Performance of emergency triage prediction of an open access natural language processing based chatbot application (ChatGPT): A preliminary, scenario-based cross-sectional study.

OBJECTIVES: Artificial intelligence companies have been increasing their initiatives recently to improve the results of chatbots, which are software programs that can converse with a human in natural language. The role of chatbots in health care is deemed worthy of research. OpenAI's ChatGPT is a supervised and empowered machine learning-based chatbot. The aim of this study was to determine the performance of ChatGPT in emergency medicine (EM) triage prediction. METHODS: This was a preliminary, cross-sectional study conducted with case scenarios generated by the researchers based on the emergency severity index (ESI) handbook v4 cases. Two independent EM specialists who were experts in the ESI triage scale determined the triage categories for each case. A third independent EM specialist was consulted as arbiter, if necessary. Consensus results for each case scenario were assumed as the reference triage category. Subsequently, each case scenario was queried with ChatGPT and the answer was recorded as the index triage category. Inconsistent classifications between the ChatGPT and reference category were defined as over-triage (false positive) or under-triage (false negative). RESULTS: Fifty case scenarios were assessed in the study. Reliability analysis showed a fair agreement between EM specialists and ChatGPT (Cohen's Kappa: 0.341). Eleven cases (22%) were over triaged and 9 (18%) cases were under triaged by ChatGPT. In 9 cases (18%), ChatGPT reported two consecutive triage categories, one of which matched the expert consensus. It had an overall sensitivity of 57.1% (95% confidence interval [CI]: 34-78.2), specificity of 34.5% (95% CI: 17.9-54.3), positive predictive value (PPV) of 38.7% (95% CI: 21.8-57.8), negative predictive value (NPV) of 52.6 (95% CI: 28.9-75.6), and an F1 score of 0.461. In high acuity cases (ESI-1 and ESI-2), ChatGPT showed a sensitivity of 76.2% (95% CI: 52.8-91.8), specificity of 93.1% (95% CI: 77.2-99.2), PPV of 88.9% (95% CI: 65.3-98.6), NPV of 84.4 (95% CI: 67.2-94.7), and an F1 score of 0.821. The receiver operating characteristic curve showed an area under the curve of 0.846 (95% CI: 0.724-0.969, P < 0.001) for high acuity cases. CONCLUSION: The performance of ChatGPT was best when predicting high acuity cases (ESI-1 and ESI-2). It may be useful when determining the cases requiring critical care. When trained with more medical knowledge, ChatGPT may be more accurate for other triage category predictions.

Prevalence and associated risk factors of post-traumatic stress disorder among survivors of the 2023 Turkey earthquake.

BACKGROUND: Earthquakes can cause psychological trauma among survivors as well as physical trauma. This study aims to determine the prevalence of post-traumatic stress disorder (PTSD) and identify associated risk factors among earthquake survivors after the 2023 Turkey earthquake. METHODS: This prospective cross-sectional study was conducted in the emergency department of a tertiary university hospital between May 6, 2023, and May 16, 2023. An online questionnaire was sent to the participants. The questionnaire form consisted of three parts. In the first part, the sociodemographic characteristics of the participants were included. In the second part, the participants were asked about their experiences with the earthquake. In the third part, the post-traumatic stress disorder checklist for the Diagnostic and Statistical Manual of Mental Disorders, which screens PTSD among the participants, was included. Participants who did not complete all the questions and had severe communication disorders were excluded. RESULTS: The prevalence of probable PTSD among the participants was 51.4% (n = 197). Age (OR: 0.96 95% CI: 0.93-99), female gender (OR: 4.54 95% CI: 2.39-8.61), being the head of the family (OR: 2.00 95% CI: 1.04-3.82), bereavement (OR: 1.71 95% CI: 1.03-2.82), lost loved ones (OR: 3.15 95% CI: 1.67-5.92), low social support (OR: 1.80 95% CI: 1.12-2.90) and receiving emergency care at the field (OR: 6.67 95% CI: 1.03-43.2) were the associated risk factors of PTSD among earthquake survivors. CONCLUSIONS: The prevalence of PTSD among survivors three months after the 2023 Turkey earthquake is over half of the survivors. Younger age, female gender, being the head of the family, bereavement, lost loved ones, low social support, and receiving emergency care in the field were the associated risk factors of PTSD among earthquake survivors. Considering survivors may visit EDs until other outpatient clinics are re-established and the high rate of PTSD, rapid psychological evaluations can be performed in emergency departments. Emergency physicians should be aware of possible risk factors and high rate of PTSD.

CHAMPS score in predicting mortality of patients with acute nonvariceal upper gastrointestinal bleeding.

