CHLA 2023 CONFERENCE CONTRIBUTED PAPERS / ABSC CONGRÈS 2023 COMMUNICATIONS LIBRES

Abstract

Introduction: The use of controlled vocabulary to identify relevant articles is a central element of bibliographic database instruction in health sciences. Students learning to search MEDLINE are taught that MeSH yields precise results, and that MeSH indexing increases an article's findability, reliably describing an article's contents. Indexing for MEDLINE was done completely by human indexers until 2011. Since April 2022, all articles are assigned MeSH via automated indexing (AI). Per the NLM, MeSH assigned by AI are determined based on terms in title, abstract, and terms and indexing of 'related records', with human review and curation of results “as appropriate”. As MEDLINE instruction typically starts with teaching learners to identify key elements or concepts in their research question and find appropriate MeSH for them, we sought to explore the following: how well does AI identify key concepts of an article? Are concepts missed more or less when compared to human indexers? Drawing on the PICO framework, are missing concepts more often any particular PICO element? Methods: We reviewed samples of automated and human-indexed records from shortly before April 2022, and some entirely-automated from later, to determine whether their main concepts were adequately represented with MeSH. Working in pairs, our team used a web form to assign key concepts (based on the PICO framework) that, per our experience, would be used to find it and similar articles based on title and abstract. Assigned MeSH were then displayed and analyzed to determine whether they adequately represented the key concepts of each record. Results & Conclusion: As the study is ongoing, results are forthcoming. Potential impacts of Automated Indexing on library instruction and basic searching will be discussed. Introduction: As we pass three years since the declaration of the COVID-19 pandemic, the question persists as to how 'normal' has evolved since then. Recent research into health information-seeking behaviour (ISB) has found evidence of differences related to the COVID-19 pandemic. Our pre-pandemic research into undergraduate students' ISB identified a mismatch between the rating of a resource's credibility, and the frequency of its use. In contrast, since the pandemic started, students report taking a more critical approach to selecting and appraising health information, including checking information against other sources, verifying if information is up-to-date, and assessing the author/institution to check for bias. Here, we present a longitudinal study of health ISB before and after the pandemic started to determine: 1) if/how health ISB has changed from before the pandemic, and 2) whether health ISB differs for everyday health issues, and those specifically related to COVID-19. Methods: Data collection involves two online surveys of McGill University undergraduate students, the first before the pandemic, and a replicate survey in early 2023. The pre-pandemic survey (in the context of everyday health issues) yielded over 3500 responses. The follow-up survey includes the same questions regarding frequency of resource use, and assessment of resource credibility. The questions are repeated for two contexts: everyday health issues, and specifically related to COVID-19. Results: Data collection and analysis are in progress; preliminary results of the post-pandemic survey, with comparison to the pre-pandemic survey will be presented. Discussion: We will discuss the implications of the preliminary findings for health librarianship, with focus on how health ISB has evolved during the pandemic, and the implications for the 'new normal' of the future. Introduction: There are various automated electronic methods for deduplicating references after searching multiple resources during the process of conducting an evidence synthesis including deduplication features in citation managers and systematic review tools. This research examines the performance of default algorithms for deduplicating references in RefWorks (using the revamped deduplication feature released by ProQuest in November 2022), Deduklick (introduced in 2022 as a stand-alone deduplication AI tool), and Systematic Review Accelerator (a suite of automation tools that includes a “Deduplicator” feature). Methods: Database search strategies from a previously conducted evidence synthesis were used to collect a heterogenous sample of references from the following Ovid databases: MEDLINE, Embase, PsycINFO and Cochrane CENTRAL. Manual abstraction of the sample of references was used to identify duplicates and develop a benchmark for comparing deduplicated references from RefWorks, Deduklick, and Systematic Review Accelerator. The number of false negative and false positive references generated by each program was determined to calculate accuracy, sensitivity, and specificity. Results: The performance of default settings for deduplicating references in RefWorks, Deduklick, and Systematic Review Accelerator will be presented and directly compared to other citation managers and systematic review software programs, which have been previously evaluated and reported on using the same sample of references and process of analysis. Conclusions: The performance of electronic methods for deduplicating references should be taken into consideration when conducting evidence syntheses to avoid unintentionally removing eligible studies, which could introduce bias. The cost and availability of programs that deduplicate references may limit and/or exclude their use among researchers, which may have implications for evidence synthesis quality.