{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,30]],"date-time":"2026-07-30T06:53:09Z","timestamp":1785394389574,"version":"3.55.0"},"reference-count":31,"publisher":"JMIR Publications Inc.","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["JMIR Cancer"],"abstract":"Abstract<\/jats:title>\n \n Background<\/jats:title>\n Randomized controlled trials (RCTs) are the gold standard for evaluating interventions in oncology, but reporting can be subject to \u201cspin\u201d\u2014presenting results in ways that mislead readers about true efficacy.<\/jats:p>\n <\/jats:sec>\n \n Objective<\/jats:title>\n This study aimed to investigate whether large language models (LLMs) could provide a standardized approach to detect spin, particularly in the conclusions, where it most commonly occurs.<\/jats:p>\n <\/jats:sec>\n \n Methods<\/jats:title>\n \n We randomly sampled 250 two-arm, single\u2013primary end point oncology RCTs from 7 major medical journals published between 2005 and 2023. Two authors independently annotated trials as positive or negative based on whether they met their primary end point. Three commercial LLMs (GPT-3.5 Turbo, GPT-4o, and GPT-o1) were tasked with classifying trials as positive or negative when provided with (1) conclusions only; (2) methods and conclusions; (3) methods, results, and conclusions; or (4) title and full abstract. LLM performance was evaluated against human annotations. Afterward, trials incorrectly classified as positive when the model was provided only with the conclusions but correctly classified as negative when provided with the whole abstract were analyzed for patterns that may indicate the presence of spin. Model performance was assessed using accuracy, precision, recall, and\n F<\/jats:italic>\n 1<\/jats:sub>\n -score calculated from confusion matrices.\n <\/jats:p>\n <\/jats:sec>\n \n Results<\/jats:title>\n \n Of the 250 trials, 146 (58.4%) were positive, and 104 (41.6%) were negative. The GPT-o1 model demonstrated the highest performance across all conditions, with\n F<\/jats:italic>\n 1<\/jats:sub>\n -scores of 0.932 (conclusions only; 95% CI 0.90-0.96), 0.96 (methods and conclusions; 95% CI 0.93-0.98), 0.98 (methods, results, and conclusions; 95% CI 0.96-0.99), and 0.97 (title and abstract; 95% CI 0.95-0.99). Analysis of trials incorrectly classified as positive when the model was provided only with the conclusions revealed shared patterns, including absence of primary end point results, emphasis on subgroup improvements, or unclear distinction between primary and secondary end points. These patterns were almost never found in trials correctly classified as negative.\n <\/jats:p>\n <\/jats:sec>\n \n Conclusions<\/jats:title>\n LLMs can effectively detect potential spin in oncology RCT reporting by identifying discrepancies between how trials are presented in the conclusions vs the full abstracts. This approach could serve as a supplementary tool for improving transparency in scientific reporting, although further development is needed to address more complex trial designs beyond those examined in this feasibility study.<\/jats:p>\n <\/jats:sec>","DOI":"10.2196\/78221","type":"journal-article","created":{"date-parts":[[2026,1,21]],"date-time":"2026-01-21T20:21:01Z","timestamp":1769026861000},"page":"e78221-e78221","source":"Crossref","is-referenced-by-count":3,"title":["Large Language Models for Supporting Clear Writing and Detecting Spin in Randomized Controlled Trials in Oncology: Comparative Analysis of GPT Models and