Browsing by Subject "Emergencies"

Now showing 1 - 6 of 6
  • Thumbnail Image
    Item
    Accuracy of a Commercial Large Language Model (ChatGPT) to Perform Disaster Triage of Simulated Patients Using the Simple Triage and Rapid Treatment (START) Protocol: Gage Repeatability and Reproducibility Study
    (JMIR, 2024-09-30) Franc, Jeffrey Micheal; Hertelendy, Attila Julius; Cheng, Lenard; Hata, Ryan; Verde, Manuela; Emergency Medicine, School of Medicine
    Background: The release of ChatGPT (OpenAI) in November 2022 drastically reduced the barrier to using artificial intelligence by allowing a simple web-based text interface to a large language model (LLM). One use case where ChatGPT could be useful is in triaging patients at the site of a disaster using the Simple Triage and Rapid Treatment (START) protocol. However, LLMs experience several common errors including hallucinations (also called confabulations) and prompt dependency. Objective: This study addresses the research problem: "Can ChatGPT adequately triage simulated disaster patients using the START protocol?" by measuring three outcomes: repeatability, reproducibility, and accuracy. Methods: Nine prompts were developed by 5 disaster medicine physicians. A Python script queried ChatGPT Version 4 for each prompt combined with 391 validated simulated patient vignettes. Ten repetitions of each combination were performed for a total of 35,190 simulated triages. A reference standard START triage code for each simulated case was assigned by 2 disaster medicine specialists (JMF and MV), with a third specialist (LC) added if the first two did not agree. Results were evaluated using a gage repeatability and reproducibility study (gage R and R). Repeatability was defined as variation due to repeated use of the same prompt. Reproducibility was defined as variation due to the use of different prompts on the same patient vignette. Accuracy was defined as agreement with the reference standard. Results: Although 35,102 (99.7%) queries returned a valid START score, there was considerable variability. Repeatability (use of the same prompt repeatedly) was 14% of the overall variation. Reproducibility (use of different prompts) was 4.1% of the overall variation. The accuracy of ChatGPT for START was 63.9% with a 32.9% overtriage rate and a 3.1% undertriage rate. Accuracy varied by prompt with a maximum of 71.8% and a minimum of 46.7%. Conclusions: This study indicates that ChatGPT version 4 is insufficient to triage simulated disaster patients via the START protocol. It demonstrated suboptimal repeatability and reproducibility. The overall accuracy of triage was only 63.9%. Health care professionals are advised to exercise caution while using commercial LLMs for vital medical determinations, given that these tools may commonly produce inaccurate data, colloquially referred to as hallucinations or confabulations. Artificial intelligence-guided tools should undergo rigorous statistical evaluation-using methods such as gage R and R-before implementation into clinical settings.
  • Thumbnail Image
    Item
    Effect of introducing the mucosal atomization device for fentanyl use in out-of-hospital pediatric trauma patients
    (Cambridge Journals, 2013-10) O'Donnell, Daniel; Schafer, Luke; Stevens, Andrew; Weinstein, Elizabeth; Miramonti, Charles; Kozak, Mary Ann
    Background: Pain associated with pediatric trauma is often under-assessed and undertreated in the out-of-hospital setting. Administering an opioid such as fentanyl via the intranasal route is a safe and efficacious alternative to traditional routes of analgesic delivery and could potentially improve pain management in pediatric trauma patients. Objective: The study sought to examine the effect of introducing the mucosal atomization device (MAD) on analgesia administration as an alternative to intravenous fentanyl delivery in pediatric trauma patients. The hypothesis for the study is that the introduction of the MAD would increase the administration of fentanyl in pediatric trauma patients. Methods: The research utilized a 2-group design (pre-MAD and post-MAD) to study 946 pediatric trauma patients (age ,16) transported by a large, urban EMS agency to one of eight hospitals in Marion County, which is located in Indianapolis Indiana. Two emergency medicine physicians independently determined whether the patient met criteria for pain medication receipt and a third reviewer resolved any disagreements. A comparison of the rates of fentanyl administration in both groups was then conducted. Results: There was no statistically significant difference in the rate of fentanyl administration between the pre-MAD (30.4%) and post-MAD groups (37.8%) (P5.238). A subgroup analysis showed that age and mechanism of injury were stronger predictors of fentanyl administration. Conclusion: Contrary to the hypothesis, the addition of the MAD device did not increase fentanyl administration rates in pediatric trauma patients. Future research is needed to address the barriers to analgesia administration in pediatric trauma patients.
