1) (P = 0.005); at 60 days, the difference was 2.5 fewer deaths per 1000 patients (CI, 0.9 to 4.0) (P = 0.002). Patients in COVID Watch had more telemedicine encounters, ED visits, and hospitalizations and presented to the ED sooner (mean, 1.9 days sooner [CI, 0.9 to 2.9 days]; all P < 0.001). LIMITATION: Observational study with the potential for unobserved confounding. CONCLUSION: Enrollment of outpatients with COVID-19 in an automated remote monitoring service was associated with reduced mortality, potentially explained by more frequent telemedicine encounters and more frequent and earlier presentation to the ED. PRIMARY FUNDING SOURCE: Patient-Centered Outcomes Research Institute.

The combined effects of physical frailty and cognitive impairment on emergency department- versus direct-admission hospitalizations.

BACKGROUND: We aimed to study whether physical frailty and cognitive impairment (CI) increase the risk of recurrent hospitalizations in older adults, independent of comorbidity, and disability. METHODS: Two thousand five hundred forty-nine community-dwelling participants from the National Health and Aging Trends Study (NHATS) with 3 + years of continuous Medicare coverage from linked claims data were included. We used the marginal means/rates recurrent events model to investigate the association of baseline CI (mild CI or dementia) and physical frailty, separately and synergistically, with the number of all-source vs. Emergency Department (ED)-admission vs. direct admission hospitalizations over 2 years. RESULTS: 17.8% of participants had at least one ED-admission hospitalization; 12.7% had at least one direct admission hospitalization. Frailty and CI, modeled separately, were both significantly associated with risk of recurrent all-source (Rate Ratio (RR) = 1.24 for frailty, 1.21 for CI; p < .05) and ED-admission (RR = 1.49 for frailty, 1.41 for CI; p < .05) hospitalizations but not direct admission, adjusting for socio-demographics, obesity, comorbidity and disability. When CI and frailty were examined together, 64.3% had neither (Unimpaired); 28.1% CI only; 3.5% Frailty only; 4.1% CI + Frailty. Compared to those Unimpaired, CI alone and CI + Frailty were predictive of all-source (RR = 1.20, 1.48, p < .05) and ED-admission (RR = 1.36, 2.14, p < .05) hospitalizations, but not direct admission, in our adjusted model. CONCLUSIONS: Older adults with both CI and frailty experienced the highest risk for recurrent ED-admission hospitalizations. Timely recognition of older adults with CI and frailty is needed, paying special attention to managing cognitive impairment to mitigate preventable causes of ED admissions and potentiate alternatives to hospitalization.

Comparing resource use between paediatric emergency department visits by triage level.

OBJECTIVE: The majority of paediatric ED visits result in discharge but little is known about what ED resources are deployed for these visits. The goal of this study was to understand the utilisation of diagnostic testing, procedures and hospital admission for paediatric ED visits triaged as 'non-urgent'. STUDY DESIGN: We examined US ED visits for children aged 0-17 years from 1 January 2009 to 31 December 2011 in the National Hospital Ambulatory Medical Care Survey. Visits triaged on arrival as 'non-urgent' (level 5) were compared with urgent visits (triage levels 1-4) for resource use and disposition. Sensitivity and specificity of triage for predicting resource use and disposition were assessed. RESULTS: Among 21 052 observations, representing 86 620 988 visits, 11.1% were triaged as 'non-urgent'. Diagnostic services were provided during 37.6% (95% CI 33.9% to 41.4%) of non-urgent and 55.2% (95% CI 53.3% to 57.2%) of urgent visits. Procedures were performed in 23.9% (95% CI 20.4% to 27.3%) of non-urgent and 33.9% (95% CI 31.2% to 35.9%) of urgent visits. 1.7% (95% CI 0.09% to 2.6%) of the non-urgent visits resulted in admission, with 0.08% (95% CI 0% to 0.2%) to critical care units, compared with 4.4% (95% CI 3.6% to 5.2%) of the urgent visits, with 0.3% (95% CI 0.2% to 0.4%) to critical care. Despite some substantial differences in the rates of resource use, triage score had poor sensitivity for identifying patients who did not receive ED tests, procedures or admission. CONCLUSION: A significant percentage of ED patients with non-urgent ED triage scores received ED testing and procedures. More work is needed to improve methods of prospectively identifying patients with low acuity complaints who do not need significant ED resources.

The Scope and Legal Implications of Tracking Technologies on Hospital Websites.

This Viewpoint analyzes the scope and legal implications of tracking on hospital websites, including potential HIPAA and state privacy law violations, and suggests that hospitals limit such tracking.

Access to Primary Care Appointments Following 2014 Insurance Expansions.