Prompts"],"prefix":"10.2196","volume":"12","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-9448-0786","authenticated-orcid":false,"given":"Carole","family":"Koechli","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-5374-8720","authenticated-orcid":false,"given":"Fabio","family":"Dennst\u00e4dt","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4168-2795","authenticated-orcid":false,"given":"Christina","family":"Schr\u00f6der","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-9493-3834","authenticated-orcid":false,"given":"Daniel M","family":"Aebersold","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-7664-9207","authenticated-orcid":false,"given":"Robert","family":"F\u00f6rster","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-4359-7735","authenticated-orcid":false,"given":"Daniel R","family":"Zwahlen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-1040-4888","authenticated-orcid":false,"given":"Paul","family":"Windisch","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1010","published-online":{"date-parts":[[2026,1,21]]},"reference":[{"issue":"33","key":"R1","doi-asserted-by":"publisher","first-page":"5458","DOI":"10.1200\/JCO.2008.16.5456","article-title":"Evolution of the randomized controlled trial in oncology over three decades","volume":"26","author":"Booth","journal-title":"J Clin Oncol"},{"issue":"6","key":"R2","doi-asserted-by":"publisher","first-page":"201","DOI":"10.1136\/ebmed-2016-110570","article-title":"\u201cSpin\u201d in reports of clinical research","volume":"21","author":"Mahtani","journal-title":"Evid Based Med"},{"issue":"36","key":"R3","doi-asserted-by":"publisher","first-page":"4120","DOI":"10.1200\/JCO.2014.56.7503","article-title":"Impact of spin in the abstracts of articles reporting results of randomized controlled trials in the field of cancer: the SPIIN randomized controlled trial","volume":"32","author":"Boutron","journal-title":"J Clin Oncol"},{"issue":"12","key":"R4","doi-asserted-by":"publisher","first-page":"e2135765","DOI":"10.1001\/jamanetworkopen.2021.35765","article-title":"Misleading reporting (spin) in noninferiority randomized clinical trials in oncology with statistically not significant results: a systematic review","volume":"4","author":"Ito","journal-title":"JAMA Netw Open"},{"key":"R5","doi-asserted-by":"publisher","DOI":"10.1016\/j.critrevonc.2019.102821","article-title":"Evaluation of spin in oncology clinical trials","volume":"144","author":"Wayant","journal-title":"Crit Rev Oncol Hematol"},{"issue":"12","key":"R6","doi-asserted-by":"publisher","DOI":"10.7759\/cureus.75748","article-title":"The impact of temperature on extracting information from clinical trial publications using large language models","volume":"16","author":"Windisch","journal-title":"Cureus"},{"issue":"7","key":"R7","doi-asserted-by":"publisher","first-page":"706","DOI":"10.1200\/JCO.2015.62.1474","article-title":"Volasertib versus chemotherapy in platinum-resistant or -refractory ovarian cancer: a randomized phase II Groupe des Investigateurs Nationaux pour l\u2019Etude des Cancers de l\u2019Ovaire Study","volume":"34","author":"Pujade-Lauraine","journal-title":"J Clin Oncol"},{"issue":"3","key":"R8","doi-asserted-by":"publisher","first-page":"301","DOI":"10.1200\/JCO.2012.44.7920","article-title":"Results of a randomized trial of chlorambucil versus fludarabine for patients with untreated Waldenstr\u00f6m macroglobulinemia, marginal zone lymphoma, or lymphoplasmacytic lymphoma","volume":"31","author":"Leblond","journal-title":"J Clin Oncol"},{"issue":"14","key":"R9","doi-asserted-by":"publisher","first-page":"2607","DOI":"10.1200\/JCO.22.01805","article-title":"Bortezomib-dexamethasone, rituximab, and cyclophosphamide as first-line treatment for Waldenstr\u00f6m\u2019s macroglobulinemia: a prospectively randomized trial of the European Consortium for Waldenstr\u00f6m\u2019s macroglobulinemia","volume":"41","author":"Buske","journal-title":"J Clin