Western Journal of Emergency Medicine最新文献

Introduction: Prior research demonstrates that emergency department (ED) patients with suicidal ideation (SI), substance use (SUD), and/or intimate partner violence (IPV) have disproportionate adverse outcomes for both women and infants. The 2013 Hague Protocol suggested that children with caregivers with the above characteristics are also more likely to suffer from child maltreatment. Of all pregnancies in this group, as many as 90% are unintended. We hypothesized that women with SI/SUD/IPV have gaps in care access, high levels of unscheduled care use, and reduced ED contraceptive inquiry, which if addressed could potentially improve outcomes.

Methods: We conducted a chart review of 62,284 ED visits from 2018-2021 from a suburban four-hospital system in the Southern United States. We compared women of reproductive age (15-44) with SI/SUD/IPV (4,776) against controls (57,508). The exposures were defined as women with SI, SUD, and/or IPV. We analyzed results using the chi-square test (χ²) with Bonferroni adjustment to test for independence and logistic regression.

Results: Women suffering from SI/SUD/IPV who present to the ED have contraceptive status less frequently documented compared to controls without these factors (39.5 vs 51.7%, RR 0.77, CI, 0.74-0.79, P < .001). They also have reduced access to care, with higher rates of uninsurance (32.7 vs 26.1%, P < .001), more care in the acute care environment, longer ED length of stay (LOS) (mean was 10.38 vs 3.87 hours, P < .001), higher hospitalization rates (61.0 vs 8.7%, P < .001), and higher 30-day ED revisits (11.8 vs 8.8%, P < .001), even after adjusting for the Social Vulnerability Index, acuity, age, and obesity (adjusted odds ratio 1.52 95% CI 1.36-1.70 P < .001).

Conclusion: Despite significant morbidity coupled with reduced access to ambulatory care and disproportionately increased ED use, little ED contraceptive documentation exists. This practice contributes to inequity, given the increased number of unintended pregnancies and greater need of contraceptives in women with suicidal ideation/substance use disorder/intimate partner violence.

Introduction: Patients leaving without being seen is a critical quality metric for emergency department (ED) performance and is associated with negative patient outcomes and operational inefficiencies. In this study we aimed to systematically assess patient- and system-level factors influencing leaving-without-being-seen behavior.

Methods: We conducted a retrospective cohort study at The Ottawa Hospital, a tertiary-care ED with 85,000 annual ED visits in Ottawa, Canada. We analyzed all patient encounters for two years from May 2022-April 2024. Variables included demographics characteristics (age, sex), visit specifics (arrival day and time, Canadian Triage and Acuity Scale [CTAS] scores, presenting complaints), and operational metrics (ED occupancy metrics). Multivariate logistic regression analyses evaluated the influence of these factors on rates of leaving without being seen.

Results: Of 170,536 ED visits, 15,473 (9.1%) patients left without being seen, and 2,716 (1.6%) left before triage. Each additional 10 years of age reduced the adjusted odds of leaving without being seen by 20.2% (older patients left less frequently). Male patients had 9.4% higher adjusted odds of leaving without being seen compared to females. For every five patients waiting to be seen, the adjusted odds of leaving increased by 16.9% for a newly arriving patient. For every five patients already seen but awaiting disposition, the adjusted odds of leaving increased by 9.6% for a newly arriving patient. Compared to CTAS 2 patients (high acuity), CTAS 3 patients had 67.1% higher adjusted odds of leaving, CTAS 4 patients had 134% higher adjusted odds, and CTAS 5 patients (lowest acuity) had 176% higher adjusted odds of leaving.

Conclusion: Younger age, male sex, lower acuity, and ED crowding independently and significantly increase rates of leaving without being seen. Importantly, both crowding and volume of patients waiting impact left-without-being-seen behaviour. Optimizing patient flow through strategic movement within the ED may enhance the perception of progress, encouraging patients to remain for care.

Introduction: Pulmonary embolism (PE) is common with potential for morbidity and mortality. Several PE risk-stratification tools exist; however, more granular and patient-specific indicators of potential decompensation or short-term mortality that can be easily obtained are needed for the bedside clinician to further sub-stratify risk and inform management decisions. We sought to determine the association of early emergency department (ED) measurement of the shock index (SI) and SI variants (modified SI, SI to peripheral oxygen saturation ratio, age-adjusted SI, respiratory-adjusted SI, and double product) and mortality among patients with acute PE.

