RESUMO
BACKGROUND: Limited capacity in the emergency department (ED) secondary to boarding and crowding has resulted in patients receiving care in hallways to provide access to timely evaluation and treatment. However, there are concerns raised by physicians and patients regarding a decrease in patient centered care and quality resulting from hallway care. We sought to explore social risk factors associated with hallway placement and operational outcomes. STUDY DESIGN/METHODS: Observational study between July 2017 and February 2020. Primary outcome was the adjusted odds ratio (aOR) of patient placement in a hallway treatment space adjusting for patient demographics and ED operational factors. Secondary outcomes included left without being seen (LWBS), discharge against medical advice (AMA), elopement, 72-h ED revisit, 10-day ED revisit and escalation of care during boarding. RESULTS: Among 361,377 ED visits, 100,079 (27.7%) visits were assigned to hallway beds. Patient insurance coverage (Medicaid (aOR 1.04, 95% CI 1.01,1.06) and Self-pay/Other (1.08, (1.03, 1.13))) with comparison to private insurance, and patient sex (Male (1.08, (1.06, 1.10))) with comparison to female sex are associated with higher odds of hallway placement but patient age, race, and language were not. These associations are adjusted for ED census, triage assigned severity, ED staffing, boarding level, and time effect, with social factors mutually adjusted. Additionally adjusting for patients' social factors, patients placed in hallways had higher odds of elopement (1.23 (1.07,1.41)), 72-h ED revisit (1.33 (1.08, 1.64)) and 10-day ED revisit (1.23 (1.11, 1.36)) comparing with patients placed in regular ED rooms. We did not find statistically significant associations between hallway placement and LWBS, discharge AMA, or escalation of care. CONCLUSION: While hallway usage is ad hoc, we find consistent differences in care delivery with those insured by Medicaid and self-pay or male sex being placed in hallway beds. Further work should examine how new front-end processes such as provider in triage or split flow may be associated with inequities in patient access to emergency and hospital care.
Assuntos
Serviço Hospitalar de Emergência , Pacientes , Estados Unidos , Humanos , Masculino , Feminino , Admissão do Paciente , Triagem , Alta do Paciente , Estudos RetrospectivosRESUMO
BACKGROUND: Discharge instructions are a key form of documentation and patient communication in the time of transition from the emergency department (ED) to home. Discharge instructions are time-consuming and often underprioritized, especially in the ED, leading to discharge delays and possibly impersonal patient instructions. Generative artificial intelligence and large language models (LLMs) offer promising methods of creating high-quality and personalized discharge instructions; however, there exists a gap in understanding patient perspectives of LLM-generated discharge instructions. OBJECTIVE: We aimed to assess the use of LLMs such as ChatGPT in synthesizing accurate and patient-accessible discharge instructions in the ED. METHODS: We synthesized 5 unique, fictional ED encounters to emulate real ED encounters that included a diverse set of clinician history, physical notes, and nursing notes. These were passed to GPT-4 in Azure OpenAI Service (Microsoft) to generate LLM-generated discharge instructions. Standard discharge instructions were also generated for each of the 5 unique ED encounters. All GPT-generated and standard discharge instructions were then formatted into standardized after-visit summary documents. These after-visit summaries containing either GPT-generated or standard discharge instructions were randomly and blindly administered to Amazon MTurk respondents representing patient populations through Amazon MTurk Survey Distribution. Discharge instructions were assessed based on metrics of interpretability of significance, understandability, and satisfaction. RESULTS: Our findings revealed that survey respondents' perspectives regarding GPT-generated and standard discharge instructions were significantly (P=.01) more favorable toward GPT-generated return precautions, and all other sections were considered noninferior to standard discharge instructions. Of the 156 survey respondents, GPT-generated discharge instructions were assigned favorable ratings, "agree" and "strongly agree," more frequently along the metric of interpretability of significance in discharge instruction subsections regarding diagnosis, procedures, treatment, post-ED medications or any changes to medications, and return precautions. Survey respondents found GPT-generated instructions to be more understandable when rating procedures, treatment, post-ED medications or medication changes, post-ED follow-up, and return precautions. Satisfaction with GPT-generated discharge instruction subsections was the most favorable in procedures, treatment, post-ED medications or medication changes, and return precautions. Wilcoxon rank-sum test of Likert responses revealed significant differences (P=.01) in the interpretability of significant return precautions in GPT-generated discharge instructions compared to standard discharge instructions but not for other evaluation metrics and discharge instruction subsections. CONCLUSIONS: This study demonstrates the potential for LLMs such as ChatGPT to act as a method of augmenting current documentation workflows in the ED to reduce the documentation burden of physicians. The ability of LLMs to provide tailored instructions for patients by improving readability and making instructions more applicable to patients could improve upon the methods of communication that currently exist.