OBJECTIVE: The aim of this study was to evaluate the performance of the Charlson Comorbidity Index ≥2, in-hospital onset, albumin <2.5 g/dL, altered mental status, Eastern Cooperative Oncology Group performance status ≥2, steroid use score in predicting mortality in patients with nonvariceal upper gastrointestinal bleeding and compare it with the Glasgow-Blatchford score; the albumin, international normalized ratio; alteration in mental status, systolic blood pressure, and age 65 score; the age, blood tests, and comorbidities score; and Complete Rockall score. METHODS: The data of patients with acute upper gastrointestinal bleeding who visited the emergency department during the study period were obtained from the hospital automation system by using the classification of disease codes and analyzed in this retrospective study. Adult patients with endoscopically confirmed nonvariceal upper gastrointestinal bleeding were included in the study. Patients with bleeding from the tumor, bleeding after endoscopic resection, or missing data were excluded. The prediction accuracy of the Charlson Comorbidity Index ≥ 2, in-hospital onset, albumin < 2.5 g/dL, altered mental status, Eastern Cooperative Oncology Group performance status ≥ 2, steroid use score was calculated using the area under the receiver operating characteristic curve and compared with that of Glasgow-Blatchford score, the albumin, international normalized ratio; alteration in mental status, systolic blood pressure, and age 65 score, the age, blood tests, and comorbidities score, and Complete Rockall score. RESULTS: A total of 805 patients were included in the study, and the in-hospital mortality rate was 6.6%. The performance of the Charlson Comorbidity Index ≥ 2, in-hospital onset, albumin < 2.5 g/dL, altered mental status, Eastern Cooperative Oncology Group performance status ≥ 2, steroid use score (area under the receiver operating characteristic curve 0.812, 95%CI 0.783-0.839) was better than Glasgow-Blatchford score (area under the receiver operating characteristic curve 0.683, 95%CI 0.650-0.713, p=0.008), and similar to the the age, blood tests, and comorbidities score (area under the receiver operating characteristic curve 0.829, 95%CI 0.801-0.854, p=0.563), the albumin, international normalized ratio; alteration in mental status, systolic blood pressure, and age 65 score (area under the receiver operating characteristic curve 0.794, 95%CI 0.764-0.821, p=0.672), and Complete Rockall score (area under the receiver operating characteristic curve 0.761, 95%CI 0.730-0.790, p=0.106). CONCLUSION: The performance of the Charlson Comorbidity Index ≥ 2, in-hospital onset, albumin < 2.5 g/dL, altered mental status, Eastern Cooperative Oncology Group performance status ≥ 2, steroid use score in predicting in-hospital mortality for our study population is better than Glasgow-Blatchford score and similar to the the age, blood tests, and comorbidities score, the albumin, international normalized ratio; alteration in mental status, systolic blood pressure, and age 65 score, and Complete Rockall score.

Antioxidant and Neuroprotective Effects of L-arginine Administration After Traumatic Brain Injury and Hemorrhagic Shock in Rats.

AIM: To investigate the effect of fluid resuscitation and L-arginine administration on oxidant status markers, blood gases, lactate and apoptosis in the brain tissue of a rat model of TBI with hemorrhagic shock. MATERIAL AND METHODS: A total of 60 rats were divided into six groups: control, isotonic saline-treated, 7.5% NaCl-treated (hypertonic saline), L-arginine-treated (100 mg/kg), saline + L-arginine-treated and 7.5% NaCl + L-arginine-treated groups. Closed head contusive weight-drop injuries were performed with hemorrhagic shock in all of the groups. Mean arterial pressure (MAP), pulse rate, lactate, malondialdehyde (MDA), total antioxidant capacity (TAC) and apoptosis were investigated. RESULTS: In a total of 48 rats, MAP levels remained higher than 60 mmHg for 3 hours in all of the treatment groups. The highest MAP values in each group were recorded. Higher MDA and lower TAC levels were observed in the control group than in all of the treatment groups (all p < 0.05). The number of apoptotic cells was highest in the control group and lowest in the L-arginine group. CONCLUSION: L-arginine administration may be an alternative treatment option for individualized fluid resuscitation in patients with TBI and hemorrhagic shock.

CHAMPS score in predicting mortality of patients with acute nonvariceal upper gastrointestinal bleeding

