Not so weak PICO: leveraging weak supervision for participants, interventions, and outcomes recognition for systematic review automation.
Open accessThe DOI matches its Crossref record; 34 of the 34 references it cites were found in a registry or an index.
Abstract
Objective: The aim of this study was to test the feasibility of PICO (participants, interventions, comparators, outcomes) entity extraction using weak supervision and natural language processing. Methodology: We re-purpose more than 127 medical and nonmedical ontologies and expert-generated rules to obtain multiple noisy labels for PICO entities in the evidence-based medicine (EBM)-PICO corpus. These noisy labels are aggregated using simple majority voting and generative modeling to get consensus labels. The resulting probabilistic labels are used as weak signals to train a weakly supervised (WS) discriminative model and observe performance changes. We explore mistakes in the EBM-PICO that could have led to inaccurate evaluation of previous automation methods. Results: In total, 4081 randomized clinical trials were weakly labeled to train the WS models and compared against full supervision. The models were separately trained for PICO entities and evaluated on the EBM-PICO test set. A WS approach combining ontologies and expert-generated rules outperformed full supervision for the participant entity by 1.71% macro-F1. Error analysis on the EBM-PICO subset revealed 18–23% erroneous token classifications. Discussion: Automatic PICO entity extraction accelerates the writing of clinical systematic reviews that commonly use PICO information to filter health evidence. However, PICO extends to more entities—PICOS (S—study type and design), PICOC (C—context), and PICOT (T—timeframe) for which labelled datasets are unavailable. In such cases, the ability to use weak supervision overcomes the expensive annotation bottleneck. Conclusions: We show the feasibility of WS PICO entity extraction using freely available ontologies and heuristics without manually annotated data. Weak supervision has encouraging performance compared to full supervision but requires careful design to outperform it.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Figures





Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Although used more frequently in the past, the 15 publications published between 2017 and now that use this approach alongside other architectures such as LLM, 148 Transformer, 33,37 -39,148,156,174 conditional random fields (CRF), 40,156 use it with SVM 41 or other binary classifiers.”
From Data extraction methods for systematic review (semi)automation: Update of a living systematic review (2025) · cites this paper as [37]
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“Neither human nor machine can instantly perform perfect data extraction or labelling, 37”
From Data extraction methods for systematic review (semi)automation: Update of a living systematic review (2025) · cites this paper as [37]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Methley, A. M., Campbell, S., Chew-Graham, C., McNally, R., & Cheraghi-Sohi, S. (2014) · BMC Health Services Research · doi:10.1186/s12913-014-0579-0 · Cited by 2,706
Domain-specific language model pretraining for biomedical natural language processing
Gu, Y., Tinn, R., Cheng, H., Lucas, M., Usuyama, N., Liu, X., Naumann, T., Gao, J., & Poon, H. (2022) · ACM Transactions on Computing for Healthcare · doi:10.1145/3458754 · Cited by 2,170 · cited 1×
“PubMedBERT was trained on PubMed literature and was chosen because of its domain similarity to our training data (PubMed abstracts) and task.(39) It was tuned on fixed parameters listed in the experimental details section in the Supplementary Material.”
MethodThe human phenotype ontology: a tool for annotating and analyzing human hereditary disease
Robinson, P. N., Köhler, S., Bauer, S., Seelow, D., Horn, D., & Mundlos, S. (2008) · The American Journal of Human Genetics · doi:10.1016/j.ajhg.2008.09.017 · Cited by 1,034
Disease Ontology: a backbone for disease semantic integration
Schriml, L. M., Arze, C., Nadendla, S., Chang, Y.-W. W., Mazaitis, M., Felix, V., Feng, G., & Kibbe, W. A. (2012) · Nucleic Acids Research · doi:10.1093/nar/gkr972 · Cited by 929
Snorkel: rapid training data creation with weak supervision
Ratner, A., Bach, S. H., Ehrenberg, H., Fries, J., Wu, S., & Ré, C. (2017) · Proceedings of the VLDB Endowment · doi:10.14778/3157794.3157797 · Cited by 777
Borah, R., Brown, A. W., Capers, P. L., & Kaiser, K. A. (2017) · BMJ Open · doi:10.1136/bmjopen-2016-012545 · Cited by 737 · cited 1×
“Manually analyzing PICO information from thousands of publications for a single SR often takes 2–8 months of 2 medical experts’ time.(2) It can be automated using machine learning (ML) by directly pointing the human reviewers to the PICO descriptions, facilitating quick decision-making for the study’s relevance.”
MethodCharacteristics of Clinical Trials Registered in ClinicalTrials.gov, 2007-2010
Califf, R. M. (2012) · JAMA · doi:10.1001/jama.2012.3424 · Cited by 516
The Unified Medical Language System: An Informatics Research Collaboration
Humphreys, B. L., Lindberg, D. A. B., Schoolman, H. M., & Barnett, G. O. (1998) · Journal of the American Medical Informatics Association · doi:10.1136/jamia.1998.0050001 · Cited by 514
Snorkel: rapid training data creation with weak supervision
Ratner, A., Bach, S. H., Ehrenberg, H., Fries, J., Wu, S., & Ré, C. (2020) · The VLDB Journal · doi:10.1007/s00778-019-00552-1 · Cited by 510
A clinical text classification paradigm using weak supervision and deep representation
Wang, Y., Sohn, S., Liu, S., Shen, F., Wang, L., Atkinson, E. J., Amin, S., & Liu, H. (2019) · BMC Medical Informatics and Decision Making · doi:10.1186/s12911-018-0723-6 · Cited by 405
Research works printed but not linked: 9
- Abaho MBD, Williamson P, Dodd S. Correcting crowdsourced annotations to improve detection of outcome types in evidence based medicine. In: CEUR workshop proceedings, Vol. 2429; January 1, 2019: 1– 5.
