Emergency Department Dataset Research Articles

Comparing NEWS2, TRISS, RTS, SI, GAP, and MGAP in predicting early and total mortality rates in trauma patients based on emergency department data set: A diagnostic study

IntroductionIn the global context, trauma remains a significant cause of mortality, particularly among individuals under 50 years old with mortality and morbidity. Timely assessment, efficient triage, and appropriate post-trauma care can help mitigate long-term mortality and morbidity risks. While the trauma and injury severity score (TRISS) is considered accurate, its complexity limits clinical utility, prompting the development of new tools like the National Early Warning Score 2 (NEWS2) and others. Thus, this study aims to evaluate the predictive abilities of NEWS2, RTS (revised trauma score), TRISS, SI (shock index), GAP (Glasgow coma score, age, and systolic blood pressure), and MGAP (mechanism of injury, Glasgow coma score, age, and systolic blood pressure) scoring systems in trauma patient care to enhance outcomes and reduce mortality and morbidity rates. MethodsThis retrospective diagnostic study included a total of 6894 trauma patients with data collected from the emergency department. Patient information, including vital signs and demographic data, was recorded on datasheets completed by ED staff. Ethical guidelines were followed, and patient consent forms were not required due to the study's retrospective nature. Scoring systems such as NEWS2, RTS, TRISS, SI, GAP, and MGAP were calculated and interpreted based on specific formulas and scoring charts. ResultsOf 6894 patients, 2.83% died during hospitalization, and 0.72% within the first 24 hours. Factors associated with mortality included age, cause of injury, duration of hospitalization, and various physiological parameters. The majority of scoring systems are reliable in predicting mortality, except for the sock index (SI), with NEWS2 demonstrating the highest sensitivity and accuracy for predicting mortality rates, while TRISS and RTS were the most specific. GAP was also the most competent predictor of overall mortality. ConclusionMost scoring systems effectively predict mortality, with TRISS and NEWS2 demonstrating high sensitivity and specificity together in forecasting in-hospital mortality. NEWS2 stands out as a precise tool for mortality prediction in the emergency department due to its simplicity and quick application. However, further research is needed to identify the most effective predictive system in trauma patients.

Sex bias in pain management decisions

In the pursuit of mental and physical health, effective pain management stands as a cornerstone. Here, we examine a potential sex bias in pain management. Leveraging insights from psychological research showing that females' pain is stereotypically judged as less intense than males' pain, we hypothesize that there may be tangible differences in pain management decisions based on patients' sex. Our investigation spans emergency department (ED) datasets from two countries, including discharge notes of patients arriving with pain complaints (N = 21,851). Across these datasets, a consistent sex disparity emerges. Female patients are less likely to be prescribed pain-relief medications compared to males, and this disparity persists even after adjusting for patients' reported pain scores and numerous patient, physician, and ED variables. This disparity extends across medical practitioners, with both male and female physicians prescribing less pain-relief medications to females than to males. Additional analyses reveal that female patients' pain scores are 10% less likely to be recorded by nurses, and female patients spend an additional 30 min in the ED compared to male patients. A controlled experiment employing clinical vignettes reinforces our hypothesis, showing that nurses (N = 109) judge pain of female patients to be less intense than that of males. We argue that the findings reflect an undertreatment of female patients' pain. We discuss the troubling societal and medical implications of females' pain being overlooked and call for policy interventions to ensure equal pain treatment.

Potentially Avoidable Emergency Department Transfers for Acute Pediatric Respiratory Illness

BackgroundAcute pediatric respiratory illness is one of the most common reasons for emergency department (ED) transfer; however, few studies have examined predictors of potentially avoidable ED transfer (PAT) in this subpopulation. This study aimed to characterize patterns and predictors of PATs in children with acute respiratory illness. MethodsCross-sectional analysis of 8,402,577 visits for patients ≤17 years from 2018 to 2019 Health Care Utilization Project State ED and Inpatient Datasets from New York, Maryland, Wisconsin, and Florida. ED transfers matched to a visit at a receiving facility with a primary diagnosis of pneumonia, croup/other upper respiratory infection (URI), bronchiolitis, or asthma were included. PAT was defined as discharge from receiving ED or within 24 hours of inpatient admission without specialized procedures, as previously described. PATs were compared with necessary transfers using a 3-level generalized linear mixed model with adjustment for patient and hospital covariates. ResultsAmong 4409 matched respiratory transfers, 25.5% were potentially avoidable. Most PATs originated from EDs within the third highest quartile of annual pediatric ED visits (n = 472, 42.0%). In the multivariable model, the likelihood of PAT was higher for patients with croup/other URI ((odds ratio) OR 2.72 (2.09–3.5) and if referring ED was in the highest quartile of annual pediatric ED volumes (OR 0.48 95% (confidence interval) CI 0.26–0.88). ConclusionsPediatric respiratory transfers with a diagnosis of croup/other URI were the most likely to be potentially avoidable. Future implementation efforts to reduce PATs should consider focusing on croup management in EDs in the lower 3 quartiles of pediatric volume.