Assuntos
Serviço Hospitalar de Emergência , Alta do Paciente , Humanos , Serviço Hospitalar de Emergência/estatística & dados numéricos , Alta do Paciente/estatística & dados numéricos , Feminino , Masculino , Inquéritos e Questionários , Adulto , Pessoa de Meia-Idade , Inteligência ArtificialRESUMO
STUDY OBJECTIVE: We evaluate the impact of the COVID-19 pandemic on care for survivors of sexual assault in three urban Emergency Departments (ED) in the United States. METHODS: A retrospective chart review was conducted on patients who presented after sexual assault to three EDs during 6-month intervals before and during the COVID-19 pandemic. We excluded individuals <18 years old. We performed a structured chart review to ascertain demographics, ED treatments, and adherence to guidelines for care of sexual assault survivors. RESULTS: Of 105 patients who received care after a sexual assault, 57 presented during the COVID-19 pandemic. The majority were female, White/Caucasian, and presented within 120 h of sexual assault. There was an increase in ED presentations for sexual assault during the pandemic. While there was no difference in medical care, there were fewer sexual assault advocates called during the pandemic. In addition, there was an increase in non-White survivors in the first 3 months of the pandemic that did not remain at 6 months. CONCLUSION: The care of survivors in the ED was disrupted by the COVID-19 pandemic. While medical care remained similar, fewer calls to sexual assault advocates, a key component of ED and long-term care of survivors, demonstrate a disruption in their care.
Assuntos
COVID-19 , Delitos Sexuais , Humanos , Masculino , Estados Unidos , Feminino , Adolescente , Pandemias , Connecticut/epidemiologia , Estudos Retrospectivos , COVID-19/epidemiologia , Serviço Hospitalar de Emergência , SobreviventesRESUMO
The increasing complexity of ED physician performance measures has resulted in significant challenges, including duplicative and conflicting measures that fail to account for different ED settings. We performed a cross sectional analysis of correlations between measures to characterize their relationships and determine if differences exist between academic versus non-academic ED settings. Pearson correlations were calculated for 12 measures among 220 ED physicians at 11 EDs. Higher admission rate was strongly correlated with higher CT utilization rate (R = 0.7, p < 0.01) and longer room to discharge time (R = 0.7, p < 0.01). Higher patients per hour was strongly correlated with shorter room to doctor time (R = -0.7, p < 0.01). Stronger measure correlations were found in the academic setting compared to the non-academic setting. Strong correlations between ED measures imply opportunities to reduce competing performance demands on clinicians. Differences in correlations at academic versus non-academic settings suggest that it may be inappropriate to apply the same performance standards across settings.
Assuntos
Medicina de Emergência , Médicos , Humanos , Serviço Hospitalar de Emergência , Estudos TransversaisRESUMO
BACKGROUND: In March 2020, the U.S. Department of Health and Human Services Office for Civil Rights stated that they would use discretion when enforcing the Health Insurance Portability and Accountability Act regarding remote communication technologies that promoted telehealth delivery during the COVID-19 pandemic. This was in an effort to protect patients, clinicians, and staff. More recently, smart speakers-voice-activated, hands-free devices-are being proposed as productivity tools within hospitals. OBJECTIVE: We aimed to characterize the novel use of smart speakers in the emergency department (ED). METHODS: A retrospective observational study of Amazon Echo Show® utilization from May 2020 to October 2020 in a large academic Northeast health system ED. Voice commands and queries were classified as either patient care-related or non-patient care-related, and then further subcategorized to explore the content of given commands. RESULTS: Of 1232 commands analyzed, 200 (16.23%) were determined to be patient care-related. Of these commands, 155 (77.5%) were clinical in nature (i.e., "drop in on triage") and 23 (11.5%) were environment-enhancing commands (i.e., "play calming sounds"). Among non-patient care-related commands, 644 (62.4%) were for entertainment. Among all commands, 804 (65.3%) were during night-shift hours, which was statistically significant (p < 0.001). CONCLUSIONS: Smart speakers showed notable engagement, primarily being used for patient communication and entertainment. Future studies should examine content of patient care conversations using these devices, effects on frontline staff wellbeing, productivity, patient satisfaction, and even explore opportunities for "smart" hospital rooms.