  • Thumbnail Image
    Item
    Epidemiologic Trends In Children With Toxicologic Exposures Requiring Intensive Care Before and During the COVID-19 Pandemic
    (Pediatric Pharmacy Association, 2024) Hughes, Kaitlin; Overberg, Adam; Satterfield, Kennedi; Voss, Hannah; Rogerson, Colin; Pediatrics, School of Medicine
    Objective: Pediatric poison exposures are a common reason for pediatric intensive care unit (PICU) -admission. The purpose of this study was to examine the exposure trends and patient outcomes in 2018-2019 compared with 2020-2021 amidst the COVID-19 pandemic. Methods: This was a retrospective cohort study of patients 18 years of age or younger with a suspected toxicologic exposure from January 2018 to March 2021. The primary endpoint was rate of PICU admissions between the 2 cohorts. Secondary endpoints included medical outcome stratified by severity, PICU length of stay, and need for mechanical ventilation. Results: Our study included a total of 340 patients with median age 14.5 (IQR, 11.9-16.1) years. There was no significant difference in age, sex, or race between the 2 cohorts. The percentage of patients admitted to the PICU for poison exposures was significantly higher in the COVID-19 cohort compared with the pre-COVID-19 cohort (8.4% vs 3.7%, p < 0.01). Severity of medical outcomes differed between the groups; the COVID-19 cohort had more extreme clinical presentations of no effect or death (p < 0.01). No significant difference was found among the remaining secondary outcomes. Classes of substances ingested were comparable with baseline poison center data. Conclusions: Poisoning-related PICU admissions occurred at more than twice the pre-pandemic rate. This may emphasize the effect of the COVID-19 pandemic on pediatric access and exposure to poisons.
  • Thumbnail Image
    Item
    Improvement in Kansas City Cardiomyopathy Questionnaire Scores After a Self-Care Intervention in Patients With Acute Heart Failure Discharged From the Emergency Department
    (American Heart Association, 2021) Stubblefield, William B.; Jenkins, Cathy A.; Liu, Dandan; Storrow, Alan B.; Spertus, John A.; Pang, Peter S.; Levy, Phillip D.; Butler, Javed; Chang, Anna Marie; Char, Douglas; Diercks, Deborah B.; Fermann, Gregory J.; Han, Jin H.; Hiestand, Brian C.; Hogan, Christopher J.; Khan, Yosef; Lee, Sangil; Lindenfeld, JoAnn M.; McNaughton, Candace D.; Miller, Karen; Peacock, W. Frank; Schrock, Jon W.; Self, Wesley H.; Singer, Adam J.; Sterling, Sarah A.; Collins, Sean P.; Emergency Medicine, School of Medicine
    Background: We conducted a secondary analysis of changes in the Kansas City Cardiomyopathy Questionnaire (KCCQ)-12 over 30 days in a randomized trial of self-care coaching versus structured usual care in patients with acute heart failure who were discharged from the emergency department. Methods: Patients in 15 emergency departments completed the KCCQ-12 at emergency department discharge and at 30 days. We compared change in KCCQ-12 scores between the intervention and usual care arms, adjusted for enrollment KCCQ-12 and demographic characteristics. We used linear regression to describe changes in KCCQ-12 summary scores and logistic regression to characterize clinically meaningful KCCQ-12 subdomain changes at 30 days. Results: There were 350 patients with both enrollment and 30-day KCCQ summary scores available; 166 allocated to usual care and 184 to the intervention arm. Median age was 64 years (interquartile range, 55-70), 37% were female participants, 63% were Black, median KCCQ-12 summary score at enrollment was 47 (interquartile range, 33-64). Self-care coaching resulted in significantly greater improvement in health status compared with structured usual care (5.4-point greater improvement, 95% CI, 1.12-9.68; P=0.01). Improvements in health status in the intervention arm were driven by improvements within the symptom frequency (adjusted odds ratio, 1.62 [95% CI, 1.01-2.59]) and quality of life (adjusted odds ratio, 2.39 [95% CI, 1.46-3.90]) subdomains. Conclusions: In this secondary analysis, patients with acute heart failure who received a tailored, self-care intervention after emergency department discharge had clinically significant improvements in health status at 30 days compared with structured usual care largely due to improvements within the symptom frequency and quality of life subdomains of the KCCQ-12.
  • Thumbnail Image
    Item
    Relevant academic literature, applicable federal regulations for the protection of human subjects on emergency research involving artificial/substitute blood products (including PolyHeme)
    (Indiana University Center for Bioethics, 2006-03-27) Brown, Brandon
    Federal oversight of research involving human subjects is found in two regulatory regimes within the Department of Health and Human Services (DHHS), Food and Drug Administration (FDA). 21 CFR 50, 56 - the Office of Human Research Protections (OHRP), and 45 CFR 46. Generally, any research that is testing a drug, device, or other product that will be submitted for FDA approval must follow their regulations (21 CFR 50/56), while research that is supported by federal funds (e.g., an NIH grant) must also comply at a minimum with 45 CFR 46 Subpart A (the federal policy for the protection of human subjects, also known as the Common Rule), and with Subparts B,C,D as appropriate. While most of the FDA and OHRP regulations are similar (or substantially overlap), there are a number of areas in which they differ. Further, all institutions supported by federal funds must negotiate a Federalwide Assurance with OHRP that provides for all research within an institution to be subject to the Common Rule, regardless of whether the research is federally funded.
  • Thumbnail Image
    Item
    SHEA statement on antibiotic stewardship in hospitals during public health emergencies
    (Cambridge University Press, 2022) Barlam, Tamar F.; Al Mohajer, Mayar; Al-Tawfiq, Jaffar A.; Auguste, Antonie J.; Cunha, Cheston B.; Forrest, Graeme N.; Gross, Alan E.; Lee, Rachael A.; Seo, Susan K.; Suh, Kathryn N.; Volk, Stacy; Schaffzin, Joshua K.; Medicine, School of Medicine