Leveraging a clinical emergency department dataset to estimate two-dose COVID-19 vaccine effectiveness and duration of protection in Canada

BackgroundDuring the COVID-19 pandemic, clinical care shifted toward virtual and Emergency Department care. We explored the feasibility of mRNA vaccine effectiveness (VE) estimation against SARS-CoV-2-related Emergency Department visits and hospitalizations using prospectively collected Emergency Department data. MethodsWe estimated two-dose VE using a test-negative design and data from 10 participating sites of the Canadian COVID-19 Emergency Department Rapid Response Network (CCEDRRN). We included Emergency Department patients presenting with COVID-19 symptoms and nucleic acid amplification testing for SARS-CoV-2 between July 19 and December 31, 2021. We excluded patients with unclear vaccination and one or more than 2 vaccine doses by their Emergency Department visit. ResultsAmong 3,405 eligible patients, adjusted two-dose mRNA VE against SARS-CoV-2-related Emergency Department visits was 93.3 % (95 % CI 87.9–96.3 %) between 7–55 days, sustained over 80 % through 139 days post-vaccination. In stratified analyses, VE was similar among patients with select immune-compromising conditions, chronic kidney disease, lung disease, unstable housing, and reported illicit substance use. ConclusionsTwo-dose mRNA VE against SARS-CoV-2-related Emergency Department visit was high and sustained, including among vulnerable subgroups. Compared to administrative datasets, active Emergency Department enrolment enables standardization for testing access and indication and supports separate VE assessment among special population subgroups. Compared to other active enrolment settings, Emergency Departments more consistently function during crises when alternate healthcare sectors become variably closed.Trial registration: Clinicaltrials.gov, NCT0470294.

Quantifying disparities in intimate partner violence: a machine learning method to correct for underreporting

The first step towards reducing the pervasive disparities in women’s health is to quantify them. Accurate estimates of the relative prevalence across groups—capturing, for example, that a condition affects Black women more frequently than white women—facilitate effective and equitable health policy that prioritizes groups who are disproportionately affected by a condition. However, it is difficult to estimate relative prevalence when a health condition is underreported, as many women’s health conditions are. In this work, we present PURPLE, a method for accurately estimating the relative prevalence of underreported health conditions which builds upon the literature in positive unlabeled learning. We show that under a commonly made assumption—that the probability of having a health condition given a set of symptoms remains constant across groups—we can recover the relative prevalence, even without restrictive assumptions commonly made in positive unlabeled learning and even if it is impossible to recover the absolute prevalence. We conduct experiments on synthetic and real health data which demonstrate PURPLE’s ability to recover the relative prevalence more accurately than do previous methods. We then use PURPLE to quantify the relative prevalence of intimate partner violence (IPV) in two large emergency department datasets. We find higher prevalences of IPV among patients who are on Medicaid, not legally married, and non-white, and among patients who live in lower-income zip codes or in metropolitan counties. We show that correcting for underreporting is important to accurately quantify these disparities and that failing to do so yields less plausible estimates. Our method is broadly applicable to underreported conditions in women’s health, as well as to gender biases beyond healthcare.

Patient-related Factors Associated with Potentially Unnecessary Transfers for Pediatric Patients with Asthma: A Retrospective Cohort Study.

