Clinical trials

This retrospective diagnostic accuracy study evaluates the ability of two large language models (LLMs) - GPT-4o (gpt-4o-2024-11-20; OpenAI) and Claude 4.6 Sonnet (claude-sonnet-4-6; Anthropic) - to generate correct diagnoses from anonymized Turkish-language emergency department (ED) anamnesis notes, and compares their performance with the diagnosis entered by the treating emergency physician. A consensus gold standard is established by three independent board-certified emergency medicine specialists who blindly review each note and vote on the primary diagnosis using ICD-10 three-character codes; the majority vote (at least 2 of 3 specialists agreeing) constitutes the reference standard. Both LLMs are evaluated using a standardized zero-shot direct prompting strategy (temperature=0, stateless API sessions). The primary outcome is diagnostic accuracy (proportion of ICD-10 chapter-level matches) and Cohen's kappa for each LLM against the gold standard. Secondary outcomes include top-3 accuracy, treating physician accuracy, inter-model agreement, and subgroup analyses by ESI triage level and ICD-10 chapter. Inter-rater reliability among the three specialists is quantified using Fleiss' kappa. Analyses are performed in Jamovi. This study represents the first evaluation of LLM diagnostic accuracy using Turkish-language clinical notes and the first to benchmark LLM performance against an independent three-specialist majority-vote gold standard rather than against the treating physician's own diagnosis.

This prospective observational diagnostic accuracy study evaluates whether large language models (LLMs) - GPT-4o (OpenAI, gpt-4o-2024-11-20) and Claude (Anthropic, claude-sonnet-4-6) - can accurately calculate HEART scores from unstructured Turkish clinical notes and predict 30-day major adverse cardiac events (MACE) in emergency department patients presenting with non-traumatic chest pain. The study will enroll 600 consecutive adult patients. For each patient, the same anonymized data (free-text anamnesis, ECG report text, troponin value, and age) will be independently processed by both LLMs via separate API calls with deterministic settings (temperature=0, JSON format). A three-expert consensus HEART score - derived through blinded independent scoring by three emergency medicine physicians with majority-vote adjudication - serves as the reference standard for agreement analysis. Actual 30-day MACE (all-cause death, AMI Type 1/2/4b, unplanned revascularization) determined via national health database and telephone follow-up serves as the outcome for diagnostic accuracy analysis. A secondary documentation-quality sub-study will quantify how spontaneously Turkish emergency anamnesis notes capture HEART score parameters.

Cleft lip and palate surgeries present unique anesthetic challenges due to shared airway access with the surgical field, frequent anatomical abnormalities, and a predominantly infant and toddler population. These factors substantially increase the risk of perioperative respiratory adverse events (PRAEs), including laryngospasm, bronchospasm, desaturation, post-extubation stridor, and unanticipated re-intubation. This prospective single-center observational cohort study aims to determine the true incidence of PRAEs in pediatric patients undergoing elective cleft lip and/or palate repair under general anesthesia, and to identify independent predictive risk factors using standardized airway assessment tools including the Han Mask Ventilation Score and the Intubation Difficulty Score (IDS). No interventions beyond routine clinical practice will be applied. All airway management decisions will remain at the discretion of the attending anesthesiologist.