- Lee GE, Sun A. A study on agreement in PICO span annotations. In: proceedings of the 42nd international ACM SIGIR conference on research and development in information retrieval, Paris, France; ACM; July 18, 2019: 1149– 52.
- Meng Y, Shen J, Zhang C, Han J. Weakly-supervised neural text classification. In: proceedings of the 27th ACM international conference on information and knowledge management, Torino, Italy; ACM; October 17, 2018: 983– 92.
- Mintz M, Bills S, Snow R, Jurafsky D. Distant supervision for relation extraction without labeled data. In: proceedings of the joint conference of the 47th annual meeting of the ACL and the 4th international joint conference on natural language processing of the AFNLP, Suntec, Singapore; August 2009: 1003– 11.
- Imamura H, Sato I, Sugiyama M. Analysis of minimax error rate for crowdsourcing and its application to worker clustering model. In: proceedings of 35th international conference on machine learning (ICML), Stockholm, Sweden; July 3, 2018: 2147– 56.
- Foncubierta Rodríguez A, Müller H. Ground truth generation in medical imaging: a crowdsourcing-based iterative approach. In: proceedings of the ACM multimedia 2012 workshop on crowdsourcing for multimedia, New York, United States; October 29, 2012: 9– 14.
- Dhrangadhariya A, Müller H. DISTANT-CTO: a zero cost, distantly supervised approach to improve low-resource entity extraction using clinical trials literature. In: proceedings of the 21st workshop on biomedical language processing, Dublin, Ireland; Association for Computational Linguistics; May 2022: 345– 58.
- Liu S, Sun Y, Li B, Wang W, Bourgeois FT, Dunn AG. Sent2Span: span detection for PICO extraction in the biomedical text without span annotations. In: findings of the association for computational linguistics: EMNLP 2021, Punta Cana, Dominican Republic; Association for Computational Linguistics; 2021: 1705– 15.
- Mohammed O, Benlamri R, Fong S. Building a diseases symptoms ontology for medical diagnosis: an integrative approach. In: the first international conference on future generation communication technologies, London, United Kingdom; IEEE; December 12–14, 2012: 104– 8.
Cited by
Data extraction methods for systematic review (semi)automation: a living systematic review
Schmidt, L., Olorisade, B. K., McGuinness, L. A., Thomas, J., & Higgins, J. P. T. (2021) · F1000Research · doi:10.12688/f1000research.51117.1 · Cited by 106
Deep learning with noisy labels in medical prediction problems: a scoping review
Wei, Y., Deng, Y., Sun, C., Lin, M., Jiang, H., & Peng, Y. (2024) · Journal of the American Medical Informatics Association · doi:10.1093/jamia/ocae108 · Cited by 62
Zero-Shot Information Extraction for Clinical Meta-Analysis using Large Language Models
Kartchner, D., Ramalingam, S. S., Al-Hussaini, I., Kronick, O., & Mitchell, C. S. (2023) · Workshop on Biomedical Language Processing (BioNLP) · doi:10.18653/v1/2023.bionlp-1.37 · Cited by 59
Data extraction methods for systematic review (semi)automation: Update of a living systematic review
Schmidt, L., Finnerty Mutlu, A. N., Elmore, R., Olorisade, B. K., Thomas, J., & Higgins, J. P. T. (2023) · F1000Research · doi:10.12688/f1000research.51117.2 · Cited by 44
Hsu, E., & Roberts, K. E. (2025) · Scientific Reports · doi:10.1038/s41598-024-68168-2 · Cited by 44
Metadata
- Authors
- Dhrangadhariya, A., & Müller, H.
- Affiliations
- University of Geneva (CH)
- Corresponding author
- Dhrangadhariya, Anjani
- Venue
- JAMIA Open
- Publisher
- Oxford University Press (OUP)
- Published
- 2023-01-06
- Volume
- 6
- Issue
- 1
- Pages
- ooac107-ooac107
- ISSN
- 2574-2531 (ISSN-L)
- PMC
- PMC9828146
- OpenAlex
- W4315754517
- Cited by
- 18
- References
- 34
- Citations by year
- 2026: 8 · 2025: 3 · 2024: 5 · 2023: 2
- FWCI
- 0.47
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- JAMIA Open (publishedVersion, cc-by)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- ArODES (HES-SO (https://www.hes-so.ch/)) (submittedVersion)
- Funders
- HES-SO Valais-Wallis, Sierre, Switzerland
- Topics
- Health Sciences › Medicine › Health Informatics › Biomedical Text Mining and Ontologies
- Physical Sciences › Computer Science › Artificial Intelligence › Topic Modeling
- Social Sciences › Decision Sciences › Statistics, Probability and Uncertainty › Meta-analysis and systematic reviews
- Keywords
- weakly supervised learning, systematic review automation, label noise, medical ontologies, majority voting, generative models, randomized controlled trials, PICO
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.