{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/inferring-which-medical-treatments-work-from","title":"Inferring Which Medical Treatments Work from Reports of Clinical Trials","arxiv_id":"1904.01606","date":"2019-04-02","proceeding":"NAACL 2019 6","authors":["Eric Lehman","Jay DeYoung","Regina Barzilay","Byron C. Wallace"],"abstract":"How do we know if a particular medical treatment actually works? Ideally one\nwould consult all available evidence from relevant clinical trials.\nUnfortunately, such results are primarily disseminated in natural language\nscientific articles, imposing substantial burden on those trying to make sense\nof them. In this paper, we present a new task and corpus for making this\nunstructured evidence actionable. The task entails inferring reported findings\nfrom a full-text article describing a randomized controlled trial (RCT) with\nrespect to a given intervention, comparator, and outcome of interest, e.g.,\ninferring if an article provides evidence supporting the use of aspirin to\nreduce risk of stroke, as compared to placebo.\n  We present a new corpus for this task comprising 10,000+ prompts coupled with\nfull-text articles describing RCTs. Results using a suite of models --- ranging\nfrom heuristic (rule-based) approaches to attentive neural architectures ---\ndemonstrate the difficulty of the task, which we believe largely owes to the\nlengthy, technical input texts. To facilitate further work on this important,\nchallenging problem we make the corpus, documentation, a website and\nleaderboard, and code for baselines and evaluation available at\nhttp://evidence-inference.ebm-nlp.com/.","url_abs":"http://arxiv.org/abs/1904.01606v2","url_pdf":"http://arxiv.org/pdf/1904.01606v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"inferring-which-medical-treatments-work-from","repo_url":"https://github.com/jayded/evidence-inference","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"inferring-which-medical-treatments-work-from","repo_url":"https://github.com/bepnye/evidence_extraction","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}}],"tasks":[{"task_slug":"articles","task_name":"Articles"}],"methods":[],"datasets_introduced":[{"slug":"evidence-inference","name":"Evidence Inference","full_name":""}],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1904.01606","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}