PURPOSE: The Patient Protection and Affordable Care Act (ACA) expanded coverage to roughly 12 million individuals by mid-2014 and 20 million by 2016, raising concern about the capacity of the primary care system to absorb these individuals. We set out to determine how justified the concern was. METHODS: We used an audit design in which simulated patients called primary care practices seeking new-patient appointments in 10 diverse states (Arkansas, Georgia, Iowa, Illinois, Massachusetts, Montana, New Jersey, Oregon, Pennsylvania, and Texas) from November 2012 through March 2013 and from May 2014 through July 2014, before and after the major ACA insurance expansions. Callers were randomly assigned to scripts specifying either private or Medicaid insurance and called only offices identified as "in network" for each plan. RESULTS: We completed 5,385 private insurance and 4,352 Medicaid calls in 2012-2013 and 2,424 private insurance and 2,474 Medicaid calls in 2014. Overall appointment rates for private insurance remained stable from 2012 (84.7%) to 2014 (85.8%) with Massachusetts and Pennsylvania experiencing significant increases. Overall, Medicaid appointment rates increased 9.7 percentage points (57.9% to 67.6%) with substantial variation by state. Across all callers, median wait times for those obtaining an appointment were 7 days in 2012 and 5 days in 2014, but the difference was not statistically significant. CONCLUSIONS: Contrary to widespread concern, we find no evidence that the millions of individuals newly insured through the ACA decreased new-patient appointment availability across 10 states as shown by stable wait times and appointment rates for private insurance as of mid-2014.

Forgiven but not Relieved: US Physician Workforce Consequences of Changes to Public Service Loan Forgiveness.

Public Service Loan Forgiveness (PSLF) was established in 2007 for public sector and nonprofit enterprise employees to pursue educational loan forgiveness. Under PSLF, graduates are offered complete loan forgiveness after 120 qualifying monthly payments while employed at public or nonprofit institutions, including payments made during residency for physicians. In response to concerns that PSLF will heavily subsidize lawyers, doctors, and other professionals, the President's 2017 budget proposes limiting maximum forgiveness. Using data from the Association of American Medical Colleges Graduation Questionnaire (n = 55,905; response rate of 80 %), we found that intended participation in PSLF among medical school graduates grew 20 % per year since 2010. Future primary care physicians intend to use PSLF more than programs that were historically designed to promote primary care, such as the National Health Service Corp (NHSC). The federal government's projected cost of PSLF will reach over $316 million for 2014 graduates (net present value), approximately seven times the annual contributions from the NHSC. The proposed cap will reduce the total anticipated forgiveness by nearly two-thirds and substantially reduce subsidies for physicians. More targeted measures of loan forgiveness could be considered, such as making forgiveness contingent on pursuing specialties that society needs or practicing in shortage areas.

Trends in hospital ED closures nationwide and across Medicaid expansion, 2006-2013.

STUDY HYPOTHESIS: Low reimbursement from the uninsured has been claimed to threaten hospital finances and even hospital emergency department (ED) closure. We hypothesized in advance of beginning data collection that states that expanded Medicaid ("expansion states") under the 2010 Patient Protection and Affordable Care Act would experience a reduced rate of ED closure compared with states that did not. METHODS: We compiled a national census of EDs from 2006 through 2013 from federal databases and manually confirmed each closure. We used difference-in-differences regression on this longitudinal panel to compare the probability over time that a hospital was in operation in expansion states to nonexpansion states. RESULTS: The number of hospitals grew every year nationally and in nonexpansion states. In expansion states, the number fell from 2027 in 2009 to 2019 in 2010, not surpassing the 2009 peak until 2012. In regression estimates, hospitals in expansion states were 2.2% (95% confidence interval, 0.3%-4.1%) less likely to be in operation after 2010 compared with the trend in nonexpansion states. CONCLUSIONS: States that expanded Medicaid experienced increased, rather than reduced, ED closure rates from 2010 through 2013. The financial benefits of the Affordable Care Act may be poorly targeted to the hospitals most vulnerable to closure.

ED, heal thyself.

Emergency department (ED) wait times have continued to worsen despite receiving considerable attention for more than 2 decades and despite the availability of a variety of methods to restructure care in a more streamlined fashion. This article offers an economic framework that abstracts away from the details of operations research to understand the fundamental disincentives to improving wait times. Hospitals that reduce wait times are financially penalized if they must provide more uncompensated care as a result. Pending changes under the Patient Protection and Affordable Care Act are considered. We find that the likely effect of the Patient Protection and Affordable Care Act's insurance expansion is to reduce this penalty for improving ED wait times. Consequently, mandating adoption of solutions to ED crowding may be unnecessary and counterproductive. If the insurance expansion is insufficient to fully solve the problem, the hospital value-based purchasing initiative should adopt wait times as a goal in its next iteration.

Identifying low acuity Emergency Department visits with a machine learning approach: The low acuity visit algorithms (LAVA).

