Rashid Latif Medical College

Acute ischemic stroke (AIS) is a leading cause of morbidity and mortality worldwide. While alteplase has been widely used for acute management, recent clinical trials suggest that tenecteplase (TNK) may offer improved clinical outcomes. This study aims to compare the efficacy and safety of TNK compared with alteplase.

A comprehensive literature search was conducted using PubMed, Embase and Cochrane Library from inception to October 2024 to identify randomized controlled trials that compared TNK at 0.25 mg/kg dosage with alteplase. Data about clinical outcomes was extracted from both groups and assessed by generating forest plots using the random-effects model and pooling odds ratios (ORs).

A total of 11 RCTs with 7,546 patients were included in the analysis. TNK showed statistically significant improvement in excellent functional outcome (mRS 0-1) compared with alteplase (OR= 1.14, 95 % CI= 1.03-1.25). No statistically significant difference was observed for good functional outcome (mRS 0-2) (OR= 1.11, 95 % CI= 0.9-1.25), early neurological improvement (OR=1.08, 95 % CI= 0.93-1.26), all-cause death (OR=0.99, 95 % CI= 0.81-1.19), symptomatic intracranial hemorrhage (OR=1.11, 95 % CI=0.81-1.52) and poor functional outcome (mRS=4-6) (OR=0.95, 95 % CI=0.79-1.14).

In patients with acute ischemic stroke, TNK demonstrated a significant advantage over alteplase in achieving excellent functional outcomes. The incidence of early neurological improvement, symptomatic intracranial hemorrhage, all-cause death, and poor functional outcome remained comparable across the two groups.

Acute pancreatitis (AP) is a common and potentially life-threatening condition, with gallstones being a major etiology.

This study aims to determine how common AP is among patients with gallstones and to identify the associated risk factors, diagnostic laboratory markers, and potential complications.

This cross-sectional study was conducted at Shalamar Hospital, Lahore, a tertiary care center, from June 2024 to January 2025. Data were collected from 165 patients presenting in the Department of Gastroenterology and General Surgery. Patients were included if they had gallstones confirmed by abdominal ultrasound and met the revised Atlanta criteria for AP, including clinical symptoms (e.g., epigastric pain, nausea) and laboratory findings (serum lipase/amylase ≥3× upper limit of normal).

The prevalence of AP in patients with gallstones was 37.6% (62/165). Risk factors such as obesity (72%, 45/62), age (≥50 years, 72%, 45/62), and female gender (65%, 40/62) were significantly associated with the condition. Elevated serum amylase (225.4 ± 112.3 U/L) and lipase (317.8 ± 150.7 U/L), along with increased white blood cell (WBC) count (13.2 ± 4.6 x 10⁹/L), C-reactive protein (CRP) (45.7 ± 19.2 mg/L), and bilirubin levels (2.4 ± 1.6 mg/dL), were observed, particularly in severe cases of AP. Pancreatic necrosis occurred in 13% (8/62), organ failure in 8% (5/62), and systemic inflammatory response syndrome (SIRS) in 16% (10/62). Cholecystectomy was performed in 69% (43/62) of cases, and the mortality rate was 3.2% (2/62).

Gallstones are a major cause of AP, with obesity, age, and female gender being key risk factors. Elevated laboratory markers correlate with the severity of the condition and can aid in disease assessment.

Data extraction from the published literature is the most laborious step in conducting living systematic reviews (LSRs). We aim to build a generalizable, automated data extraction workflow leveraging large language models (LLMs) that mimics the real-world 2-reviewer process.

A dataset of 10 trials (22 publications) from a published LSR was used, focusing on 23 variables related to trial, population, and outcomes data. The dataset was split into prompt development (n = 5) and held-out test sets (n = 17). GPT-4-turbo and Claude-3-Opus were used for data extraction. Responses from the 2 LLMs were considered concordant if they were the same for a given variable. The discordant responses from each LLM were provided to the other LLM for cross-critique. Accuracy, ie, the total number of correct responses divided by the total number of responses, was computed to assess performance.

In the prompt development set, 110 (96%) responses were concordant, achieving an accuracy of 0.99 against the gold standard. In the test set, 342 (87%) responses were concordant. The accuracy of the concordant responses was 0.94. The accuracy of the discordant responses was 0.41 for GPT-4-turbo and 0.50 for Claude-3-Opus. Of the 49 discordant responses, 25 (51%) became concordant after cross-critique, increasing accuracy to 0.76.

Concordant responses by the LLMs are likely to be accurate. In instances of discordant responses, cross-critique can further increase the accuracy.

Large language models, when simulated in a collaborative, 2-reviewer workflow, can extract data with reasonable performance, enabling truly “living” systematic reviews.