Asthma is a common chronic medical condition among children and the most common diagnosis associated with interfacility transports for pediatric patients. As many as 40% of pediatric transfers may be unnecessary, resulting in potential delays in care and unnecessary costs. Our objective was to identify the patient-related factors associated with potentially unnecessary transfers for pediatric patients with asthma. We used patient care data from the California Department of Health Care Access and Information patient discharge and emergency department (ED) datasets to capture ED visits where a pediatric patient (age 2-17years) presented with asthma and was transferred to another ED or acute care hospital. The outcome of interest was a potentially unnecessary transfer, defined as a visit where length of stay after transfer was <24 hours and no advanced services were used, such as respiratory therapy or critical care. Patient-related characteristics were extracted, including age, gender, race/ethnicity, primary language, insurance status, and clinical characteristics. First, we used descriptive statistics to compare necessary vs unnecessary transfers. Second, we used generalized estimating equations accounting for clustering by ED to estimate odds ratios (OR) and identify factors associated with potentially unnecessary transfers. A total of 4,233 pediatric ED patients were transferred with a diagnosis of asthma, including 461 (11%) transfers that met criteria as potentially unnecessary. Median age was 12years (interquartile range 7-15), and 46% were female. Factors associated with increased odds of potentially unnecessary transfer while controlling for key factors included younger age (eg, 2-5years, OR 2.0, 95% confidence interval [CI] 1.4-2.9), male gender (OR 1.4, 95% CI 1.1-1.7), and Hispanic ethnicity (OR 1.6, 95% CI 1.2-2.1), while multiple hospitalizations for asthma per year was associated with decreased odds (OR 0.2, 95% CI 0.1-0.4). Several patient-related factors were associated with increased or decreased odds of potentially unnecessary transfers among pediatric patients presenting to the ED with asthma. These factors can be considered in future work to better understand, predict, and reduce unnecessary transfers and their negative consequences.

Using Primary Care and Emergency Department Datasets for Researching Violence Victimisation in the UK: A Methodological Review of Four Sources

Violence is recognised as a cause of health harm, but it is not consistently or adequately captured in healthcare data systems. While administrative health records could be valuable sources of information for measuring violence, they remain underutilised in violence-related research. The present research aims to examine the suitability of violence indicators in emergency care, primary care, and linked healthcare datasets. Descriptive analyses were conducted with the 2015/16 Hospital Episode Statistics Accident and Emergency (HES A&amp;E) and the 2021/22 Emergency Care Data Set (ECDS). The potential of the Clinical Practice Research Datalink (CPRD) and the South Wales Violence Surveillance dataset (a police and emergency department (ED) dataset linked by Public Health Wales) were shown using available evidence. Among the discussed datasets, the South Wales Violence Surveillance dataset has the most detail about violent acts and their contexts, while the CPRD includes a more extensive range of socioeconomic factors about patients and extensive linkage with other datasets. Currently, detailed safeguarding information is routinely removed from the ECDS extracts provided to researchers, limiting its utility for violence research. In the HES A&amp;E, only physical violence was consistently recorded. Addressing these issues has the potential to improve the use of health administrative data in research on violence.

Assessing the use of HL7 FHIR for implementing the FAIR guiding principles: a case study of the MIMIC-IV Emergency Department module.

To provide a real-world example on how and to what extent Health Level Seven Fast Healthcare Interoperability Resources (FHIR) implements the Findable, Accessible, Interoperable, and Reusable (FAIR) guiding principles for scientific data. Additionally, presents a list of FAIR implementation choices for supporting future FAIR implementations that use FHIR. A case study was conducted on the Medical Information Mart for Intensive Care-IV Emergency Department (MIMIC-ED) dataset, a deidentified clinical dataset converted into FHIR. The FAIRness of this dataset was assessed using a set of common FAIR assessment indicators. The FHIR distribution of MIMIC-ED, comprising an implementation guide and demo data, was more FAIR compared to the non-FHIR distribution. The FAIRness score increased from 60 to 82 out of 95 points, a relative improvement of 37%. The most notable improvements were observed in interoperability, with a score increase from 5 to 19 out of 19 points, and reusability, with a score increase from 8 to 14 out of 24 points. A total of 14 FAIR implementation choices were identified. Our work examined how and to what extent the FHIR standard contributes to FAIR data. Challenges arose from interpreting the FAIR assessment indicators. This study stands out for providing a real-world example of a dataset that was made more FAIR using FHIR. To the best of our knowledge, this is the first study that formally assessed the conformance of a FHIR dataset to the FAIR principles. FHIR improved the accessibility, interoperability, and reusability of MIMIC-ED. Future research should focus on implementing FHIR in research data infrastructures.

Improved interpretable machine learning emergency department triage tool addressing class imbalance.