SUMMARY OBJECTIVE: The aim of this study was to evaluate the performance of the Charlson Comorbidity Index ≥2, in-hospital onset, albumin <2.5 g/dL, altered mental status, Eastern Cooperative Oncology Group performance status ≥2, steroid use score in predicting mortality in patients with nonvariceal upper gastrointestinal bleeding and compare it with the Glasgow-Blatchford score; the albumin, international normalized ratio; alteration in mental status, systolic blood pressure, and age 65 score; the age, blood tests, and comorbidities score; and Complete Rockall score. METHODS: The data of patients with acute upper gastrointestinal bleeding who visited the emergency department during the study period were obtained from the hospital automation system by using the classification of disease codes and analyzed in this retrospective study. Adult patients with endoscopically confirmed nonvariceal upper gastrointestinal bleeding were included in the study. Patients with bleeding from the tumor, bleeding after endoscopic resection, or missing data were excluded. The prediction accuracy of the Charlson Comorbidity Index ≥ 2, in-hospital onset, albumin < 2.5 g/dL, altered mental status, Eastern Cooperative Oncology Group performance status ≥ 2, steroid use score was calculated using the area under the receiver operating characteristic curve and compared with that of Glasgow-Blatchford score, the albumin, international normalized ratio; alteration in mental status, systolic blood pressure, and age 65 score, the age, blood tests, and comorbidities score, and Complete Rockall score. RESULTS: A total of 805 patients were included in the study, and the in-hospital mortality rate was 6.6%. The performance of the Charlson Comorbidity Index ≥ 2, in-hospital onset, albumin < 2.5 g/dL, altered mental status, Eastern Cooperative Oncology Group performance status ≥ 2, steroid use score (area under the receiver operating characteristic curve 0.812, 95%CI 0.783-0.839) was better than Glasgow-Blatchford score (area under the receiver operating characteristic curve 0.683, 95%CI 0.650-0.713, p=0.008), and similar to the the age, blood tests, and comorbidities score (area under the receiver operating characteristic curve 0.829, 95%CI 0.801-0.854, p=0.563), the albumin, international normalized ratio; alteration in mental status, systolic blood pressure, and age 65 score (area under the receiver operating characteristic curve 0.794, 95%CI 0.764-0.821, p=0.672), and Complete Rockall score (area under the receiver operating characteristic curve 0.761, 95%CI 0.730-0.790, p=0.106). CONCLUSION: The performance of the Charlson Comorbidity Index ≥ 2, in-hospital onset, albumin < 2.5 g/dL, altered mental status, Eastern Cooperative Oncology Group performance status ≥ 2, steroid use score in predicting in-hospital mortality for our study population is better than Glasgow-Blatchford score and similar to the the age, blood tests, and comorbidities score, the albumin, international normalized ratio; alteration in mental status, systolic blood pressure, and age 65 score, and Complete Rockall score.

Retrospective Analysis of Intracranial Hemorrhages in the COVID-19 Pandemic.

AIM: To identify the characteristics of patients admitted to the emergency department with intracranial hemorrhages in the era of the COVID-19 pandemic. MATERIAL AND METHODS: Seventy-eight patients with spontaneous intracranial bleeding who were admitted to emergency departments and treated in neurosurgery clinics between March 11, 2020, and September 11, 2020, were included in the study. RESULTS: The most frequent symptom was the loss of consciousness (32.1%), followed by headache (15.4%), syncope (10.3%), motor loss (9%), and seizures (9%). Antiaggregant and anticoagulant drug use were detected in 37.2% of the patients. Intraparenchymal hematoma was the most common type of intracranial hemorrhage (59%). Viral pneumonia was detected in 52.6% of the patients in thorax CTs. Surgical treatment was applied to 23.1% of the patients. There was no significant difference between patients with pneumonia and patients without pneumonia in the treatment modalities or 30-day mortality. CONCLUSION: In this study, we found that low Glasgow Coma Scores affected mortality and that mechanical ventilation needs are higher in ICH patients with COVID-19. Nevertheless, the treatment differences may not have affected the outcomes.

An extremely rare complication of acute pancreatitis: Intraventricular thrombus.

Acute pancreatitis is a frequent reason for emergency admission, which has seen its numbers increase over the years. This condition has systemic, local, and vascular complications. A 73-year-old male patient presented to our emergency department complaining of abdominal pain, nausea, and vomiting. During imaging, intraventricular thrombus was discovered, and following completion of diagnostic testing, he was diagnosed with acute pancreatitis. Herein, we present the first case of intraventricular thrombus related to acute pancreatitis prothrombotic process in the literature.

Diagnosis of Acute Coronary Syndrome with a Support Vector Machine.

Acute coronary syndrome (ACS) is a serious condition arising from an imbalance of supply and demand to meet myocardium's metabolic needs. Patients typically present with retrosternal chest pain radiating to neck and left arm. Electrocardiography (ECG) and laboratory tests are used indiagnosis. However in emergency departments, there are some difficulties for physicians to decide whether hospitalizing, following up or discharging the patient. The aim of the study is to diagnose ACS and helping the physician with his decisionto discharge or to hospitalizevia machine learning techniques such as support vector machine (SVM) by using patient data including age, sex, risk factors, and cardiac enzymes (CK-MB, Troponin I) of patients presenting to emergency department with chest pain. Clinical, laboratory, and imaging data of 228 patients presenting to emergency department with chest pain were reviewedand the performance of support vector machine. Four different methods (Support vector machine (SVM), Artificial neural network (ANN), Naïve Bayes and Logistic Regression) were tested and the results of SVM which has the highest accuracy is reported. Among 228 patients aged 19 to 91 years who were included in the study, 99 (43.4 %) were qualified as ACS, while 129 (56.5 %) had no ACS. The classification model using SVM attained a 99.13 % classification success. The present study showed a 99.13 % classification success for ACS diagnosis attained by Support Vector Machine. This study showed that machine learning techniques may help emergency department staff make decisions by rapidly producing relevant data.