Methods: This was an observational case-control study of adult patients who presented to the ED at a single health system (January 2021-April 2023) and had PE response team (PERT) activation for newly diagnosed acute PE. We evaluated the association of 30-day in-hospital mortality with the SI (heart rate/systolic blood pressure) and variants of the SI-modified SI = heart rate/mean arterial pressure; SI to peripheral oxygen saturation ratio = SI/peripheral oxygen saturation; age-adjusted SI = age x SI; respiratory-adjusted SI = SI x (respiratory rate/10); double product = systolic blood pressure x heart rate-in addition to the Simplified Pulmonary Embolism Severity Index (sPESI) and European Society of Cardiology (ESC) risk schema. We used the area under the receiver operating characteristic curve (AUC) to assess discriminatory efficiency of the SI and each variant with the primary outcome. Multivariable logistic regression measured the association between SI and variants with 30-day mortality.

Results: Of 121 patients included in the study, 12 (9.9%) died. The SI and variants were all significantly different between survivors and non-survivors (P < .05), while the sPESI was not different (P = .30). The age-adjusted SI had the highest discriminatory efficiency for mortality (AUC 0.82; 95% CI, 0.71-0.93), followed by the SI (AUC 0.78; 0.67-0.89), the SI/peripheral oxygen saturation (AUC 0.77; 0.65-0.90), double product (AUC 0.76; 0.61-0.91), modified SI (AUC 0.75; 0.61-0.90), ESC risk schema (AUC 0.71; 0.52-0.90), and the respiratory-adjusted SI (AUC 0.70; 0.54-0.87).

Conclusion: Among patients presenting to the ED who had a PERT activation for acute PE, the age-adjusted SI had the highest discriminatory efficiency for mortality, followed by the SI and its other variants. Further investigation regarding use of the age-adjusted SI for prognostication of acute PE and implications on PE management is warranted.

Introduction: The coral snake is the only native elapid in North America. Their venom contains potent neurotoxins. Historically, all confirmed/presumed bites were treated with antivenom whether or not symptoms were present. Production of antivenom ceased in 2003. The resultant national shortage prompted clinicians to investigate alternative treatment strategies such as a wait-and-see approach where antivenom is held until signs of systemic toxicity manifest. Now that production has resumed there is limited research available comparing these two treatment paradigms, empiric administration vs the wait-and-see approach. Our objective in this study was to compare outcomes of the two treatment paradigms to determine whether one is associated with better patient outcomes.

Methods: This was a retrospective analysis of coral snakebite cases reported to the Florida Poison Information Center Network from January 1, 1998-December 31, 2021. We collected demographic, clinical, and outcome variables. Patients were stratified into two groups, empiric antivenom administration vs the wait-and-see approach in patients who were asymptomatic in terms of systemic symptoms at the time of initial presentation to the emergency department. We used multivariable logistic regression models, controlling for whether the bite occurred during the North American Coral Snake Antivenin (NACSA) shortage period (yes/no), age, sex, and whether systemic effects developed (yes/no), to determine differences between study groups in the incidence of the main outcomes: intensive care unit (ICU) admission; intubation; and death, as well as ICU and hospital length of stay.

Results: We analyzed 301 cases: 171 (56.8%) empiric; and 130 (43.2%) wait-and-see. Patients in the empiric treatment group had approximately three times higher likelihood of ICU admission (empiric 121 [75.2%] and wait-and-see 71 [56.8%]), odds ratio [OR} 3.047, P = .05). There was no difference in the incidence of intubation (empiric 2 [1.2%] and wait-and-see 1 [<1%]), OR 2.486, P = .63) or in ICU length of stay (OR 0.485, P = .08). Of the patients treated with NACSA (191), adverse reactions to the antivenom occurred in 38 (19.9%) patients-35 patients in the empiric group and three in the wait-and-see group who later received antivenom. Of these 38 patients, eight (21.1%) experienced an anaphylactic reaction.

Conclusion: Empiric North American Coral Snake Antivenin administration was associated with higher ICU admissions and with a considerably higher risk of adverse reactions, which may serve to impose caution when treating empirically.

Introduction: The COVID-19 pandemic disrupted care-seeking and respiratory disease epidemiology across healthcare settings, notably for emergency department (ED) care. The scope of this disruption and whether patterns of ED visits have returned to predictable seasonal patterns is of interest in planning ED staffing and resource availability for future illness surges, pandemic or not. We evaluated ED visits for acute respiratory illness among children in a large, integrated healthcare delivery system to describe illness and patient characteristics in the years before, during, and after the pandemic peak.

Methods: We conducted a cross-sectional study of ED visits among patients 0-17 years of age to the 21 EDs of Kaiser Permanente Northern California, from January 1, 2018-December 31, 2019, pre-pandemic; January 1, 2020-December 31, 2021, pandemic; and January 1, 2022-March 31, 2024, post-vaccine (vaccines for children > 5 years of age approved and available). We electronically extracted eligible ED visits with acute respiratory infection diagnoses and a range of sociodemographic, medical comorbidity, and utilization characteristics.

