{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,17]],"date-time":"2026-10-17T00:11:13Z","timestamp":1792195873549,"version":"4.4.0"},"reference-count":18,"publisher":"PeerJ","license":[{"start":{"date-parts":[[2025,4,30]],"date-time":"2025-04-30T00:00:00Z","timestamp":1745971200000},"content-version":"unspecified","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"abstract":"<jats:sec>\n<jats:title>Background<\/jats:title>\n<jats:p>Large language models (LLMs) offer a potential solution to the labor-intensive nature of systematic reviews. This study evaluated the ability of the GPT model to identify articles that discuss perioperative risk factors for esophagectomy complications. To test the performance of the model, we tested GPT-4 on narrower inclusion criterion and by assessing its ability to discriminate relevant articles that solely identified preoperative risk factors for esophagectomy.<\/jats:p>\n<\/jats:sec>\n<jats:sec>\n<jats:title>Methods<\/jats:title>\n<jats:p>A literature search was run by a trained librarian to identify studies (<jats:italic>n<\/jats:italic> = 1,967) discussing risk factors to esophagectomy complications. The articles underwent title and abstract screening by three independent human reviewers and GPT-4. The Python script used for the analysis made Application Programming Interface (API) calls to GPT-4 with screening criteria in natural language. GPT-4\u2019s inclusion and exclusion decision were compared to those decided human reviewers.<\/jats:p>\n<\/jats:sec>\n<jats:sec>\n<jats:title>Results<\/jats:title>\n<jats:p>The agreement between the GPT model and human decision was 85.58% for perioperative factors and 78.75% for preoperative factors. The AUC value was 0.87 and 0.75 for the perioperative and preoperative risk factors query, respectively. In the evaluation of perioperative risk factors, the GPT model demonstrated a high recall for included studies at 89%, a positive predictive value of 74%, and a negative predictive value of 84%, with a low false positive rate of 6% and a macro-F1 score of 0.81. For preoperative risk factors, the model showed a recall of 67% for included studies, a positive predictive value of 65%, and a negative predictive value of 85%, with a false positive rate of 15% and a macro-F1 score of 0.66. The interobserver reliability was substantial, with a kappa score of 0.69 for perioperative factors and 0.61 for preoperative factors. Despite lower accuracy under more stringent criteria, the GPT model proved valuable in streamlining the systematic review workflow. Preliminary evaluation of inclusion and exclusion justification provided by the GPT model were reported to have been useful by study screeners, especially in resolving discrepancies during title and abstract screening.<\/jats:p>\n<\/jats:sec>\n<jats:sec>\n<jats:title>Conclusion<\/jats:title>\n<jats:p>This study demonstrates promising use of LLMs to streamline the workflow of systematic reviews. The integration of LLMs in systematic reviews could lead to significant time and cost savings, however caution must be taken for reviews involving stringent a narrower and exclusion criterion. Future research is needed and should explore integrating LLMs in other steps of the systematic review, such as full text screening or data extraction, and compare different LLMs for their effectiveness in various types of systematic reviews.<\/jats:p>\n<\/jats:sec>","DOI":"10.7717\/peerj-cs.2822","type":"journal-article","created":{"date-parts":[[2025,4,30]],"date-time":"2025-04-30T08:27:11Z","timestamp":1746001631000},"page":"e2822","source":"Crossref","is-referenced-by-count":5,"title":["Validation of automated paper screening for esophagectomy systematic review using large language models"],"prefix":"10.7717","volume":"11","author":[{"given":"Rashi","family":"Ramchandani","sequence":"first","affiliation":[{"name":"Department of Medicine, University of Ottawa, Ottawa, Ontario, Canada"},{"name":"Institute of Health Policy, Management and Evaluation, University of Toronto, Toronto, Ontario, Canada"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Eddie","family":"Guo","sequence":"additional","affiliation":[{"name":"Cumming School of Medicine, University of Calgary, Calgary, Alberta, Canada"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Esra","family":"Rakab","sequence":"additional","affiliation":[{"name":"Department of Medicine, University of Ottawa, Ottawa, Ontario, Canada"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jharna","family":"Rathod","sequence":"additional","affiliation":[{"name":"Department of Medicine, University of