Doi, Suhail A

Objective To examine the potential errors of a general large language model (LLM) (ie, Claude 3.5 Sonnet) on data extraction from randomised controlled trials (RCTs). Design and setting An empirical study comparing Claude 3.5 Sonnet extractions against a human-performed verification dataset. The extraction tasks for Claude 3.5 Sonnet were based solely on original RCT portable document format (PDF) files. For PDFs that could not be directly extracted by Claude 3.5 Sonnet, optical character recognition was employed to convert them into text format before extraction. Participants A random sample of 664 trials was selected from a well-established trial bank and a final data pool was established based on rigorous manual cross-checking as a reference standard. Data sources PubMed, EMBASE, Scopus, Web of Science (all databases) and the Cochrane Central Register of Controlled Trials (CENTRAL) up to February 2023. Eligibility criteria for selecting studies RCTs on children involving medication and adverse events. Main outcome measures Claude 3.5 Sonnet was applied to extract the basic information (eg, trial design, population information and source of funding) and adverse outcomes (ie, name of adverse events, number of events). Claude 3.5 Sonnet outputs were compared against the final data pool and all errors were recorded. Results are presented as error rates and with 95% CI, estimated using a generalised linear mixed model. Results For the 664 trials, a total of 23 069 data cells were extracted via Claude 3.5 Sonnet, with 10 624 for basic information and 12 445 for adverse outcomes. The overall error rate for data extraction was 6.6% (95% CI 5.4% to 8.2%), with 5.7% (95% CI 5.2% to 6.1%) in basic information and 7.6% (95% CI 4.9% to 11.8%) in adverse outcomes. When stratified the 1542 total errors by error types, misallocation (assigning data to incorrect fields; 57.1%, 881/1542) and missed or omitted data (incomplete extraction of available data; 23.2%, 357/1542) accounted for the two most frequent errors, with misallocation occurring more in basic information (53.3%, 470/881), while missed or omitted data occurred more in adverse outcomes (96.1%, 343/357). Post hoc analysis examining the association between trial reporting quality (assessed using Consolidated Standards of Reporting Trials (CONSORT) 2025 and LLM data extraction error rates indicated that higher CONSORT adherence was associated with lower extraction error rates. Conclusions The data extraction error of Claude was relatively low, but it alerts LLM applications in evidence synthesis. Detailed checking for LLM outputs should be the primary consideration for evidence synthesisers.

Introduction Increasing publication of fraudulent clinical trials poses a serious threat to evidence-based medicine. In the VITALITY Study I, we demonstrated that contamination by retracted trials significantly distorts evidence synthesis. The upcoming VITALITY Study II will take a step further and investigate how such problematic evidence undermines the credibility of guideline recommendations. Methods and analyses The VITALITY Study II will adopt a retrospective cohort design and will be structured as three work packages (WPs). To start with, a cohort of clinical practice guidelines (CPGs) that were contaminated by retracted trials and/or meta-analyses that included retracted trials will be established based on forward citation searching (WP1). Then, recommendations from these CPGs that used evidence from retracted trials and/or meta-analyses that synthesised these retracted trials will be re-evaluated after removing such problematic evidence, in terms of both the direction and strength of effect sizes (WP2). Finally, the subsequent impact on patient outcomes attributable to distorted recommendations will be estimated on a hypothetical population, measured by the number of expected benefit loss and the number of expected harm increment per 100 000 patients (WP3). Ethics and dissemination Formal ethical approval is not necessary for this study as it does not involve human or animal participants, nor does it involve the collection of primary data. We will disseminate the findings through publication in peer-reviewed journals and, whenever possible, presentations at academic conferences.