Results: We observed 151,983 pediatric ED visits with eligible respiratory infection diagnoses, 49,912 (32.8%) visits pre-pandemic, 27,109 (17.8%) visits during the pandemic, and 74,962 (49.3%) visits post-vaccine. Eligible visits dropped every month from 6,361 in February 2020, just prior to the pandemic onset, to their lowest volume (243) in June 2020. In the post-vaccine period, visits peaked at 10,638 in November 2022, the highest of any month during the study period. Sex, race/ethnicity, and tobacco exposure were comparable over time, but the proportion of visits by patients with under-immunized diagnosis trended upward over time. Upper respiratory infection (30% pre-pandemic, 32% pandemic, and 33% post-vaccine periods), asthma (15% pre-pandemic, 12% pandemic, and 12% post-vaccine periods), and cough (9.9% pre-pandemic, 12% pandemic, and 12% post-vaccine periods), were the top three diagnoses across all periods.

Conclusion: In this cross-sectional study of acute respiratory illness-related ED visits in an integrated healthcare system, from 2022 onward seasonal variation in respiratory illness ED visits rebounded, with notable and unseasonal peaks in late 2022. COVID-19 appears to be a minor contributor to ED visits for pediatric respiratory illness. However, an increased overall and seasonal burden of ED visits has implications for surge planning and mitigation, with COVID-19 now being endemic and typical respiratory pathogens having resurfaced.

Introduction: ChatGPT and other large language models (LLM) have increased in popularity. Despite the rapid rise in the implementation of such technologies, frameworks for implementing appropriate prompting techniques in medical applications are limited. In this paper we establish the nomenclature of "variable" and "clause" in the prompting of a LLM, while providing example interviews that outline the utility of such an approach in medical applications.

Methods: In this study assessing the LLM ChatGPT-4, we define terms used in prompting procedures including "input prompt," "variable," "demographic variable and clause," "independent variable and clause," "dependent variable and clause," "generative clause," and "output." This methodology was implemented with three sample patient cases from both a patient and physician perspective.

Results: As demonstrated in our three cases, precise combinations of variables and clauses that consider the patient's age, gender, weight, height, and education level can yield unique outputs. The software can do so quickly and in a personalized, patient-specific manner. Our findings demonstrate that LLMs can be used to generate comprehensive sets of educational material to augment current limitations, with the potential of improving healthcare outcomes as the use of LLM is further explored.

Conclusion: The framework we describe represents a unique attempt to standardize a methodology for medical inputs into a large language model. Doing so expands the potential for outlining patient-specific information that can be implemented in a query by either a patient or a physician. Most notably, future projects should consider the specialty- and presentation-specific input changes that may yield the best outputs for the desired goals.

Introduction: Despite numerous randomized controlled trials, lung protective ventilation and prone positioning remain the only therapies shown to have a survival benefit in acute respiratory distress syndrome (ARDS). A National Heart, Lung, and Blood Institute workshop on the future of clinical research in ARDS suggested that identification of at-risk patients earlier in their clinical course would allow implementation of prevention strategies and facilitate study of these interventions. To this end, the Lung Injury Prevention Score (LIPS) was derived and validated to identify patients at risk of developing ARDS upon hospital admission, and the Emergency Department Lung Injury Prevention Score (EDLIPS) was subsequently derived and internally validated. For this study, we sought to externally validate EDLIPS.

Methods: We performed a validation study of EDLIPS, using data from a large, multicenter trial- the Vitamin D to Improve Outcomes by Leveraging Early Treatment (VIOLET) trial. After identifying patients who met VIOLET inclusion criteria while in the ED, variables comprising EDLIPS were extracted for each patient. We calculated area under the receiver operating characteristic curves (AUC) of EDLIPS for the VIOLET dataset.

Results: We identified a total of 1,270 patients. The mean age was 56, and 55% were male. The incidence of ARDS was 8.1%. EDLIPS discriminated patients who developed ARDS from those who did not with an AUC of 0.786 (95% CI, 0.740-0.832), nearly identical to its performance in the original study, which yielded an AUC of 0.784 (95% CI, 0.748-0.820).

Conclusion: We successfully validated a risk-prediction model for the identification of ED patients at risk for ARDS in a large cohort of critically ill patients. The development of ARDS prevention trials will involve collaboration with other clinical groups, such as emergency physicians, to enroll patients as early as possible in their clinical course. EDLIPS is the first tool of its kind to undergo external validation, and it can aid in the identification of ED patients at risk for the development of ARDS.