The Score for Emergency Risk Prediction (SERP) is a novel mortality risk prediction score which leverages machine learning in supporting triage decisions. In its derivation study, SERP-2d, SERP-7d and SERP-30d demonstrated good predictive performance for 2-day, 7-day and 30-day mortality. However, the dataset used had significant class imbalance. This study aimed to determine if addressing class imbalance can improve SERP's performance, ultimately improving triage accuracy. The Singapore General Hospital (SGH) emergency department (ED) dataset was used, which contains 1,833,908 ED records between 2008 and 2020. Records between 2008 and 2017 were randomly split into a training set (80%) and validation set (20%). The 2019 and 2020 records were used as test sets. To address class imbalance, we used random oversampling and random undersampling in the AutoScore-Imbalance framework to develop SERP+-2d, SERP+-7d, and SERP+-30d scores. The performance of SERP+, SERP, and the commonly used triage risk scores was compared. The developed SERP+ scores had five to six variables. The AUC of SERP+ scores (0.874 to 0.905) was higher than that of the corresponding SERP scores (0.859 to 0.894) on both test sets. This superior performance was statistically significant for SERP+-7d (2019: Z = -5.843, p < 0.001, 2020: Z = -4.548, p < 0.001) and SERP+-30d (2019: Z = -3.063, p = 0.002, 2020: Z = -3.256, p = 0.001). SERP+ outperformed SERP marginally on sensitivity, specificity, balanced accuracy, and positive predictive value measures. Negative predictive value was the same for SERP+ and SERP. Additionally, SERP+ showed better performance compared to the commonly used triage risk scores. Accounting for class imbalance during training improved score performance for SERP+. Better stratification of even a small number of patients can be meaningful in the context of the ED triage. Our findings reiterate the potential of machine learning-based scores like SERP+ in supporting accurate, data-driven triage decisions at the ED.

Characterizing Malpractice Cases Involving Emergency Department Advanced Practice Providers, Physicians in Training, and Attending Physicians.

To evaluate available characteristics and financial costs of malpractice cases among Advanced Practice Providers (Nurse Practitioners and Physician Assistants), trainees (medical students, residents, fellows) and attending physicians. A retrospective analysis of claims occurring in the emergency department from January 1, 2010 to December 31, 2019 contained in the Candello database. Cases were classified according to the provider type(s) involved: Nurse Practitioner (NP), Physician Assistant (PA), trainee, or cases which did not identify an extender as being substantially involved in the adverse event that resulted in the case ("no extender"). There were 5,854 cases identified with a total gross indemnity paid of $1,007,879,346. Of these cases, 193 (3.3%) involved an NP, 513 (8.8%) involved a PA, 535 (9.1%) involved a trainee, and 4,568 (78.0%) were no extender. Cases where a trainee was involved account for the highest average gross indemnity paid whereas no extender cases are the lowest. NP and PA cases differed by contributing factors compared to no extender cases: Clinical Judgment (NP 89.1% vs no extender 76.8%, p<0.0001; PA 84.6% vs no extender, p < .0001), Documentation (NP 23.3% vs no extender 17.8%, p=0.0489; PA 25.9% vs no extender, p<0.0001) and Supervision (NP 22.3% vs no extender 1.8%, p<0.0001; PA 25.7% vs no extender p < 0.0001). Cases involving NPs and PAs had a lower percentage of high severity cases such as loss of limb or death (NP 45.6% vs no extender 50.2%, p=0.0004; PA 48.3% vs no extender, p < 0.0001). Advanced Practice Providers (APPs) and trainees comprise approximately 21% of malpractice cases and 33% of total gross indemnity paid in this large national emergency department dataset. Understanding differences in characteristics of malpractice claims that occur in emergency care settings can be used to help to mitigate provider risk.

