Cabrera, Daniel

Objectives: Electronic health records (EHR)-based triggers (eTriggers) have been used to study diagnostic errors in the emergency department (ED), often with suboptimal performance. Our objective was to investigate incremental value of multi-factor machine learning (ML) approaches to improve eTrigger performance. Methods: Patients presenting to an academic ED were categorized into trigger-positive and trigger-negative using standard trigger (T) definitions: (T1) ED return visits resulting in admission within 10 days; (T2) care escalation from the inpatient unit to the ICU within 24 hours; and (T3) deaths within 24 hours of admission. We trained and evaluated 6 supervised ML models. Results: A total of 124,053 consecutive encounters (5791 T-positive and 118,262 T-negative) were included. Among the T-positive, 4159 (72%) were associated with T1, 1415 (24%) with T2, and 217 (4%) with T3. The T-based positive predictive values (PPV) were 5.2% for T1, 8.2% for T2, and 6.5% for T3. ML models trained and evaluated on balanced training dataset and imbalanced test set had low classification performances (accuracy: 0.72–0.95; PPV: 0.00–0.16; F1-score: 0.00–0.23). Higher performances were observed in balanced test sets (accuracy: 0.80–0.97; PPV: 0.82–1.00; F1-score: 0.79–0.97). Comparing models trained on clinically annotated data with models trained on T-based labels identified other important factors. Conclusions: Utilizing machine learning to refine e-triggers slightly improves the identification of diagnostic errors, as evidenced by an increase in PPV values. We identified new potential factors contributing to ED diagnostic errors. These findings open new avenues to construct or modify more accurate e-triggers for diagnostic error identification.

Importance Diagnostic errors represent a major patient safety concern, with the potential to significantly impact patient outcomes. To address this, various trigger-based strategies have been developed to identify diagnostic errors, aiming to enhance clinical decision-making and improve patient safety. Objective To evaluate the performance of three pre-established triggers (T) in the emergency department (ED) setting and assess their effectiveness in detecting diagnostic errors. Design Consecutive cohort, retrospective observational design. Setting Academic ED with 80 000 annual visits. Participants Adults and children presenting to a single ED in the USA between 1 May 2018 and 1 January 2020. Intervention/outcomes Electronic health records (EHRs) were retrieved and categorised into trigger-positive and trigger-negative cases using the following criteria: T1—unscheduled returnvisits to the ED with admission within 7–10 days of theinitial visit; T2—care escalation from the inpatient unitto the intensive care unit (ICU) within 6, 12 or 24 hoursof ED admission; and T3—all deaths in the ED or within24 hours of ED admission, excluding palliative care. A random sample of trigger-positive cases was reviewed using the SaferDx tool to determine the presence or absence of a diagnostic error. Results A total of 5791 trigger-positive and 118262 trigger-negative cases were identified. Among trigger-positive cases, 4159 (72%) were associated with T1, 1415 (24%) with T2, and 217 (4%) with T3. A preliminary chart review of 462 trigger-positive and 251 trigger-negative cases showed most were error-negative (279 and 217, respectively). Detailed reviews found 32 diagnostic errors among 183 trigger-positive cases, yielding PPVs of 5.4% (T1), 8.9% (T2), and 6.9% (T3). No errors were found in 34 reviewed trigger-negative cases, resulting in a 100% NPV. Sepsis was the most common diagnosis among error-positive cases (n=11, 34.4%). Those with non-specific chief complaints like altered mental status or shortness of breath had higher diagnostic error risk. Conclusion and relevance While previously proposed EHR-based triggers can identify some diagnostic errors, they are insufficient for detecting all cases. To improve error detection performance, we recommend exploring data-driven strategies, such as machine learning techniques, to more effectively identify underlying contributing factors to diagnostic errors and enhance detection accuracy in the ED.

Abstract Purpose Rapid sequence intubation (RSI) is a common emergency department (ED) procedure with an associated complication of postintubation hypotension (PIH). It has not been clearly established whether the selection and dose of induction agent affect risk of PIH. The objective of this study was to determine the incidence of PIH in patients receiving full-dose compared to reduced-dose induction agent for RSI in the ED. Methods This was a health system–wide, retrospective cohort study comparing incidence of PIH based on the induction medication and dose given for RSI in the ED. Patients were included if they underwent RSI from July 1, 2018, through December 31, 2020, were 18 years of age or older, and received etomidate or ketamine. A reduced dose was defined as a ketamine dose of 1.25 mg/kg or less and an etomidate dose of 0.2 mg/kg or less. Results A total of 909 patients were included in the final analysis, with most receiving etomidate (n = 764; 84%) and a smaller number receiving ketamine (n = 145; 16%). Patients who received ketamine had a higher mean pre-intubation shock index (full dose, 1.08; reduced dose, 1.04) than those who received etomidate (full dose, 0.89; reduced dose, 0.92) (P ≤ 0.001). Reduced doses of induction agent were observed for 107 patients receiving etomidate (14.0%) and 60 patients receiving ketamine (41.4%). Patients who received full-dose ketamine for induction had the highest rate of PIH (n = 31; 36.5%), and the difference was statistically significant compared to patients receiving reduced-dose ketamine (16.7%; P = 0.021) and full-dose etomidate (22.8%; P = 0.010). Conclusion We observed that full-dose ketamine was associated with the highest rate of PIH; however, this group had the poorest baseline hemodynamics, confounding interpretation. Our results do not support broad use of a reduced-dose induction agent.

ABSTRACT Chilaiditi sign is an incidental radiological finding where the intestine is interposed between the diaphragm and liver. Chilaiditi syndrome (CS), characterized by gastrointestinal symptoms and Chilaiditi sign on imaging, is of important clinical significance despite its rarity given associated complications including intestinal obstruction, bowel ischemia, and perforation. While most cases involve the large intestine, we report a rare case of CS with ileal involvement complicated by small bowel obstruction, managed conservatively. Failure to recognize Chilaiditi sign or CS may prompt unnecessary surgical interventions, emphasizing the need for physician awareness to ensure accurate timely diagnosis and appropriate management.

Abstract Background Precision medicine, sometimes referred to as personalized medicine, is rapidly changing the possibilities for how people will engage health care in the near future. As technology to support precision medicine exponentially develops, there is an urgent need to proactively improve our understanding of precision medicine and pose important research questions (RQs) related to its inclusion in the education and training of future emergency physicians. Methods A seven‐step process was employed to develop a research agenda exploring the intersection of precision and emergency medicine education/training. A literature search of articles about precision medicine was conducted first, which informed the creation of future four scenarios in which trainees and practicing physicians regularly discuss and incorporate precision medicine tools into their discussions and work. Based on these futurist narratives, potential education RQs were generated by an expert panel. A total of 59 initial questions were subsequently categorized and refined to a priority list through a nominal group voting method. The top/priority questions were presented at the 2023 SAEM Consensus Conference on Precision Medicine, Austin, Texas, for further input. Results Eight high‐value education RQs were developed, reflecting a holistic view of the challenges and opportunities for precision medicine education in the knowledge, skills, and attitudes relevant to emergency medicine. These questions contend with topics such as most effective pedagogical methods; intended resulting outcomes and behaviors; the generational differences between practicing emergency physicians, educators, and future trainees; and the desires and expectations of patients. Conclusions Emergency medicine and emergency physicians must be prepared to understand precision medicine and incorporate this information into their “toolbox” of thinking, problem solving, and communication with patients and colleagues. This research agenda on how best to educate future emergency physicians in the use of personalized data to provide optimal health care is the focus of this article.