Oncol"},{"issue":"16","key":"R10","doi-asserted-by":"publisher","first-page":"3697","DOI":"10.1200\/JCO.2005.05.112","article-title":"Addition of bevacizumab to bolus fluorouracil and leucovorin in first-line metastatic colorectal cancer: results of a randomized phase II trial","volume":"23","author":"Kabbinavar","journal-title":"J Clin Oncol"},{"issue":"8","key":"R11","doi-asserted-by":"publisher","first-page":"786","DOI":"10.1200\/JCO.2015.62.4734","article-title":"Efficacy and safety of trabectedin or dacarbazine for metastatic liposarcoma or leiomyosarcoma after failure of conventional chemotherapy: results of a phase III randomized multicenter clinical trial","volume":"34","author":"Demetri","journal-title":"J Clin Oncol"},{"issue":"9762","key":"R12","doi-asserted-by":"publisher","first-page":"321","DOI":"10.1016\/S0140-6736(10)62312-4","article-title":"Adjuvant tamoxifen and exemestane in early breast cancer (TEAM): a randomised phase 3 trial","volume":"377","author":"van de Velde","journal-title":"Lancet"},{"issue":"29","key":"R13","doi-asserted-by":"publisher","first-page":"4738","DOI":"10.1200\/JCO.2006.06.0483","article-title":"Bcl-2 antisense (oblimersen sodium) plus dacarbazine in patients with advanced melanoma: the Oblimersen Melanoma Study Group","volume":"24","author":"Bedikian","journal-title":"J Clin Oncol"},{"issue":"36","key":"R14","doi-asserted-by":"publisher","first-page":"9227","DOI":"10.1200\/JCO.2005.03.0551","article-title":"High-dose therapy and autologous blood stem-cell transplantation compared with conventional treatment in myeloma patients aged 55 to 65 years: long-term results of a randomized control trial from the Group Myelome-Autogreffe","volume":"23","author":"Fermand","journal-title":"J Clin Oncol"},{"issue":"4","key":"R15","doi-asserted-by":"publisher","first-page":"295","DOI":"10.1200\/JCO.20.02529","article-title":"Total body irradiation or chemotherapy conditioning in childhood ALL: a multinational, randomized, noninferiority phase III study","volume":"39","author":"Peters","journal-title":"J Clin Oncol"},{"issue":"1","key":"R16","doi-asserted-by":"publisher","first-page":"114","DOI":"10.1016\/S1470-2045(13)70539-4","article-title":"Oral ibandronic acid versus intravenous zoledronic acid in treatment of bone metastases from breast cancer: a randomised, open label, non-inferiority phase 3 trial","volume":"15","author":"Barrett-Lee","journal-title":"Lancet Oncol"},{"issue":"34","key":"R17","doi-asserted-by":"publisher","first-page":"4039","DOI":"10.1200\/JCO.2015.61.4578","article-title":"Randomized, double-blind, phase II study of ruxolitinib or placebo in combination with capecitabine in patients with metastatic pancreatic cancer for whom therapy with gemcitabine has failed","volume":"33","author":"Hurwitz","journal-title":"J Clin Oncol"},{"issue":"6","key":"R18","doi-asserted-by":"publisher","first-page":"799","DOI":"10.1016\/S1470-2045(18)30193-1","article-title":"NGR-hTNF in combination with best investigator choice in previously treated malignant pleural mesothelioma (NGR015): a randomised, double-blind, placebo-controlled phase 3 trial","volume":"19","author":"Gregorc","journal-title":"Lancet Oncol"},{"issue":"19","key":"R19","doi-asserted-by":"publisher","first-page":"3089","DOI":"10.1200\/JCO.2005.04.5252","article-title":"Placebo-controlled phase III trial of immunologic therapy with sipuleucel-T (APC8015) in patients with metastatic, asymptomatic hormone refractory prostate cancer","volume":"24","author":"Small","journal-title":"J Clin Oncol"},{"issue":"21","key":"R20","doi-asserted-by":"publisher","first-page":"2367","DOI":"10.1200\/JCO.20.02824","article-title":"Comparison of radiation with or without concurrent trastuzumab for HER2-positive ductal carcinoma in situ resected by lumpectomy: a phase III clinical