Association Between Hospital-Ascertained Atrial Fibrillation And Central Retinal Artery Occlusion: A Study Of 12 Million Patients (P2-9.003)

<h3>Objective:</h3> To determine whether hospital-ascertained atrial fibrillation (AF) is associated with subsequent central retinal artery occlusion (CRAO). <h3>Background:</h3> Atrial fibrillation is a major risk factor for cerebral ischemic stroke. However, it is not known whether AF predicts the development of central retinal artery occlusion. <h3>Design/Methods:</h3> A retrospective cohort study was undertaken using data from the California Healthcare Cost and Utilization Project (HCUP) State Inpatient and State Emergency Department Datasets (SID/SEDDs). Patients 18 years and older discharged from non-federal hospitals between 2005 and 2011 who did not have a history of CRAO were analyzed. AF and CRAO were identified using validated ICD-9-CM diagnosis codes. Association between AF and CRAO was modeled using a Fine-Gray method with death as a competing risk with adjustment for age, biological sex, race, and vascular comorbidities. <h3>Results:</h3> A total of 12,181,778 patients were included, 806,397 with AF and 11,375,381 without AF. In total, 309 patients had CRAO. In an unadjusted analysis, there was a higher risk of CRAO in patients with versus without AF (HR 2.24 (95% CI: 1.51 to 3.32)). After adjustment for pre-specified covariates, there appeared to be a lower hazard of CRAO in patients with AF (aHR 0.61 (95% CI: 0.45 to 0.98)). Further analyses including cerebral ischemic stroke (aHR 1.16 (95% CI: 1.14 to 1.18)) and specifically embolic stroke (aHR 4.29 (95% CI 4.10–4.48)) as positive controls argued against overadjustment bias. We present sensitivity analyses including CRAO identified in any position of the ICD list, using different ascertainment windows for AF and using broader categories of retinal ischemia. <h3>Conclusions:</h3> The incidence of CRAO was higher in patients with AF than those without AF, but the hazard of CRAO was not higher for patients with AF after adjustment for measured covariates. Endpoint and exposure ascertainment may have been limited by inclusion only of inpatient and emergency department encounters. <b>Disclosure:</b> Mr. Lusk has received research support from American Heart Association. The institution of Ms. Song has received research support from Research to Prevent Blindness. Ms. Unnithan has nothing to disclose. Dr. Al-Khalidi has received personal compensation in the range of $500-$4,999 for serving on a Scientific Advisory or Data Safety Monitoring board for Medpace. Dr. Al-Khalidi has received personal compensation in the range of $500-$4,999 for serving on a Scientific Advisory or Data Safety Monitoring board for CSL Behring . The institution of Dr. Al-Khalidi has received research support from NIH/NHLBI. Jonathan Piccini has nothing to disclose. Dr. Xian has nothing to disclose. Dr. O’Brien has received personal compensation in the range of $5,000-$9,999 for serving as a Consultant for Boehringer Ingelheim. The institution of Dr. O’Brien has received research support from Pfizer. The institution of Dr. O’Brien has received research support from BMS. Dr. Mac Grory has nothing to disclose.

Abstract 30: Association Between Hospital-ascertained Atrial Fibrillation And Central Retinal Artery Occlusion: A Study Of 12 Million Patients

Introduction: Atrial fibrillation (AF) is a major risk factor for cerebral ischemic stroke. However, it is not known whether AF predicts the development of central retinal artery occlusion (CRAO), a form of ischemic stroke affecting the retina. Methods: A retrospective cohort study was undertaken using data from the California Healthcare Cost and Utilization Project (HCUP) State Inpatient and State Emergency Department Datasets (SID/SEDDs). Patients 18 years and older discharged from non-federal hospitals between 2005 and 2011 who did not have a history of CRAO were analyzed. The exposure variable was AF and the endpoint was CRAO, each identified using validated ICD-9-CM diagnosis codes. Association between AF and CRAO was modeled using a Fine-Gray method with death as a competing risk with adjustment for age, biological sex, race, and vascular co-morbidities. Results: A total of 12,181,778 patients were included, among these 806,397 with AF and 11,375,381 without AF. In total, 309 patients had CRAO. In an unadjusted analysis, there was a higher risk of CRAO in patients with versus without AF (HR 2.24 (95% CI: 1.51 to 3.32)). After adjustment for pre-specified covariates, there appeared to be a lower hazard of CRAO in patients with AF (aHR 0.61 (95% CI: 0.45 to 0.98)). Further analyses including cerebral ischemic stroke (aHR 1.16 (95% CI: 1.14 to 1.18)) and specifically embolic stroke (aHR 4.29 (95% CI 4.10-4.48)) as positive controls argued against overadjustment bias. We present sensitivity analyses including CRAO identified in any position of the discharge ICD list, using different ascertainment windows for AF and using broader categories of retinal ischemia. Conclusion: The incidence of CRAO was higher in patients with AF than those without AF, but the hazard of CRAO was not higher for patients with AF after adjustment for measured covariates. Endpoint and exposure ascertainment may have been limited by inclusion only of inpatient and emergency department encounters.