BACKGROUND: Hemolysis is the most common preanalytical error in emergency department (ED) laboratories, affecting 17-26% of blood samples collected in the ED and leading to test cancellations, repeat venipuncture, delayed diagnoses, and increased healthcare costs. Venous stasis created by tourniquet application during phlebotomy is a recognized contributing factor to hemolysis. While clinical guidelines recommend releasing the tourniquet once blood flow is established, the optimal timing of tourniquet release in relation to tube filling sequence has not been systematically evaluated. OBJECTIVE: The primary objective of this trial is to determine whether early release of sphygmomanometer-applied venous stasis (released after the first tube fills) reduces hemolysis rates compared to late release (released after the last tube fills) during routine phlebotomy in ED patients triaged as green or yellow category. DESIGN: Single-center, parallel-group, superiority randomized controlled trial with 1:1 allocation ratio. The trial was prospectively registered prior to the enrollment of the first participant. PARTICIPANTS: Adult patients (≥18 years) presenting to the emergency department with triage category green (semi-urgent) or yellow (urgent), for whom blood collection is indicated as part of routine clinical care. Patients requiring blood collection from an intravenous catheter, those with known coagulation disorders, and those who decline to participate are excluded. INTERVENTIONS: Group A (Early Release): Sphygmomanometer inflated to 60 mmHg for venous stasis; tourniquet released immediately after the first tube (sodium citrate, blue cap) completes filling. Remaining tubes (SST/gel, yellow cap; K2-EDTA, purple cap) are collected after release. Group B (Late Release): Sphygmomanometer inflated to 60 mmHg; tourniquet maintained throughout all tube filling and released only after the last tube (K2-EDTA, purple cap) completes filling. Tube collection order follows the CLSI H03-A6 standard for both groups. PRIMARY OUTCOME: Hemolysis rate, defined as the proportion of serum separator tube (SST/yellow cap) samples with a Hemolysis Index (HI) ≥ 1+ (corresponding to free hemoglobin ≥50 mg/dL), is assessed by the clinical chemistry laboratory analyzer. The outcome assessor (laboratory technician) is blinded to group assignment. SECONDARY OUTCOMES: (1) Distribution of ordinal hemolysis index categories (-, 1+, 2+, 3+, 4+, 5+) in SST samples; (2) Proportion of hemolyzed samples requiring repeat blood collection; (3) Total blood collection duration (seconds) from sphygmomanometer inflation to last tube filling completion; (4) Complication rate (hematoma, nerve injury, vasovagal reaction, arterial puncture, multiple puncture attempts). SAMPLE SIZE: A total of 792 participants (396 per group) are required based on an assumed hemolysis rate of 12% in the late release group and 6% in the early release group (50% relative risk reduction), α=0.05 (two-tailed), 80% power (Fleiss with pooled variance), plus 10% dropout buffer. RANDOMIZATION: Simple randomization using a computer-generated random number list (randomizer.org). The allocation sequence is maintained by a designated person not involved in enrollment or data collection. Allocation is revealed sequentially at the point of care. STATISTICAL ANALYSIS: Primary analysis: Chi-square test comparing hemolysis rates between groups (intention-to-treat population). Secondary analyses: Mann-Whitney U test for ordinal HI distribution; logistic regression for adjusted odds ratio. Bonferroni correction applied to multiple secondary comparisons (adjusted α = 0.017). Per-protocol analysis performed as a sensitivity analysis. Missing data handled using complete case analysis with sensitivity analysis.

Rapid sequence intubation is commonly performed in the emergency department to secure the airway of critically ill patients. However, blood pressure may decrease during or shortly after intubation because of the patient's underlying illness, the effects of induction medications, and the transition to positive-pressure ventilation. This complication is known as peri-intubation hypotension and is associated with worse clinical outcomes. This study will evaluate whether a preemptive low dose of intravenous adrenaline, also known as epinephrine, can reduce peri-intubation hypotension in adult emergency department patients undergoing rapid sequence intubation. Participants will be randomly assigned to receive either 10 mcg of intravenous adrenaline or 1 mL of 0.9% sodium chloride placebo immediately before induction. The study will compare hemodynamic changes, vasopressor requirements, safety outcomes, and mortality between the two groups.