Ottawa, Ottawa, Ontario, Canada"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4023-5715","authenticated-orcid":true,"given":"Jamie","family":"Strain","sequence":"additional","affiliation":[{"name":"Ottawa Hospital Research Institute, Ottawa, Ontario, Canada"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"William","family":"Klement","sequence":"additional","affiliation":[{"name":"Ottawa Hospital Research Institute, Ottawa, Ontario, Canada"},{"name":"Faculty of Computer Sciences, Dalhousie University, Dalhousie, Halifax, Canada"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Risa","family":"Shorr","sequence":"additional","affiliation":[{"name":"Library and Learning Services, The Ottawa Hospital, Ottawa, Ontario, Canada"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-5418-814X","authenticated-orcid":true,"given":"Erin","family":"Williams","sequence":"additional","affiliation":[{"name":"Division of General Surgery, Department of Surgery, The Ottawa Hospital, Ottawa, Ontario, Canada"},{"name":"Division of Thoracic Surgery, Department of Surgery, The Ottawa Hospital, Ottawa, Ontario, Canada"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Daniel","family":"Jones","sequence":"additional","affiliation":[{"name":"Division of General Surgery, Department of Surgery, The Ottawa Hospital, Ottawa, Ontario, Canada"},{"name":"Division of Thoracic Surgery, Department of Surgery, The Ottawa Hospital, Ottawa, Ontario, Canada"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sebastien","family":"Gilbert","sequence":"additional","affiliation":[{"name":"Division of General Surgery, Department of Surgery, The Ottawa Hospital, Ottawa, Ontario, Canada"},{"name":"Division of Thoracic Surgery, Department of Surgery, The Ottawa Hospital, Ottawa, Ontario, Canada"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"4443","published-online":{"date-parts":[[2025,4,30]]},"reference":[{"issue":"3","key":"10.7717\/peerj-cs.2822\/ref-1","doi-asserted-by":"publisher","first-page":"1565","DOI":"10.1007\/s00405-023-08337-7","article-title":"How ChatGPT works: a mini review","volume":"281","author":"Briganti","year":"2024","journal-title":"European Archives of Oto-Rhino-Laryngology: Official Journal of the European Federation of Oto-Rhino-Laryngological Societies (EUFOS): Affiliated with the German Society for Oto-Rhino-Laryngology\u2014Head and Neck Surgery"},{"issue":"1","key":"10.7717\/peerj-cs.2822\/ref-2","doi-asserted-by":"publisher","first-page":"93","DOI":"10.1186\/s13643-021-01635-3","article-title":"Research Screener: a machine learning tool to semi-automate abstract screening for systematic reviews","volume":"10","author":"Chai","year":"2021","journal-title":"Systematic Reviews"},{"issue":"2","key":"10.7717\/peerj-cs.2822\/ref-3","doi-asserted-by":"publisher","first-page":"e24418","DOI":"10.2196\/24418","article-title":"The impact of systematic review automation tools on methodological quality and time taken to complete systematic review tasks: case study","volume":"7","author":"Clark","year":"2021","journal-title":"JMIR Medical Education"},{"issue":"6","key":"10.7717\/peerj-cs.2822\/ref-4","doi-asserted-by":"publisher","first-page":"1534","DOI":"10.1007\/s10439-023-03336-5","article-title":"Role of ChatGPT-4 for medical researchers","volume":"52","author":"Datt","year":"2024","journal-title":"Annals of Biomedical Engineering"},{"issue":"1","key":"10.7717\/peerj-cs.2822\/ref-5","doi-asserted-by":"publisher","first-page":"45","DOI":"10.1186\/s13643-018-0707-8","article-title":"Technology-assisted title and abstract screening for systematic reviews: a retrospective evaluation of the Abstrackr machine learning tool","volume":"7","author":"Gates","year":"2018","journal-title":"Systematic Reviews"},{"key":"10.7717\/peerj-cs.2822\/ref-6","doi-asserted-by":"publisher","first-page":"e48996","DOI":"10.2196\/48996","article-title":"Automated paper screening for clinical reviews using large language models: data analysis study","volume":"26","author":"Guo","year":"2024","journal-title":"Journal of Medical Internet Research"},{"issue":"1","key":"10.7717\/peerj-cs.2822\/ref-7","doi-asserted-by":"publisher","first-page":"285","DOI":"10.1186\/s12874-021-01451-2","article-title":"Guidance for using artificial intelligence for title and abstract screening while conducting knowledge syntheses","volume":"21","author":"Hamel","year":"2021","journal-title":"BMC Medical Research Methodology"},{"issue":"1","key":"10.7717\/peerj-cs.2822\/ref-8","doi-asserted-by":"publisher","first-page":"256","DOI":"10.1186\/s12874-020-01129-1","article-title":"An evaluation of DistillerSR\u2019s machine