Identifying patients at risk of unplanned re-hospitalisation using statewide electronic health records

Preventing unplanned hospitalisations, including readmissions and re-presentations to the emergency department, is an important strategy for addressing the growing demand for hospital care. Significant successes have been reported from interventions put in place by hospitals to reduce their incidence. However, there is limited use of data-driven algorithms in hospital services to identify patients for enrolment into these intervention programs. Here we present the results of a study aiming to develop algorithms deployable at scale as part of a state government’s initiative to address rehospitalizations and which fills several gaps identified in the state-of-the-art literature. To the best of our knowledge, our study involves the largest-ever sample size for developing risk models. Logistic regression, random forests and gradient boosted techniques were explored as model candidates and validated retrospectively on five years of data from 27 hospitals in Queensland, Australia. The models used a range of predictor variables sourced from state-wide Emergency Department(ED), inpatient, hospital-dispensed medications and hospital-requested pathology databases. The investigation leads to several findings: (i) the advantage of looking at a longer patient data history, (ii) ED and inpatient datasets alone can provide useful information for predicting hospitalisation risk and the addition of medications and pathology test results leads to trivial performance improvements, (iii) predicting readmissions to the hospital was slightly easier than predicting re-presentations to ED after an inpatient stay, which was slightly easier again than predicting re-presentations to ED after an EDstay, (iv) a gradient boosted approach (XGBoost) was systematically the most powerful modelling approach across various tests.

Validation study of machine-learning chest radiograph software in primary and emergency medicine

To evaluate the performance of a machine learning based algorithm tool for chest radiographs (CXRs), applied to a consecutive cohort of historical clinical cases, in comparison to expert chest radiologists. The study comprised 1,960 consecutive CXR from primary care referrals and the emergency department (992 and 968 cases respectively), obtained in 2015 at a UK hospital. Two chest radiologists, each with >20 years of experience independently read all studies in consensus to serve as a reference standard. A chest artificial intelligence (AI) algorithm, Lunit INSIGHT CXR, was run on the CXRs, and results were correlated with those by the expert readers. The area under the receiver operating characteristic curve (AUROC) was calculated for the normal and 10 common findings: atelectasis, fibrosis, calcification, consolidation, lung nodules, cardiomegaly, mediastinal widening, pleural effusion, pneumothorax, and pneumoperitoneum. The ground truth annotation identified 398 primary care and 578 emergency department datasets containing pathologies. The AI algorithm showed AUROC of 0.881-0.999 in the emergency department dataset and 0.881-0.998 in the primary care dataset. The AUROC for each of the findings between the primary care and emergency department datasets did not differ, except for pleural effusion (0.954 versus 0.988, p<0.001). The AI algorithm can accurately and consistently differentiate normal from major thoracic abnormalities in both acute and non-acute settings, and can serve as a triage tool.

Multi-population generalizability of a deep learning-based chest radiograph severity score for COVID-19.

To tune and test the generalizability of a deep learning-based model for assessment of COVID-19 lung disease severity on chest radiographs (CXRs) from different patient populations.A published convolutional Siamese neural network-based model previously trained on hospitalized patients with COVID-19 was tuned using 250 outpatient CXRs. This model produces a quantitative measure of COVID-19 lung disease severity (pulmonary x-ray severity (PXS) score). The model was evaluated on CXRs from 4 test sets, including 3 from the United States (patients hospitalized at an academic medical center (N = 154), patients hospitalized at a community hospital (N = 113), and outpatients (N = 108)) and 1 from Brazil (patients at an academic medical center emergency department (N = 303)). Radiologists from both countries independently assigned reference standard CXR severity scores, which were correlated with the PXS scores as a measure of model performance (Pearson R). The Uniform Manifold Approximation and Projection (UMAP) technique was used to visualize the neural network results.Tuning the deep learning model with outpatient data showed high model performance in 2 United States hospitalized patient datasets (R = 0.88 and R = 0.90, compared to baseline R = 0.86). Model performance was similar, though slightly lower, when tested on the United States outpatient and Brazil emergency department datasets (R = 0.86 and R = 0.85, respectively). UMAP showed that the model learned disease severity information that generalized across test sets.A deep learning model that extracts a COVID-19 severity score on CXRs showed generalizable performance across multiple populations from 2 continents, including outpatients and hospitalized patients.