Peripheral intravenous catheter (PIVC) insertion is one of the most frequently performed invasive procedures in both medical and nursing practice. It is essential for blood sampling, intravenous (IV) fluid resuscitation, medication administration, blood product transfusion, and IV contrast-enhanced radiological imaging. In emergency department (ED) settings, where patients may deteriorate rapidly, timely and adequate IV access is of critical importance. Larger-bore cannulas allow faster fluid and contrast delivery and are less prone to occlusion; however, Larger-bore cannulas are often perceived to cause greater pain and procedural difficulty compared to smaller-bore cannulas. This perception may lead clinicians to prefer smaller cannulas, potentially compromising care quality in patients who may unexpectedly deteriorate. Cannula insertion inherently causes discomfort, pain, and anxiety for patients, and multiple failed attempts further amplify these effects. Patient satisfaction, an important dimension of overall care quality, is directly influenced by pain experience and the number of insertion attempts. Patients requiring two or more attempts for successful cannulation are generally classified as having difficult intravenous access (DIVA). Repeated failed attempts are associated with vascular injury, increased patient distress, delays in diagnosis and treatment, and greater consumption of healthcare resources. The A-DIVA score, developed by Van Loon et al., enables risk stratification for IV access difficulty at triage, classifying patients as low, moderate, or high risk, and facilitating proactive measures such as ultrasound guidance or involvement of experienced personnel for high-risk patients. This single-center, prospective, single-blind, randomized controlled trial aims to compare 18-gauge (18G) and 20-gauge (20G) PIVCs in terms of patient-reported pain intensity and nurse-reported procedural difficulty in adult ED patients with low-to-moderate A-DIVA risk. The investigators hypothesize that 18G cannulas cause similar pain and procedural difficulty compared to 20G cannulas in this patient population. The study will be conducted at Marmara University Pendik Training and Research Hospital Emergency Department. Ethics approval has been granted by the Marmara University Faculty of Medicine Clinical Research Ethics Committee. Adult patients (18 years and older) requiring IV access for diagnostic or therapeutic purposes and classified as low or moderate A-DIVA risk will be eligible. Patients with cognitive impairment, altered mental status, or hemodynamic instability (systolic BP less than 90 mmHg or mean arterial pressure less than 65 mmHg) will be excluded. Eligible consenting patients will be randomized 1:1 using a block randomization method (block size of 4) to either 18G or 20G cannula insertion. Cannula assignment will be delivered to nurses in sealed envelopes. Patients will be blinded to cannula size using blackout goggles. Cannulation will be performed by ED nurses following standard antiseptic and securement procedures, targeting the dorsum of the hand, forearm, or antecubital fossa. If the first attempt fails, at least one additional attempt with the assigned cannula will be required before allowing free choice of cannula and site. The primary outcomes are patient-reported pain intensity and nurse-reported procedural difficulty, each assessed using a 10 cm Visual Analog Scale (VAS) immediately after the first cannulation attempt and compared between the 18G and 20G cannula groups. Secondary outcomes include the comparison of first-attempt cannulation success rate between the 18G and 20G cannula groups. This comparison will also be performed separately according to cannulation site, operator experience, A-DIVA risk category, patient age, and patient sex. A total of 204 patients (102 per group) will be enrolled, based on 90% power, 5% type I error, a minimum clinically important difference of 13 mm on the VAS, and a 10% dropout allowance.

Supraventricular tachycardia (SVT) is a common heart rhythm disorder seen in emergency departments, causing a rapid heartbeat (typically 150-250 beats per minute). The current best non-drug treatment, the modified Valsalva maneuver (mVM), successfully restores normal rhythm in about 43% of cases. When these maneuvers fail, intravenous adenosine is used, which, while effective, can cause brief but distressing side effects such as chest tightness, shortness of breath, and intense anxiety. During the COVID-19 pandemic, some patients briefly fainted during nasal swab collection. This happens because inserting a swab into the back of the nasal cavity (nasopharynx) stimulates the trigeminal nerve, which then activates the vagus nerve and slows the heart - a phenomenon called the trigeminocardiac (or nasocardiac) reflex. One published case report described a patient whose SVT was terminated within 10 seconds using a nasal swab. This study compares the nasal swab technique with the modified Valsalva maneuver in patients presenting to the emergency department with SVT. Patients are randomly assigned to one of two groups. The primary outcome is whether a normal heart rhythm is restored within 1 minute of the procedure. Patient comfort and satisfaction are also measured. The study is conducted in two phases. The first (pilot) phase (30 patients per group) will assess whether the study can be successfully conducted and collect data to finalize the required sample size. The second (main) phase will use the pilot phase's actual data to determine the final number of participants needed.

This prospective observational study aims to evaluate the impact of epidural catheter placement techniques - loss of resistance (LOR) and hanging drop (HD) - on first-attempt success rates and procedural complications among anesthesia residents. Residents with at least two years of training will perform epidural catheterization under supervision for patients scheduled for elective surgery requiring perioperative epidural analgesia. An independent observer will record procedural details, complications, and postoperative pain scores. The study will enroll 440 patients aged 18-80 with ASA I-III classification.

