Academic Journal
Introduction to Concepts in Artificial Intelligence and Machine Learning for Pharmacoepidemiologists: Large Language Models.
| Τίτλος: | Introduction to Concepts in Artificial Intelligence and Machine Learning for Pharmacoepidemiologists: Large Language Models. |
|---|---|
| Συγγραφείς: | Gwinnutt JM; AI and Technology Solutions, IQVIA., de Oliveira R; AI and Technology Solutions, IQVIA., Haviland MJ; AI and Technology Solutions, IQVIA., Shippee JH; AI and Technology Solutions, IQVIA., Eldridge E; AI and Technology Solutions, IQVIA., Mendes Pereira L; AI and Technology Solutions, IQVIA., Bratton E; AI and Technology Solutions, IQVIA., Nanavati J; AI and Technology Solutions, IQVIA., Mack CD; AI and Technology Solutions, IQVIA. |
| Πηγή: | Pharmacoepidemiology and drug safety [Pharmacoepidemiol Drug Saf] 2026 Oct; Vol. 35 (10), pp. e70452. |
| Τύπος έκδοσης: | Journal Article |
| Γλώσσα: | English |
| Στοιχεία περιοδικού: | Publisher: Wiley Country of Publication: England NLM ID: 9208369 Publication Model: Print Cited Medium: Internet ISSN: 1099-1557 (Electronic) Linking ISSN: 10538569 NLM ISO Abbreviation: Pharmacoepidemiol Drug Saf Subsets: MEDLINE |
| Imprint Name(s): | Original Publication: Chichester, West Sussex : Wiley, 1992- |
| Ιατρικοί όροι (MeSH): | Pharmacoepidemiology*/methods , Artificial Intelligence* , Machine Learning*, Humans ; Generative Artificial Intelligence ; Large Language Models ; Natural Language Processing ; Reproducibility of Results ; Review Literature as Topic |
| Περίληψη: | Large language models (LLMs) represent a type of generative artificial intelligence (GenAI) that generate and interpret text, with some LLMs able to process multimodal content (e.g., images, audio, video), and can be deployed as part of agents to perform users' tasks. LLMs can perform natural language processing functions such as summarization, translation, and extraction giving them the potential to enhance and scale pharmacoepidemiological and real-world research by performing tasks such as literature review, data extraction, and medical writing. Despite the growing integration of GenAI tools into research workflows, their technical foundations and methodological implications remain unfamiliar to many pharmacoepidemiologists, who are often responsible for the reliability and accuracy of research that relies on these tools. This paper aims to inform pharmacoepidemiologists about the capabilities and limitations of LLMs to support responsible integration into the field of pharmacoepidemiology, providing an intuitive overview of how LLMs work, focusing on training and text generation, and reviews current and emerging applications in drug effectiveness and safety research and epidemiology. The article addresses challenges associated with LLM use in real-world evidence generation, including concerns regarding reproducibility, bias, hallucinations, plagiarism, data privacy, and the need for validation. (© 2026 The Author(s). Pharmacoepidemiology and Drug Safety published by John Wiley & Sons Ltd.) |
| References: | NPJ Digit Med. 2025 Jan 31;8(1):74. (PMID: 39890970) Gastroenterology. 2024 Apr;166(4):707-709.e3. (PMID: 38151192) Appl Psychol Health Well Being. 2025 Jun;17(3):e70038. (PMID: 40377231) Am J Epidemiol. 2023 Nov 3;192(11):1904-1916. (PMID: 37139570) Mayo Clin Proc Digit Health. 2024 Nov 29;3(1):100184. (PMID: 40206998) J Am Med Inform Assoc. 2025 Jun 1;32(6):1015-1024. (PMID: 40190132) Ann Neurol. 2025 Aug;98(2):308-316. (PMID: 40277428) Front Med (Lausanne). 2019 Apr 17;6:66. (PMID: 31058150) Res Synth Methods. 