Assuntos
COVID-19 , Telemedicina , Humanos , Pandemias , Serviço Hospitalar de Emergência , Estudos RetrospectivosRESUMO
STUDY OBJECTIVE: Our institution experienced a change in SARS-CoV-2 testing policy as well as substantial changes in local COVID-19 prevalence, allowing for a unique examination of the relationship between SARS-CoV-2 testing and emergency department (ED) length of stay. METHODS: This was an observational interrupted time series of all patients admitted to an academic health system between March 15, 2020, and September 30, 2020. Given testing limitations from March 15 to April 24, all patients receiving SARS-CoV-2 tests were symptomatic. On April 24, testing was expanded to all ED admissions. The primary and secondary outcomes were ED length of stay and number needed to test to obtain a positive, respectively. RESULTS: A total of 70,856 patients were cared for in the EDs during the 7-month period. The testing change increased admission length of stay by 1.89 hours (95% confidence interval 1.39 to 2.38). The number needed to test was 2.5 patients and was highest yield on April 1, 2020, when the state positivity rate was 39.7%; however, the number needed to test exceeded 170 patients by Sept 1, 2020, at which point the state positivity rate was 0.5%. CONCLUSION: Although universal SARS-CoV-2 testing of ED admissions may meaningfully support mitigation and containment efforts, the clinical cost of testing all admissions amid low community positivity is notable. In our system, universal ED SARS-CoV-2 testing was associated with a 24% increase in admission length of stay alongside the detection of only 1 positive case every other day. Given the known harms and risks of ED boarding and crowding, solutions must be developed to support regular operational flow while balancing infection prevention needs.
Assuntos
Teste para COVID-19/métodos , COVID-19/diagnóstico , Serviço Hospitalar de Emergência/estatística & dados numéricos , Tempo de Internação/estatística & dados numéricos , COVID-19/epidemiologia , Humanos , Pandemias , SARS-CoV-2 , Estados Unidos/epidemiologiaRESUMO
STUDY OBJECTIVE: There is a continued movement toward health data transparency, accelerated by the 21st Century CURES Act, which mandated the automatic and immediate release of clinical notes, often termed "open notes." Differences in utilization among different patient demographics and disproportionately affected populations within the emergency department (ED) are not yet known. METHODS: This was an observational study of 10 EDs and 3 urgent care centers across a single health system over a 13-week period from February 1, 2021 to May 2, 2021. Primary outcomes included the proportion of patients with patient portal access to open notes at the time of encounter, the proportion of patients with access who opened the clinical note, and time from clinical note signing to patient read. RESULTS: Among 98,725 patient visits, less than half (48.9%) had patient portal access, of which 13.7% read an open note. Access was less likely in patients who were under age 18 (odds ratio 0.10, 95% confidence interval 0.08 to 0.11), older than 65 (0.82, 0.73 to 0.93), Black non-Hispanic (0.66, 0.61 to 0.73), non-English speakers, and on public insurance. Patients were less likely to read open notes if they identified as Black non-Hispanic (0.61, 0.57 to 0.66), spoke Spanish (0.70, 0.60 to 0.81), or were on public insurance. CONCLUSION: We identified substantial differences in digital access to clinical notes as well as patient utilization of open notes in a large, diverse sample. Health transparency initiatives must address not only technology adoption broadly but also the unique barriers faced by populations experiencing disadvantage to facilitate equitable access to and awareness about digital health tools without the unintended consequence of expanding disparities.