learning-based prioritization tool for title\/abstract screening\u2014impact on reviewer-relevant outcomes","volume":"20","author":"Hamel","year":"2020","journal-title":"BMC Medical Research Methodology"},{"issue":"11","key":"10.7717\/peerj-cs.2822\/ref-9","doi-asserted-by":"publisher","first-page":"2761","DOI":"10.1177\/0363546513497567","article-title":"How to write a systematic review","volume":"42","author":"Harris","year":"2014","journal-title":"The American Journal of Sports Medicine"},{"issue":"3","key":"10.7717\/peerj-cs.2822\/ref-10","doi-asserted-by":"publisher","first-page":"118","DOI":"10.1177\/014107680309600304","article-title":"Five steps to conducting a systematic review","volume":"96","author":"Khan","year":"2003","journal-title":"Journal of the Royal Society of Medicine"},{"issue":"9734","key":"10.7717\/peerj-cs.2822\/ref-11","doi-asserted-by":"publisher","first-page":"287","DOI":"10.1016\/j.jclinepi.2021.05.019","article-title":"Resource use during systematic review production varies widely: a scoping review","volume":"139","author":"Nussbaumer-Streit","year":"2021","journal-title":"Journal of Clinical Epidemiology"},{"key":"10.7717\/peerj-cs.2822\/ref-12","doi-asserted-by":"publisher","first-page":"n71","DOI":"10.1136\/bmj.n71","article-title":"The PRISMA, 2020 statement: an updated guideline for reporting systematic reviews","volume":"372","author":"Page","year":"2021","journal-title":"BMJ (Clinical Research Ed.)"},{"issue":"3","key":"10.7717\/peerj-cs.2822\/ref-13","doi-asserted-by":"publisher","first-page":"330","DOI":"10.1002\/jrsm.1354","article-title":"Best practice guidelines for abstract screening large-evidence systematic reviews and meta-analyses","volume":"10","author":"Polanin","year":"2019","journal-title":"Research Synthesis Methods"},{"issue":"1","key":"10.7717\/peerj-cs.2822\/ref-14","doi-asserted-by":"publisher","first-page":"46","DOI":"10.1186\/s41182-019-0165-6","article-title":"A step by step guide for conducting a systematic review and meta-analysis with simulation data","volume":"47","author":"Tawfik","year":"2019","journal-title":"Tropical Medicine and Health"},{"issue":"6","key":"10.7717\/peerj-cs.2822\/ref-15","doi-asserted-by":"publisher","first-page":"791","DOI":"10.7326\/M23-3389","article-title":"Sensitivity and specificity of using GPT-3.5 turbo models for title and abstract screening in systematic reviews and meta-analyses","volume":"177","author":"Tran","year":"2024","journal-title":"Annals of Internal Medicine"},{"issue":"7","key":"10.7717\/peerj-cs.2822\/ref-16","doi-asserted-by":"publisher","first-page":"700","DOI":"10.1111\/1759-7714.14795","article-title":"Comparison of neoadjuvant nab-paclitaxel plus immunotherapy versus paclitaxel plus immunotherapy for esophageal squamous cell carcinoma","volume":"14","author":"Yang","year":"2023","journal-title":"Thoracic Cancer"},{"issue":"1","key":"10.7717\/peerj-cs.2822\/ref-17","doi-asserted-by":"publisher","first-page":"189","DOI":"10.1097\/XCS.0000000000000454","article-title":"Omental flap wrapping around the esophagogastric anastomosis and association with anastomotic leak in esophagectomy for esophageal cancer: a propensity score-matching analysis","volume":"236","author":"Yoshida","year":"2023","journal-title":"Journal of the American College of Surgeons"},{"issue":"1","key":"10.7717\/peerj-cs.2822\/ref-18","doi-asserted-by":"publisher","first-page":"114","DOI":"10.3390\/cells12010114","article-title":"Research progress on the predicting factors and coping strategies for postoperative recurrence of esophageal cancer","volume":"12","author":"Zhang","year":"2022","journal-title":"Cells"}],"container-title":["PeerJ Computer Science"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/peerj.com\/articles\/cs-2822.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/peerj.com\/articles\/cs-2822.xml","content-type":"application\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/peerj.com\/articles\/cs-2822.html","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/peerj.com\/articles\/cs-2822.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,4,30]],"date-time":"2025-04-30T08:27:14Z","timestamp":1746001634000},"score":1,"resource":{"primary":{"URL":"https:\/\/peerj.com\/articles\/cs-2822"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,4,30]]},"references-count":18,"alternative-id":["10.7717\/peerj-cs.2822"],"URL":"https:\/\/doi.org\/10.7717\/peerj-cs.2822","archive":["CLOCKSS","LOCKSS","Portico"],"relation":{},"ISSN":["2376-5992"],"issn-type":[{"value":"2376-5992","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,4,30]]},"article-number":"e2822"}}