Sehgal, Priya

Abstract Background Large language models (LLMs) offer promise for systematic review data extraction, but performance in complex multidisciplinary domains and utility for clinical statement generation remain insufficiently described. Objectives To evaluate Google NotebookLM for AI-assisted data extraction and RAND/UCLA consensus statement generation in a systematic review of IBD, obesity, and cardiometabolic comorbidities. Methods Studies were organized into domain-specific notebooks; structured prompts generated standardized evidence tables. Two independent reviewers validated outputs against full-text articles using a four-category error classification. Cell-level accuracy and critical accuracy (cells free of major factual errors) were the primary metrics; workflow time was compared against a published conventional extraction benchmark. Concordance between AI-generated and expert-finalized statements was assessed. Results Across 57 articles, 1,710 data cells were extracted; 151 (8.83%) were flagged, yielding 91.17% cell-level accuracy. Major factual errors occurred in only 4 cells (0.23%), for a critical accuracy of 99.77%. Most errors were minor omissions (59.6%) or incomplete extractions (30.5%); domain error rates ranged from 7.08% to 11.33%. The pipeline required 17.7 versus a projected 165.1 person-hours (89.3% reduction). PICO-structured prompting generated 70 candidate statements; 58 of 112 finalized panel statements (51.8%) were AI-derived, and 85.7% were retained in the finalized set. Conclusion Google NotebookLM demonstrates feasibility as a primary extraction and synthesis tool in a multidisciplinary systematic review, with extractive incompleteness as the principal limitation and substantial time savings over conventional approaches. Its novel application to RAND/UCLA consensus statement generation extends AI-assisted evidence synthesis to clinical consensus generation workflow.

Abstract Introduction There are limited data on the impact of glucagon-like peptide-1 receptor agonist (GLP-1RA) on the risk of recurrent pouchitis after ileal pouch–anal anastomosis (IPAA) for ulcerative colitis (UC). Methods A retrospective cohort study was conducted in the TriNetX Research Network in adult patients with obesity and IPAA for UC with a history of pouchitis in whom a GLP-1RA was initiated compared to those patients in whom a GLP-1RA was not initiated. The primary outcome was the risk of recurrent pouchitis within 12 months. The secondary outcomes were mean number of antibiotic prescriptions, anti-diarrheal medication use, Crohns-like disease of pouch (CLDP), and pouch excision. One-to-one propensity score matching was performed. Results We identified 43 UC-IPAA patients with obesity and history of pouchitis (mean age 51.7 ± 14.7 years, mean body mass index [BMI] 34.4 ± 6 kg/m2, 60.4% female). After propensity score matching, the GLP-1RA cohort had a significantly decreased risk of recurrent pouchitis (26.3% vs 52.6%; adjusted odds ratio [aOR], 0.32; 95% CI, 0.12-0.84) compared to the control cohort. Mean antibiotic prescriptions were numerically lower in the GLP-1RA cohort (2 vs 4.4, P = .16). Anti-diarrheal medication use was lower in the GLP-1RA cohort (18.4% vs 47.3%; aOR, 0.23; 95% CI, 0.07-0.67). There was no difference in the risk of CLDP. Zero patients required pouch excision in both cohorts. Conclusion Our study in a small cohort of patients with UC-IPAA and obesity with a history of pouchitis showed that GLP-1RA use was associated with decreased risk of recurrent pouchitis and anti-diarrheal medication use.

ABSTRACT Background Obesity has been linked to a more severe phenotype in patients with ulcerative colitis (UC). Aim To evaluate the impact of obesity on outcomes of advanced therapies in UC. Methods We conducted a retrospective cohort study utilising the TriNetX database comparing the composite score of corticosteroid use, change in advanced therapy or colectomy within two years between two cohorts of patients with UC—those with obesity (BMI ≥ 30 kg/m 2 ) and those without (BMI 18.5–24.9). The risk assessment was stratified to specific advanced therapies, including tumour necrosis factor α inhibitors (TNFi), vedolizumab, ustekinumab and Janus kinase inhibitors (JAKi). We performed 1:1 propensity score matching (PSM) for demographics, co‐morbid conditions, laboratory values and IBD medications including corticosteroids. Results There were 3904, 2025, 1150 and 477 patients on TNFi, vedolizumab, ustekinumab and JAKi, respectively, in the UC obesity cohort. After PSM, the UC obesity cohort was at an increased risk of the composite outcome of corticosteroid use, change in therapy and colectomy compared to the UC control cohort in patients on TNFi (aHR 1.37, 95% CI 1.29–1.49), vedolizumab (aHR 1.29, 95% CI 1.16–1.43), ustekinumab (aHR 1.1.26, 95% CI 1.10–1.44) and JAKi (aHR 1.38, 95% CI 1.13–1.69). Sub‐group analysis based on the specific TNFi also showed an increased risk of composite outcome for infliximab (aHR 1.36, 95% CI 1.22–1.52) and adalimumab (aHR 1.26, 95% CI 1.11–1.42) within 2 years. Conclusions Obesity is associated with lower efficacy of several advanced therapies in patients with UC.