Assuntos
Acesso à Informação , Tomada de Decisão Clínica , Registros Eletrônicos de Saúde , Serviço Hospitalar de Emergência , Política de Saúde , Registros de Saúde Pessoal , Adolescente , Adulto , Idoso , Feminino , Humanos , Masculino , Pessoa de Meia-Idade , Adulto JovemRESUMO
BACKGROUND: Emergency Department (ED) boarding, the practice of holding patients in the ED after they have been admitted to the hospital due to unavailability of inpatient beds, is common and contributes to the public health crisis of ED crowding. Prior work has documented the harms of ED boarding on access and quality of care. Limited studies examine the relationship between ED boarding and an equally important domain of quality-the cost of care. This study evaluates the relationship between ED boarding, ED characteristics and risk-adjusted hospitalization costs utilizing national publicly-reported measures. METHODS: We conducted a cross-sectional analysis of two 2018 Centers for Medicare and Medicaid Services (CMS) Hospital Compare datasets: 1) Medicare Hospital Spending per Patient and 2) Timely and Effective Care. We constructed a hospital-level multivariate linear regression analysis to examine the association between ED boarding and Medicare spending per beneficiary (MSPB), adjusting for ED length of stay, door to diagnostic evaluation time, and ED patient volume. RESULTS: A total of 2903 hospitals were included in the analysis. ED boarding was significantly correlated with MSPB (r = 0.1774; p-value: < 0.0001). In multivariate regression, ED boarding was also positively associated with MSPB (Beta: 0.00015; p < 0.0001) after adjustment for other hospital level crowding indicators. CONCLUSION: We found a strong relationship between measures of ED crowding, including ED boarding, and risk-adjusted hospital spending. Future work should elucidate the mediators of this relationship. Policymakers and administrators should consider the financial harms of ED boarding when devising strategies to improve hospital care access and flow.
Assuntos
Serviço Hospitalar de Emergência/economia , Hospitalização/economia , Tempo de Internação/economia , Listas de Espera , Idoso , Estudos Transversais , Feminino , Humanos , Masculino , Medicare/economia , Estados UnidosRESUMO
OBJECTIVE: Although timely administration of antibiotics has an established benefit in serious bacterial infection, the majority of studies evaluating antibiotic delay focus only on the first dose. Recent evidence suggests that delays in redosing may also be associated with worse clinical outcome. In light of the increasing burden of boarding in Emergency Departments (ED) and subsequent need to redose antibiotic in the ED, we examined the association between delayed second antibiotic dose administration and mortality among patients admitted from the ED with a broad array of infections and characterized risk factors associated with delayed second dose administration. METHODS: We performed a retrospective cohort study of patients admitted through five EDs in a single healthcare system from 1/2018 through 12/2018. Our study included all patients, aged 18 years or older, who received two intravenous antibiotic doses within a 30-h period, with the first dose administered in the ED. Patients with end stage renal disease, cirrhosis and extremes of weight were excluded due to a lack of consensus on antibiotic dosing intervals for these populations. Delay was defined as administration of the second dose at a time-point greater than 125% of the recommended interval. The primary outcome was in-hospital mortality. RESULTS: A total of 5605 second antibiotic doses, occurring during 4904 visits, met study criteria. Delayed administration of the second dose occurred during 21.1% of visits. After adjustment for patient characteristics, delayed second dose administration was associated with increased odds of in-hospital mortality (OR 1.50, 95%CI 1.05-2.13). Regarding risk factors for delay, every one-hour increase in allowable compliance time was associated with a 18% decrease in odds of delay (OR 0.82 95%CI 0.75-0.88). Other risk factors for delay included ED boarding more than 4 h (OR 1.47, 95%CI 1.27-1.71) or a high acuity presentation as defined by emergency severity index (ESI) (OR 1.54, 95%CI 1.30-1.81 for ESI 1-2 versus 3-5). CONCLUSIONS: Delays in second antibiotic dose administration were frequent in the ED and early hospital course, and were associated with increased odds of in-hospital mortality. Several risk factors associated with delays in second dose administration, including ED boarding, were identified.
