From: Impact of large language model (ChatGPT) in healthcare: an umbrella review and evidence synthesis
Author, year | Country (corresponding author’s affiliation) | Search Strategy (databases, search terms and timeline) | RoB* assessment performed (Y/N) | Number and type of studies included | Meta-analysis performed (Y/N) | Version of ChatGPT studied | Utilization of ChatGPT (in writeup/ conducting systematic search) | Acknowledgement of ChatGPT by authors | Funding information |
---|---|---|---|---|---|---|---|---|---|
[10] | Lucknow, India | Embase, Scopus, PubMed, Google Scholar Search term: ChatGPT Data till 24th May 2023 | No | 118 articles (original articles, reviews, editorial/ commentaries, and letter to the editor) | No | Not mentioned | ChatGPT was used for analysis of records and manuscript writing ChatPDF was used to generate | Yes | Authors declare no financial support |
[27] | Leipzing, Germany | PubMed search MESH terms were developed by ChatGPT and refined by human authors January, 2020 onwards (NLP instead of ChatGPT only) | No | 5 (main concern was to evaluate the effectiveness and reliability of ChatGPT in conducting review compared with human authors) | No | 3.5 (Legacy) and 4.0 | All tasks of systematic review augmented by ChatGPT | Yes | No information in the article |
[25] | Amman, Jordan | PubMed/MEDLINE and Google Scholar Search term: ChatGPT Year: 2022–2023 | No | 60 (article, review, communication, editorial, opinion) | No | Not mentioned | No | No | Authors declare no financial support |
[18] | Essen, Germany | Pubmed Search term: ChatGPT Till March, 2023 | No | 58 (all types but the writeup done for original articles and opinion/editorial pieces separately) | No | Version released in Nov., 2022 | No | No | Medical Faculty of German Rheinisch-Westfälische Technische Hochschule Aachen University as part of the Clinician Scientist Program |
[36] | Basra, Iraq | Taylor and Francis, Google Scholar, Scopus, Web of Science, Elsevier, Springer, MDPI, IEEE Xplore digital and Wiley Keywords: ChatGPT with various keywords for healthcare and medicine between November 2022 and August 2023 | No | 82 (All types) | No (although the title suggests so) | Not mentioned | No | No | Funding by The Deanship of Scientific Research at King Khalid University, KSA |
[31] | Dublin, Ireland | PubMed, EMBASE and Web of Science Keyword not described Search till 18 June 2023 | No | 6 (prospective) | No | version 3.0 to 4.0 | No | No | Authors declare that open access funding provided by IReL |
[2] | Rajnandgaon, India | PubMed, Google Scholar, Scopus, EMBASE, Cochrane Library, and UpToDate Key terms: ChatGPT and Medical Informatics which were combined using Boolean operators “AND” and “OR” with “dental”, “specialty”, “accuracy”, “query”, “response” and “meta- analysis”. Articles published in 2023 only | Yes | 11 descriptive studies | Yes | 3.5 | No | No | No information in the article |
[3] | Zenica, Bosnia and Herzegovina | Pubmed, Embase, Scopus Keywords:(ChatGPT OR OpenAI) AND (neurosurgery OR spinal surgery) till 12th of August, 2023 | No | 13 (all types) | No | Not mentioned | No | No | Authors declare no financial support |
[17] | Quebec, Canada | PubMed, Scopus and Web of Science Keyterm: ChatGPT till 2nd June 2023 | No | 19 peer-reviewed articles | yes | 3.5 | No | No | Authors declare no financial support |
[20] | Sarajevo, Bosnia and Herzegovina | Google Scholar and PubMed Keyword: “ChatGPT applications in medicine” Till April 15, 2023 | No | 31- any type of published scientific research or preprints | No | Not mentioned | Yes (Paraphrasing and writeup) | Yes | No information in the article |
[15] | Tel Aviv, Israel | Pubmed Key words: MESH terms related to ‘ChatGPT’, and ‘Gastroenterology’ | Yes | 6 | No | Heterogenous | No | No | Authors declare no financial support |
[32] | Rosemont, USA | Pubmed, Embase, Ovid, Global Health, PsycINFO, Web of Science Using search phrases related to chatGPT and public health dentistry Between: March 31, 2018, and March 31, 2023 | No | 39 | No | Not mentioned | No | No | Authors declare no financial support |
[22] | Abu Dhabi, UAE | PubMed, Google Scholar, and Embase MESH terms for "neurosurgery" AND "ChatGPT" Till June 30, 2023 | No | 22 -All types (peer reviewed and gray literature search) | No | Not specified | No | No | Authors declare no financial support |
[4] | Cleveland, Ohio, USA | MEDLINE, EMBASE using the search terms: ChatGPT, imaging, radiology, LLM, large language models, BARD from November 2022 to August 15, 2023, | Yes | 51 (23 original research articles and 28 non-original research articles) | No | ChatGPT version 3, 3.5 and 4 | No | No | Authors declare no financial support |
[5] | Bucharest, Romania | ProQuest, Scopus, and the Web of Science with search terms including "generative artificial intelligence-based diagnostic algorithms," "disease risk detection," "personalized and targeted healthcare procedures," and "patient care safety and quality." throughout April 2023 | Yes | 32 (28 original research articles and 4 review articles) | No | Not specified | No | No | The paper is an output of the project NFP313010BWN6 “The implementation framework and business model of the Internet of Things, Industry 4.0 and smart transport.” The funder had no role in study design, data collection analysis, and interpretation, decision to submit the manuscript for publication, or the preparation and writing of this paper |
[35] | Singapore | PubMed, Europe PMC, Scopus, and Web of Science keywords related to ChatGPT and LLMs (“large language model”, “natural language processing”, “generative artificial intelligence”, “ChatGPT”, “chatbot”, “GPT-3.5”, “GPT-4.0”) with those specific to the field of ophthalmology (“ophthalmology”, “ophthalmic”, “ophthalmologist”, “ophthalmological”, “ocular”, “optical”, “eye”, “retina”, “vision science”, “vision research”). published between 1 January 2022 and 31 July 2023 | No | 32 (24 original research articles and 8 commentaries) | No | ChatGPT 3.5 & 4.0 | No | No | National Medical Research Council of Singapore |
[24] | Cleveland, USA | Google Scholar, Web of Science, PubMed, and Medline using search terms “ChatGPT” AND “Chatbot” AND “Medical Research” searched on January 21, 2023, at 9:26 PM EST to identify articles published between 2022 and 2023 | No | 6 (2 literature review articles, 1 case study, 1 editorial, 1 perspective, 1 not specific) | No | Not specified | No | No | No information in the article |