OBJECTIVE: To improve the performance of International Classification of Disease (ICD) code rule-based algorithms for identifying low acuity Emergency Department (ED) visits by using machine learning methods and additional covariates. DATA SOURCES: We used secondary data on ED visits from the National Hospital Ambulatory Medical Survey (NHAMCS), from 2016 to 2020. STUDY DESIGN: We established baseline performance metrics with seven published algorithms consisting of International Classification of Disease, Tenth Revision codes used to identify low acuity ED visits. We then trained logistic regression, random forest, and gradient boosting (XGBoost) models to predict low acuity ED visits. Each model was trained on five different covariate sets of demographic and clinical data. Model performance was compared using a separate validation dataset. The primary performance metric was the probability that a visit identified by an algorithm as low acuity did not experience significant testing, treatment, or disposition (positive predictive value, PPV). Subgroup analyses assessed model performance across age, sex, and race/ethnicity. DATA COLLECTION: We used 2016-2019 NHAMCS data as the training set and 2020 NHAMCS data for validation. PRINCIPAL FINDINGS: The training and validation data consisted of 53,074 and 9542 observations, respectively. Among seven rule-based algorithms, the highest-performing had a PPV of 0.35 (95% CI [0.33, 0.36]). All model-based algorithms outperformed existing algorithms, with the least effective-random forest using only age and sex-improving PPV by 26% (up to 0.44; 95% CI [0.40, 0.48]). Logistic regression and XGBoost trained on all variables improved PPV by 83% (to 0.64; 95% CI [0.62, 0.66]). Multivariable models also demonstrated higher PPV across all three demographic subgroups. CONCLUSIONS: Machine learning models substantially outperform existing algorithms based on ICD codes in predicting low acuity ED visits. Variations in model performance across demographic groups highlight the need for further research to ensure their applicability and fairness across diverse populations.

Patient perceptions of behavioral flags in the emergency department: A qualitative analysis.

BACKGROUND: To combat increasing levels of violence in the emergency department (ED), hospitals have implemented several safety measures, including behavioral flags. These electronic health record (EHR)-based notifications alert future clinicians of past incidents of potentially threatening patient behavior, but observed racial disparities in their placement may unintentionally introduce bias in patient care. Little is known about how patients perceive these flags and the disparities that have been found in their placement. OBJECTIVE: This study aims to investigate patient perceptions and perceived benefits and harms associated with the use of behavioral flags. METHODS: Twenty-five semistructured qualitative interviews were conducted with a convenience sample of patients in the ED of a large, urban, academic medical center who did not have a behavioral flag in their EHR. Interviews lasted 10-20 min and were recorded then transcribed. Thematic analysis of deidentified transcripts took place in NVivo 20 software (QSR International) using a general inductive approach. RESULTS: Participant perceptions of behavioral flags varied, with both positive and negative opinions being shared. Five key themes, each with subthemes, were identified: (1) benefits of behavioral flags, (2) concerns and potential harms of flags, (3) transparency with patients, (4) equity, and (5) ideas for improvement. CONCLUSIONS: Patient perspectives on the use of behavioral flags in the ED vary. While many saw flags as a helpful tool to mitigate violence, concerns around negative impacts on care, transparency, and equity were also shared. Insights from this stakeholder perspective may allow for health systems to make flags more effective without compromising equity or patient ideals.

User Information Sharing and Hospital Website Privacy Policies.

Importance: Hospital websites frequently use tracking technologies that transfer user information to third parties. It is not known whether hospital websites include privacy policies that disclose relevant details regarding tracking. Objective: To determine whether hospital websites have accessible privacy policies and whether those policies contain key information related to third-party tracking. Design, Setting, and Participants: In this cross-sectional content analysis of website privacy policies of a nationally representative sample of nonfederal acute care hospitals, hospital websites were first measured to determine whether they included tracking technologies that transferred user information to third parties. Hospital website privacy policies were then identified using standardized searches. Policies were assessed for length and readability. Policy content was analyzed using a data abstraction form. Tracking measurement and privacy policy retrieval and analysis took place from November 2023 to January 2024. The prevalence of privacy policy characteristics was analyzed using standard descriptive statistics. Main Outcomes and Measures: The primary study outcome was the availability of a website privacy policy. Secondary outcomes were the length and readability of privacy policies and the inclusion of privacy policy content addressing user information collected by the website, potential uses of user information, third-party recipients of user information, and user rights regarding tracking and information collection. Results: Of 100 hospital websites, 96 (96.0%; 95% CI, 90.1%-98.9%) transferred user information to third parties. Privacy policies were found on 71 websites (71.0%; 95% CI, 61.6%-79.4%). Policies were a mean length of 2527 words (95% CI, 2058-2997 words) and were written at a mean grade level of 13.7 (95% CI, 13.4-14.1). Among 71 privacy policies, 69 (97.2%; 95% CI, 91.4%-99.5%) addressed types of user information automatically collected by the website, 70 (98.6%; 95% CI, 93.8%-99.9%) addressed how collected information would be used, 66 (93.0%; 95% CI, 85.3%-97.5%) addressed categories of third-party recipients of user information, and 40 (56.3%; 95% CI, 44.5%-67.7%) named specific third-party companies or services receiving user information. Conclusions and Relevance: In this cross-sectional study of hospital website privacy policies, a substantial number of hospital websites did not present users with adequate information about the privacy implications of website use, either because they lacked a privacy policy or had a privacy policy that contained limited content about third-party recipients of user information.