Assuntos
Antibacterianos/administração & dosagem , Infecções Bacterianas/tratamento farmacológico , Administração Intravenosa , Antibacterianos/uso terapêutico , Infecções Bacterianas/mortalidade , Esquema de Medicação , Serviço Hospitalar de Emergência/estatística & dados numéricos , Feminino , Mortalidade Hospitalar , Humanos , Masculino , Pessoa de Meia-Idade , Estudos Retrospectivos , Fatores de Risco , Fatores de TempoRESUMO
The COVID-19 pandemic has led to rapid changes in community and healthcare delivery policies creating new and unique challenges to managing ED pandemic response efforts. One example is the practice of social distancing in the workplace as an internationally recommended non-pharmaceutical intervention to reduce transmission. While attention has been focused on public health measures, healthcare workers cannot overlook the transmission risk they present to their colleagues and patients. Our network of three EDs are all high traffic areas for both patients and staff, which makes the limitation of close person-to-person contact particularly difficult to achieve. To design, implement and communicate contact reduction changes in the ED workplace, our COVID-19 task force formalised a set of multidisciplinary recommendations that enumerated concrete ways to reduce healthcare worker transmission to coworkers and to patients from ED patient arrival to discharge. We also addressed staff-to-staff contact reduction strategies when not performing direct patient care. We describe our conceptual approach and successful implementation of workplace distancing.
Assuntos
Infecções por Coronavirus , Transmissão de Doença Infecciosa/prevenção & controle , Serviço Hospitalar de Emergência/organização & administração , Controle de Infecções , Relações Interpessoais , Pandemias , Pneumonia Viral , Local de Trabalho/organização & administração , Betacoronavirus , COVID-19 , Infecções por Coronavirus/epidemiologia , Infecções por Coronavirus/prevenção & controle , Infecções por Coronavirus/terapia , Atenção à Saúde/métodos , Atenção à Saúde/tendências , Humanos , Controle de Infecções/métodos , Controle de Infecções/organização & administração , Comunicação Interdisciplinar , Inovação Organizacional , Pandemias/prevenção & controle , Pneumonia Viral/epidemiologia , Pneumonia Viral/prevenção & controle , Pneumonia Viral/terapia , Formulação de Políticas , SARS-CoV-2 , Estados UnidosRESUMO
INTRODUCTION: Nontraumatic intracranial hemorrhage (ICH) is a neurological emergency of research interest; however, unlike ischemic stroke, has not been well studied in large datasets due to the lack of an established administrative claims-based definition. We aimed to evaluate both explicit diagnosis codes and machine learning methods to create a claims-based definition for this clinical phenotype. METHODS: We examined all patients admitted to our tertiary medical center with a primary or secondary International Classification of Disease version 9 (ICD-9) or 10 (ICD-10) code for ICH in claims from any portion of the hospitalization in 2014-2015. As a gold standard, we defined the nontraumatic ICH phenotype based on manual chart review. We tested explicit definitions based on ICD-9 and ICD-10 that had been previously published in the literature as well as four machine learning classifiers including support vector machine (SVM), logistic regression with LASSO, random forest and xgboost. We report five standard measures of model performance for each approach. RESULTS: A total of 1830 patients with 2145 unique ICD-10 codes were included in the initial dataset, of which 437 (24%) were true positive based on manual review. The explicit ICD-10 definition performed best (Sensitivityâ¯=â¯0.89 (95% CI 0.85-0.92), Specificityâ¯=â¯0.83 (0.81-0.85), F-scoreâ¯=â¯0.73 (0.69-0.77)) and improves on an explicit ICD-9 definition (Sensitivityâ¯=â¯0.87 (0.83-0.90), Specificityâ¯=â¯0.77 (0.74-0.79), F-scoreâ¯=â¯0.67 (0.63-0.71). Among machine learning classifiers, SVM performed best (Sensitivityâ¯=â¯0.78 (0.75-0.82), Specificityâ¯=â¯0.84 (0.81-0.87), AUCâ¯=â¯0.89 (0.87-0.92), F-scoreâ¯=â¯0.66 (0.62-0.69)). CONCLUSIONS: An explicit ICD-10 definition can be used to accurately identify patients with a nontraumatic ICH phenotype with substantially better performance than ICD-9. An explicit ICD-10 based definition is easier to implement and quantitatively not appreciably improved with the additional application of machine learning classifiers. Future research utilizing large datasets should utilize this definition to address important research gaps.