2024 Jul;15(4):616-626. (PMID: 38484744) Front Drug Saf Regul. 2025 Feb 13;5:1519307. (PMID: 40978504) JCO Clin Cancer Inform. 2023 Feb;7:e2200155. (PMID: 36809022) JAMA. 2025 Jan 28;333(4):319-328. (PMID: 39405325) Nat Med. 2025 Jan;31(1):60-69. (PMID: 39779929) JAMIA Open. 2025 Jul 11;8(4):ooaf058. (PMID: 40655536) J Med Syst. 2025 Jan 16;49(1):10. (PMID: 39820845) J Clin Epidemiol. 2025 May;181:111746. (PMID: 40021099) Pharmacoepidemiol Drug Saf. 2024 Nov;33(11):e70041. (PMID: 39500844) J Am Med Inform Assoc. 2024 Sep 1;31(9):1833-1843. (PMID: 38613821) JMIR Med Inform. 2023 Aug 22;11:e46159. (PMID: 37621203) Res Synth Methods. 2024 Jul;15(4):576-589. (PMID: 38432227) PLoS One. 2025 Feb 21;20(2):e0317084. (PMID: 39982913) JAMA Intern Med. 2023 Jun 1;183(6):589-596. (PMID: 37115527) J Clin Oncol. 2024 Dec 10;42(35):4134-4144. (PMID: 39226489) J Med Internet Res. 2025 Dec 1;27:e84120. (PMID: 41325597) J Clin Epidemiol. 2026 Jun;194:112221. (PMID: 41831731) Stud Health Technol Inform. 2020 Jun 16;270:173-177. (PMID: 32570369) Br J Radiol. 2023 Oct;96(1150):20220878. (PMID: 36971405) Nat Biomed Eng. 2026 Jun 17;:. (PMID: 42310130) Int J Equity Health. 2025 Feb 26;24(1):57. (PMID: 40011901) JAMIA Open. 2024 Sep 25;7(3):ooae098. (PMID: 39323560) Nat Med. 2025 Mar;31(3):943-950. (PMID: 39779926) Syst Rev. 2024 Jun 15;13(1):158. (PMID: 38879534) BMJ. 2025 Mar 24;388:e082505. (PMID: 40127903) NPJ Digit Med. 2021 Jun 3;4(1):93. (PMID: 34083689) BMJ Health Care Inform. 2025 Jan 19;32(1):. (PMID: 39832824) BMJ Lead. 2023 May 2;:. (PMID: 37192124) Front Artif Intell. 2025 Feb 05;8:1533508. (PMID: 39974356) Nat Med. 2023 Aug;29(8):1930-1940. (PMID: 37460753) Lancet Digit Health. 2024 Jul;6(7):e441-e443. (PMID: 38906607) Clin Trials. 2026 Aug;23(4):383-392. (PMID: 41776746) Mayo Clin Proc. 2023 Oct;98(10):1444-1448. (PMID: 37793723) Glob Epidemiol. 2023 Dec 16;7:100130. (PMID: 38188038) JMIR Cancer. 2025 Mar 28;11:e65984. (PMID: 40153782) |
| Contributed Indexing: | Keywords: AI for pharmacoepidemiology; artificial intelligence; generative AI; large language models; natural language processing; transformers Local Abstract: [plain-language-summary] Generative artificial intelligence refers to systems that can create new content in response to user inputs. Large language models (LLMs) are one example of this technology and are designed to generate and interpret text, with some LLMs able to process multimodal content, including images, audio, and video. In drug safety and effectiveness research, these models can help with tasks such as reviewing previously published studies, extracting information from large collections of medical records, and drafting reports about new research. LLMs are created by training them on very large amounts of text, although many contemporary models are also trained on images, audio, and other modalities. During training, the model reads text and repeatedly tries to predict the next word. By learning from its mistakes over time, the model gradually improves. Although LLMs show considerable promise, they must be used with caution in drug safety and effectiveness research. These systems can make mistakes and may reflect biases that exist in the data they were trained on. For this reason, careful testing and ongoing human oversight are essential. |
| Entry Date(s): | Date Created: 20260921 Date Completed: 20260921 Latest Revision: 20260928 |
| Update Code: | 20260929 |
| PubMed Central ID: | PMC13593388 |
| DOI: | 10.1002/pds.70452 |
| PMID: | 42767679 |
| Βάση Δεδομένων: | MEDLINE |
| ISSN: | 1099-1557 |
|---|---|
| DOI: | 10.1002/pds.70452 |