trial","volume":"39","author":"Cobleigh","journal-title":"J Clin Oncol"},{"issue":"11","key":"R21","doi-asserted-by":"publisher","first-page":"1538","DOI":"10.1001\/jamaoncol.2017.1269","article-title":"Predictors of hand-foot syndrome and pyridoxine for prevention of capecitabine-induced hand-foot syndrome: a randomized clinical trial","volume":"3","author":"Yap","journal-title":"JAMA Oncol"},{"issue":"13","key":"R22","doi-asserted-by":"publisher","first-page":"1300","DOI":"10.1001\/jama.2014.2626","article-title":"Tadalafil for prevention of erectile dysfunction after radiotherapy for prostate cancer: the Radiation Therapy Oncology Group [0831] randomized clinical trial","volume":"311","author":"Pisansky","journal-title":"JAMA"},{"issue":"6","key":"R23","doi-asserted-by":"publisher","first-page":"744","DOI":"10.1200\/JCO.2012.43.8820","article-title":"Randomized phase III trial of induction chemotherapy with docetaxel, cisplatin, and fluorouracil followed by surgery versus up-front surgery in locally advanced resectable oral squamous cell carcinoma","volume":"31","author":"Zhong","journal-title":"J Clin Oncol"},{"issue":"12","key":"R24","doi-asserted-by":"publisher","first-page":"1923","DOI":"10.1001\/jamaoncol.2020.4574","article-title":"Effect of weekly paclitaxel with or without bevacizumab on progression-free rate among patients with relapsed ovarian sex cord-stromal tumors: the ALIENOR\/ENGOT-ov7 randomized clinical trial","volume":"6","author":"Ray-Coquard","journal-title":"JAMA Oncol"},{"issue":"1","key":"R25","doi-asserted-by":"publisher","first-page":"59","DOI":"10.1016\/S1470-2045(13)70510-2","article-title":"Tecemotide (L-BLP25) versus placebo after chemoradiotherapy for stage III non-small-cell lung cancer (START): a randomised, double-blind, phase 3 trial","volume":"15","author":"Butts","journal-title":"Lancet Oncol"},{"issue":"4","key":"R26","doi-asserted-by":"publisher","first-page":"792","DOI":"10.1200\/JCO.2005.05.098","article-title":"Randomized phase III trial of capecitabine compared with bevacizumab plus capecitabine in patients with previously treated metastatic breast cancer","volume":"23","author":"Miller","journal-title":"J Clin Oncol"},{"key":"R27","doi-asserted-by":"publisher","DOI":"10.1016\/j.annemergmed.2019.03.011","article-title":"Evaluation of spin in the abstracts of emergency medicine randomized controlled trials","author":"Reynolds-Vaughn","journal-title":"Ann Emerg Med"},{"key":"R28","volume":"287","author":"Yun","journal-title":"Proc Mach Learn Res"},{"key":"R29","unstructured":"Multiple endpoints in clinical trials: guidance for industry. US Food and Drug Administration. 2022. URL: https:\/\/www.fda.gov\/files\/drugs\/published\/Multiple-Endpoints-in-Clinical-Trials-Guidance-for-Industry.pdf [Accessed 07-01-2026]"},{"key":"R30","doi-asserted-by":"publisher","DOI":"10.1038\/s43586-020-00001-2","article-title":"Bayesian statistics and modelling","volume":"1","author":"van de Schoot","journal-title":"Nat Rev Methods Primers"},{"key":"R31","unstructured":"Windisch-paul\/positive_negative. GitHub. URL: https:\/\/github.com\/windisch-paul\/positive_negative [Accessed 25-05-2025]"}],"container-title":["JMIR Cancer"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/cancer.jmir.org\/2026\/1\/e78221","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,6,30]],"date-time":"2026-06-30T18:10:47Z","timestamp":1782843047000},"score":1,"resource":{"primary":{"URL":"https:\/\/cancer.jmir.org\/2026\/1\/e78221"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,1,21]]},"references-count":31,"URL":"https:\/\/doi.org\/10.2196\/78221","relation":{},"ISSN":["2369-1999"],"issn-type":[{"value":"2369-1999","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,1,21]]},"article-number":"v12i33e78221"}}