Physicians prescribe fewer analgesics during night shifts than day shifts

Adequate pain management is one of the biggest challenges of the modern healthcare system. Physician perception of patient subjective pain, which is crucial to pain management, is susceptible to a host of potential biases. Here we explore the timing of physicians' work as a previously unrecognized source of systematic bias in pain management. We hypothesized that during night shifts, sleep deprivation, fatigue, and stress would reduce physicians' empathy for others' pain, leading to underprescription of analgesics for patient pain relief. In study 1, 67 resident physicians, either following a night shift or not, performed empathy for pain assessment tasks and simulated patient scenarios in laboratory conditions. As predicted, following a night shift, physicians showed reduced empathy for pain. In study 2, we explored this phenomenon in medical decisions in the field. We analyzed three emergency department datasets from Israel and the United States that included discharge notes of patients arriving with pain complaints during 2013 to 2020 (n = 13,482). Across all datasets, physicians were less likely to prescribe an analgesic during night shifts (compared to daytime shifts) and prescribed fewer analgesics than generally recommended by the World Health Organization. This effect remained significant after adjusting for patient, physician, type of complaint, and emergency department characteristics. Underprescription for pain during night shifts was particularly prominent for opioids. We conclude that night shift work is an important and previously unrecognized source of bias in pain management, likely stemming from impaired perception of pain. We consider the implications for hospitals and other organizations employing night shifts.

Impact of the COVID-19 Pandemic on the Pediatric Hospital Visits: Evidence from the State of Florida.

Although early evidence reported a substantial decline in pediatric hospital visits during COVID-19, it is unclear whether the decline varied across different counties, particularly in designated Medically Underserved Areas (MUA). The objective of this study is to explore the state-wide impact of COVID-19 on pediatric hospital visit patterns, including the economic burden and MUA communities. We conducted a retrospective observational study of pediatric hospital visits using the Florida State all-payer Emergency Department (ED) and Inpatient dataset during the pandemic (April–September 2020) and the same period in 2019. Pediatric Treat-and-Release ED and inpatient visit rates were compared by patient demographics, socioeconomic, diagnosis, MUA status, and hospital characteristics. Pediatric hospital visits in Florida decreased by 53.7% (62.3% in April–June, 44.2% in July–September) during the pandemic. The Treat-and-Release ED and inpatient visits varied up to 5- and 3-fold, respectively, across counties. However, changes in hospital visits across MUA counties were similar compared with non-MUA counties except for lower Treat-and-Release ED volume in April–May. The disproportional decrease in visits was notable for the underserved population, including Hispanic and African American children; Medicaid coverages; non-children’s hospitals; and diagnosed with respiratory diseases, appendicitis, and sickle-cell. Florida Hospitals experienced a USD 1.37 billion (average USD 8.3 million) decline in charges across the study period in 2020. Disproportionate decrease in hospital visits, particularly in the underserved population, suggest a combined effect of the persistent challenge of care access and changes in healthcare-seeking behavior during the pandemic. These findings suggest that providers and policymakers should emphasize alternative interventions/programs ensuring adequate care during the pandemic, particularly for high-risk children.

Association of Emergency Department Payer Mix with ED Receipt of Telehealth Services: An Observational Analysis.

IntroductionTelehealth is commonly used to connect emergency department (ED) patients with specialists or resources required for their care. Its infrastructure requires substantial upfront and ongoing investment from an ED or hospital and may be more difficult to implement in lower-resourced settings. Our aim was to examine for an association between ED payer mix and receipt of telehealth services.MethodsUsing data from the National Emergency Department Inventory (NEDI)-USA 2016 survey, we categorized EDs based on receipt of telehealth services (yes/no). The NEDI-USA data for EDs in New York state was linked with data from state ED datasets (SEDD) and state inpatient data (SID) to determine EDs’ payer mix (percent self-pay or Medicaid). Other ED characteristics of interest were rural location, academic status, and annual ED visit volume. We compared EDs with and without telehealth receipt, and used a logistic regression model to examine the relationship between ED payer mix and telehealth receipt after accounting for other ED characteristics.ResultsOf the 162 New York EDs in the SEDD-SID dataset, 160 (99%) were linked to the NEDI-USA dataset and 133 of those responded (83%) to the survey. Telehealth receipt was reported by 48 EDs (36%, 95% confidence interval [CI], 28–44%). Emergency departments with and without telehealth receipt were similar (all P >0.40) with respect to rurality (6% vs 9%, respectively), academic status (13% vs 8%), and annual volume (median 36,728 vs 43,000). By contrast, median percent of Medicaid or self-pay patients was lower in telehealth EDs (36%) vs non-telehealth EDs (45%, P = 0.02). In adjusted analysis, increasing proportion of Medicaid and self-pay patients was associated with decreased odds of telehealth receipt (odds ratio 0.87 per 5% increase; 95% CI, 0.77–0.99). Rural location, academic status, and ED volume were not significantly associated with odds of ED telehealth receipt in the adjusted model.ConclusionAmong EDs in the state of New York, increasing proportion of self-pay and Medicaid patients was associated with decreased odds of ED telehealth receipt, even after accounting for rural location, academic status, and ED volume. The findings support the need for additional infrastructural investment in EDs serving a greater proportion of disadvantaged patients to ensure equitable access.

