Coombs, Shannon

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.

Background: Home telemetry has been shown to reduce 30-day readmissions in patients at high risk for readmission. It remains unclear whether this finding is due to the telemetry itself or the nurse contact that follows detection of an abnormal parameter. Purpose: This study was a subgroup analysis from a prior trial comparing patients on home telemetry for 30 days to usual care after discharge to determine if nurse contact affected readmission rates. Methods: Patients randomized to telemetry who had at least 1 nurse contact were compared to those who had none to see if the contact reduced the rate of readmissions. Results: Of 449 home telemetry patients, 292 (65.0%) received 1 or more nurse contacts. Single and multivariable logistic regression models exploring the association of contact with 30-day readmission did not find any statistically significant associations of nurse contact with readmission (all unadjusted P ≥ .13, all adjusted P ≥ .18). Conclusion: Our subset analysis did not show that the nurse contact was the factor that reduced readmission in patients on home telemetry.

Abstract Background: Home telemonitoring has been shown to reduce 30-day readmissions patients at high risk for return to the hospital. The reason for this remains unclear. The aim of this study was to explore the association of nurse contact with 30-day hospital readmission in these high risk patients. Methods: This study was a subgroup analysis from a prior randomized control comparing home telemonitoring for readmission for 30 days to usual care at discharge. Patients randomized to the home telemonitoring who had vital signs outside a preset range, triggered an alert flag for the monitoring nurse, which generated a contact to the patient. We compared patients who had at least one nurse contact to those who had none to see if the contact reduced the rate of readmissions in patients being monitored. Results: Of 449 patients, 292 (65.0%) received 1 or more telemonitoring flags. Single variable and multivariable logistic regression models exploring the association of flag type with hospital 30-day readmission did not find any statistically significant associations of telemonitoring flags with hospital 30-day readmission (all unadjusted P≥0.13, all adjusted P≥0.18). Conclusion: While post-hospitalization home telemetry has been shown to reduce 30-day readmission, our subset analysis did not show that the nurse contact was the factor that reduced readmission in these patients. Trial registration: ClinicalTrials.gov identifier: NCT02136186, registration date May 12, 2014.