Assuntos
Demandas Administrativas em Assistência à Saúde , Mineração de Dados , Classificação Internacional de Doenças , Hemorragias Intracranianas/diagnóstico , Máquina de Vetores de Suporte , Idoso , Idoso de 80 Anos ou mais , Feminino , Pesquisa sobre Serviços de Saúde , Humanos , Hemorragias Intracranianas/classificação , Masculino , Pessoa de Meia-Idade , Fenótipo , Valor Preditivo dos Testes , Reprodutibilidade dos TestesAssuntos
Serviço Hospitalar de Emergência/organização & administração , Serviço Hospitalar de Enfermagem/normas , Satisfação do Paciente/estatística & dados numéricos , Adolescente , Adulto , Idoso , Idoso de 80 Anos ou mais , Connecticut , Feminino , Humanos , Masculino , Pessoa de Meia-Idade , Recursos Humanos de Enfermagem Hospitalar/normas , Estudos Prospectivos , Inquéritos e Questionários , Adulto JovemAssuntos
Comunicação , Serviço Hospitalar de Emergência , Tempo de Internação , Satisfação do Paciente , Relações Profissional-Paciente , Tempo para o Tratamento , Adulto , Negro ou Afro-Americano , Feminino , Humanos , Pesquisa Qualitativa , Fatores de Tempo , Triagem , População Branca , Adulto JovemRESUMO
BACKGROUND: The Geriatric Emergency Medicine Specialist (GEMS) pilot program is an innovative approach that utilizes geriatric-trained advanced practice providers to facilitate geriatric assessments and care planning for older adults in the emergency department (ED). The objective of this study was to explore the effect of GEMS on the use of observation status and final ED disposition. METHODS: This was a retrospective study under a target trial emulation framework. Geriatric patients (65+ years old) who presented to two ED sites within a large regional healthcare system between December 2020 and December 2022 were included. The primary outcome was final ED disposition (discharge, hospital inpatient admission, or hospital observation admission). Secondary outcomes included ED observation and ED length of stay. Non-GEMS patients were propensity score matched 5:1 to GEMS patients. Doubly robust regression was used to estimate the odds ratios and 95% confidence intervals of inpatient admission, discharge, hospital observation admission, ED observation admission, and estimate the mean ED length of stay. RESULTS: A total of 427 of 43,064 total patients (1.0%) received a GEMS intervention during the study period. Our analysis included 2,302 geriatric ED patients (410 GEMS, 1,892 non-GEMS) after propensity score matching. Hospital admission rates were 34.1% for GEMS compared to 56.4% for conventional treatment. GEMS patients had decreased odds of inpatient admission (OR: 0.41, 95 CI: 0.34-0.51, p < 0.001), increased odds of discharge (OR: 1.19 95 CI: 1.00-1.42, p = 0.047), hospital observation admission (OR: 2.97, 95 CI: 2.35-3.75, p < 0.001), ED observation admission (OR: 4.84 95 CI: 3.67-6.38, p < 0.001), and had a longer average ED length of stay (170 min, 95 CI: 84.6-256, p < 0.001) compared to non-GEMS patients. CONCLUSIONS: Patients seen by GEMS during their ED visit were associated with higher rates of hospital discharge and lower rates of hospital admissions.