Emergence delirium is a common complication in pediatric patients undergoing general anesthesia. The aim of this study is to investigate the incidence of postoperative awakening delirium and agitation in pediatric patients aged 3-10 years (ASA I-II) using two different target-controlled infusion (TCI) methods (Paedfusor and Eleveld) for TIVA, compared to inhalation anesthesia. This study will examine the effect of two different TCI models on postoperative awakening agitation/delirium by comparing them with each other and with inhalation anesthesia.

Endoscopic submucosal dissection (ESD) has become widely used as a minimally invasive alternative for the resection of early-stage gastrointestinal neoplasms. Due to the lengthy procedure time and intense pain caused by stretching, cutting, and dissecting the gastric wall during ESD, a deeper level of sedation is recommended compared to traditional endoscopic procedures (1). While ensuring adequate patient immobilisation during ESD, preserving respiratory function and rapid recovery are important clinical goals. Total intravenous anaesthesia (TIVA) is an alternative method to inhalation anaesthesia, achieved through the combination of agents such as propofol and remifentanil. TIVA applications can be performed using manual or target-controlled infusion (TCI) systems. TCI systems aim to achieve and maintain the targeted plasma or effect site concentration based on pharmacokinetic models. These systems have been shown to provide advantages such as more stable depth of sedation during endoscopy, less haemodynamic fluctuation, and faster recovery (2-4). Preserving spontaneous breathing is preferred during ESD procedures, which requires careful monitoring of respiration. The Capnostream® device records four variables (SpO₂, RR, non-invasive EtCO₂, heart rate) every 30 seconds via a nasal cannula and integrates them into a single, dimensionless value called the integrated pulmonary index (IPI). The IPI can range from 1 to 10, with 4 and below requiring intervention and 8 to 10 representing the normal range. Furthermore, the use of the bispectral index (BIS) enables objective monitoring of anaesthesia depth by analysing EEG waves and can increase the safety of the recovery process (5). This study aims to compare the effects of manual TIVA and TCI applications on recovery time, BIS, and respiratory parameters during ESD procedures in the stomach or colon region performed under sedation in the endoscopy unit. The findings will contribute to the safer and more effective planning of sedation applications. References; 1. Sasaki T, Tanabe S, Azuma M, Sato A, Naruke A, Ishido K, et al. Propofol sedation with bispectral index monitoring is useful for endoscopic submucosal dissection: a randomised prospective phase II clinical trial. Endoscopy. 2012 Jun;44(6):584-9. 2. Chang YT, Tsai TC, Hsu H, Chen YM, Chi KP, Peng SY. Sedation for gastrointestinal endoscopy with the application of target-controlled infusion. Turk J Gastroenterol Off J Turk Soc Gastroenterol. 2015 Sep;26(5):417-22. 3. Sarraj R, Theiler L, Vakilzadeh N, Krupka N, Wiest R. Propofol sedation in routine endoscopy: A case series comparing target controlled infusion vs manually controlled bolus concept. World J Gastrointest Endosc. 2024 Jan 16;16(1):11-7. 4. García Guzzo ME, Fernandez MS, Sanchez Novas D, Salgado SS, Terrasa SA, Domenech G, et al. Deep sedation using propofol target-controlled infusion for gastrointestinal endoscopic procedures: a retrospective cohort study. BMC Anaesthesiol. 10 August 2020;20(1):195. 5. Sandler NA, Hodges J, Sabino M. Assessment of recovery in patients undergoing intravenous conscious sedation using bispectral analysis. J Oral Maxillofac Surg Off J Am Assoc Oral Maxillofac Surg. 2001 Jun;59(6):603-11; discussion 611-612. 6. Ding Y, White PF. Simplified quality of anaesthesia scoring system. Anaesthesia. 1992 Oct;47(10):906-7.

This clinical study aims to investigate the effects of breathing tubes, which we use to connect children aged 1-11 to a respiratory device under general anesthesia in the operating room, on intracranial pressure after they are inserted. No changes will be made to routine anesthesia practices for this purpose, and eye examinations will be performed with an USG device. This study has no undesirable effects or risks. This study does not include any invasive procedures. You have the right to refuse to participate in the study. You have the right to withdraw from the study at any time. If you decide not to participate in the study for any reason, there will be no disruption to your treatment. You will not be liable for any financial loss if you participate in the study, and you will not be paid. Your identity will be kept confidential if you participate in the study.