Variation in Seeking Care for Cardiovascular Disease and Ambulance Utilization among Migrants in Australia: Time, Ethnicity, and Delay (TED) Study III.

Insight into differences in seeking medical care for chest pain among migrant populations is limited. This study aimed to determine ethnic differences in seeking care behaviors and using ambulances among migrants compared to an Australian-born group. A total of 607 patients presenting with chest pain to a tertiary hospital between 1 July 2012 and 30 June 2014 were randomly selected. Data from the emergency department dataset and medical record reviews were collected and linked for analysis. The migrant group was stratified into nine ethnic groups for analysis based on the Australian Standard Classification of Cultural and Ethnic Groups. The overall median prehospital delay time was 3.7 (1.5, 10.7) h, which ranged from 2.5 (1.0, 10.7) (Southern and Eastern European group) to 6.0 (2.3, 20.6) (Sub-Saharan African group). The median decision time was 2.0 (0.8, 7.9) h, which ranged from 1.5 (Australian-born group) to 4.5 h (Sub-Saharan African group). Five ethnic groups had significantly longer decision times compared to the Australian-born group. Decision time accounted for 58.4% of pre-hospital delay time. Migrant patients were 60% less likely to seek care for chest pain within one hour (odds ratio 0.40, (0.23–0.68), p = 0.001). There was no significant difference in ambulance utilization between migrant and Australian-born groups. In conclusion, ethnic differences in seeking care for chest pain do exist, and ethnicity plays a vital role in a longer delay in seeking care. To reduce the delays and improve patient outcomes, appropriate health campaigns focusing on ethnic differences among migrant populations and normalizing cultural competency into practice are recommended.

California Children Presenting to an Emergency Department for Mental Health Emergencies: Trajectories of Care.

Pediatric emergency department (ED) mental health visits are increasing in the United States. At the same time, child/adolescent psychiatric services are limited. This study examines the trajectory of pediatric patients presenting with mental health emergencies to better understand availability of specialty care resources in regional networks. This retrospective cohort study used a California Office of Statewide Health Planning and Development linked ED and Inpatient Discharge Dataset (2005-2015) to study pediatric patients (5-17 years) who presented to an ED with a primary mental health diagnosis. Outcomes were disposition: discharge, admission, or transfer.Patients transferred were further analyzed for disposition. Regression models to identify characteristics associated with disposition were created. There were 384,339 pediatric patients presented for a primary mental health emergency from 2005 to 2015; 287,997 were discharged, 17,564 were admitted, and 78,725 were transferred. Among those not discharged, patients with public (odds ratio [OR], 1.28; P < 0.01) or self-pay insurance (OR, 5.64; P < 0.01), Black (OR, 2.15; P < 0.01), or Native American race (OR, 2.32; P < 0.01), and who presented to rural EDs (OR, 3.10; P < 0.01), nonteaching hospitals (OR, 3.06; P < 0.01), or hospitals in counties without dedicated child/adolescent psychiatric beds (OR, 5.59; P < 0.01) had higher odds of transfer.Among those not discharged from the second hospital, Black patients (OR, 2.47; P < 0.03) and those who were transferred to a teaching hospital (OR, 1.9; P < 0.01) had higher odds of second transfer. Pediatric patients with mental health emergencies experience different trajectories of care. Transfer protocols and regionalized networks may help streamline services and decrease inefficiencies in care.