Assuntos
Serviço Hospitalar de Emergência , Avaliação Geriátrica , Tempo de Internação , Alta do Paciente , Humanos , Serviço Hospitalar de Emergência/estatística & dados numéricos , Idoso , Feminino , Masculino , Estudos Retrospectivos , Alta do Paciente/estatística & dados numéricos , Tempo de Internação/estatística & dados numéricos , Idoso de 80 Anos ou mais , Avaliação Geriátrica/métodos , Geriatria , Pontuação de Propensão , Medicina de Emergência , Hospitalização/estatística & dados numéricos , Projetos Piloto , Admissão do Paciente/estatística & dados numéricosRESUMO
Objectives: This study presents a design framework to enhance the accuracy by which large language models (LLMs), like ChatGPT can extract insights from clinical notes. We highlight this framework via prompt refinement for the automated determination of HEART (History, ECG, Age, Risk factors, Troponin risk algorithm) scores in chest pain evaluation. Methods: We developed a pipeline for LLM prompt testing, employing stochastic repeat testing and quantifying response errors relative to physician assessment. We evaluated the pipeline for automated HEART score determination across a limited set of 24 synthetic clinical notes representing four simulated patients. To assess whether iterative prompt design could improve the LLMs' ability to extract complex clinical concepts and apply rule-based logic to translate them to HEART subscores, we monitored diagnostic performance during prompt iteration. Results: Validation included three iterative rounds of prompt improvement for three HEART subscores with 25 repeat trials totaling 1200 queries each for GPT-3.5 and GPT-4. For both LLM models, from initial to final prompt design, there was a decrease in the rate of responses with erroneous, non-numerical subscore answers. Accuracy of numerical responses for HEART subscores (discrete 0-2 point scale) improved for GPT-4 from the initial to final prompt iteration, decreasing from a mean error of 0.16-0.10 (95% confidence interval: 0.07-0.14) points. Conclusion: We established a framework for iterative prompt design in the clinical space. Although the results indicate potential for integrating LLMs in structured clinical note analysis, translation to real, large-scale clinical data with appropriate data privacy safeguards is needed.
Assuntos
Acidentes de Trânsito , Aorta/lesões , Tórax Fundido/complicações , Adulto , Tórax Fundido/etiologia , Humanos , Masculino , Veículos Automotores , Fraturas das Costelas/complicações , Fraturas das Costelas/diagnóstico por imagem , Cintos de Segurança , Tomografia Computadorizada por Raios X/métodosRESUMO
BACKGROUND: Despite improving supplies, SARS-CoV-2 nucleic acid amplification tests remain limited during surges and more so given concerns around COVID-19/influenza co-occurrence. Matching clinical guidelines to available supplies ensures resources remain available to meet clinical needs. We report a change in clinician practice after an electronic health record (EHR) order redesign to impact emergency department (ED) testing patterns. METHODS: We included all ED visits between December 1, 2021 and January 18, 2022 across a hospital system to assess the impact of EHR order changes on provider behavior 3 weeks before and after the change. The EHR order redesign included embedded symptom-based order guidance. Primary outcomes were the proportion of COVID-19 + flu/respiratory syncytial virus (RSV) testing performed on symptomatic, admitted, and discharged patients, and the proportion of COVID-19 + flu testing on symptomatic, discharged patients. RESULTS: A total of 52 215 ED visits were included. For symptomatic, discharged patients, COVID-19 + flu/RSV testing decreased from 11.4 to 5.8 tests per 100 symptomatic visits, and the rate of COVID-19 + flu testing increased from 7.4 to 19.1 before and after the intervention, respectively. The rate of COVID-19 + flu/RSV testing increased from 5.7 to 13.1 tests per 100 symptomatic visits for symptomatic patients admitted to the hospital. All changes were significant (P < 0.0001). CONCLUSIONS: A simple EHR order redesign was associated with increased adherence to institutional guidelines for SARS-CoV-2 and influenza testing amidst supply chain limitations necessitating optimal allocation of scarce testing resources. With continually shifting resource availability, clinician education is not sufficient. Rather, system-based interventions embedded within exiting workflows can better align resources and serve testing needs of the community.
Assuntos
COVID-19 , Influenza Humana , Humanos , COVID-19/diagnóstico , SARS-CoV-2 , Hospitalização , Teste para COVID-19RESUMO
BACKGROUND: Prior work on opioid prescribing has examined dosing defaults, interruptive alerts, or "harder" stops such as electronic prescribing of controlled substances (EPCS), which has become increasingly required by state policy. Given that real-world opioid stewardship policies are concurrent and overlapping, the authors examined the effect of such policies on emergency department (ED) opioid prescriptions. METHODS: The researchers performed observational analysis of all ED visits discharged between December 17, 2016, and December 31, 2019, across seven EDs of a hospital system. Four interventions were examined in chronological order, with each successive intervention added on top of all previous interventions: 12-pill prescription default, EPCS, electronic health record (EHR) pop-up alert, and 8-pill prescription default. The primary outcome was opioid prescribing, which was described as number of opioid prescriptions per 100 discharged ED visits and modeled as a binary outcome for each visit. Secondary outcomes included prescription morphine milligram equivalents (MME) and non-opioid analgesia prescriptions. RESULTS: A total of 775,692 ED visits were included in the study. Compared to the preintervention period, cumulative reductions in opioid prescribing were seen with incremental interventions, including after adding a 12-pill default (odds ratio [OR] 0.88, 95% confidence interval [CI] 0.82-0.94), after adding EPCS (OR 0.7, 95% CI 0.63-0.77), after adding pop-up alerts (OR 0.67, 95% CI 0.63-0.71), and after adding an 8-pill default (OR 0.61, 95% CI 0.58-0.65). CONCLUSION: EHR-implemented solutions such as EPCS, pop-up alerts, and pill defaults had varying but significant effects on reducing ED opioid prescribing. Policy makers and quality improvement leaders might achieve sustainable improvements in opioid stewardship while balancing clinician alert fatigue through policy efforts promoting implementation of EPCS and default dispense quantities.
Assuntos
Analgésicos Opioides , Padrões de Prática Médica , Humanos , Analgésicos Opioides/uso terapêutico , Hospitais , Registros Eletrônicos de Saúde , Serviço Hospitalar de Emergência , Estudos RetrospectivosRESUMO
Importance: Emergency department (ED) triage models are intended to queue patients for treatment. In the absence of higher acuity, patients of the same acuity should room in order of arrival. Objective: To characterize disparities in ED care access as unexplained queue jumps (UQJ), or instances in which acuity and first come, first served principles are violated. Design, Setting, and Participants: Retrospective, cross-sectional study between July 2017 and February 2020. Participants were all ED patient arrivals at 2 EDs within a large Northeast health system. Data were analyzed from July to September 2022. Exposure: UQJ was defined as a patient being placed in a treatment space ahead of a patient of higher acuity or of a same acuity patient who arrived earlier. Main Outcomes and Measures: Primary outcomes were odds of a UQJ and association with ED outcomes of hallway placement, leaving before treatment complete, escalation to higher level of care while awaiting inpatient bed placement, and 72-hour ED revisitation. Secondary analysis examined UQJs among high acuity ED arrivals. Regression models (zero-inflated Poisson and logistic regression) adjusted for patient demographics and ED operational variables at time of triage. Results: Of 314â¯763 included study visits, 170â¯391 (54.1%) were female, the mean (SD) age was 50.46 (20.5) years, 132â¯813 (42.2%) patients were non-Hispanic White, 106â¯401 (33.8%) were non-Hispanic Black, and 66â¯465 (21.1%) were Hispanic or Latino. Overall, 90â¯698 (28.8%) patients experienced a queue jump, and 78â¯127 (24.8%) and 44â¯551 (14.2%) patients were passed over by a patient of the same acuity or lower acuity, respectively. A total of 52â¯959 (16.8%) and 23â¯897 (7.6%) patients received care ahead of a patient of the same acuity or higher acuity, respectively. Patient demographics including Medicaid insurance (incident rate ratio [IRR], 1.11; 95% CI, 1.07-1.14), Black non-Hispanic race (IRR, 1.05; 95% CI, 1.03-1.07), Hispanic or Latino ethnicity (IRR, 1.05; 95% CI, 1.02-1.08), and Spanish as primary language (IRR, 1.06; 95% CI, 1.02-1.10) were independent social factors associated with being passed over. The odds of a patient receiving care ahead of others were lower for ED visits by Medicare insured (odds ratio [OR], 0.92; 95% CI, 0.88-0.96), Medicaid insured (OR, 0.81; 95% CI, 0.77-0.85), Black non-Hispanic (OR, 0.94; 95% CI, 0.91-0.97), and Hispanic or Latino ethnicity (OR, 0.87; 95% CI, 0.83-0.91). Patients who were passed over by someone of the same triage severity level had higher odds of hallway bed placement (OR, 1.01; 95% CI, 1.00-1.02) and leaving before disposition (OR, 1.02; 95% CI, 1.01-1.04). Conclusions and Relevance: In this cross-sectional study of ED patients in triage, there were consistent disparities among marginalized populations being more likely to experience a UQJ, hallway placement, and leaving without receiving treatment despite being assigned the same triage acuity as others. EDs should seek to standardize triage processes to mitigate conscious and unconscious biases that may be associated with timely access to emergency care.