The Effect of Librarian Collaboration on the Quality of Otolaryngology Systematic Reviews and Meta-Analyses

Wednesday, May 4, 2022

2:30 PM – 2:45 PM CT

Location: 354

Presenter and Author(s)

Emily Brennan, MLIS

Research and Education Informationist
Medical University of South Carolina Library
Charleston, South Carolina

Slides
Rachel Whitney, AHIP

Research & Education Informationist
Medical University of South Carolina
Charleston, South Carolina

Author(s)

TG

Tamar Gordis

Research Fellow
Medical University of South Carolina
MS

Michael Shih

Clinical Research Fellow & Senior Manager of Clinical Research Operations
Medical University of South Carolina
SN

Shaun A. Nguyen, M.D.,FAPCR

Professor & Director of Clinical Research
Medical University of South Carolina

Objectives: This study aims to determine if librarian collaboration is associated with better search quality, reproducibility, and reporting in otolaryngology systematic reviews and meta-analyses. Otolaryngology researchers at the Medical University of South Carolina observed an improvement in the quality of systematic reviews and meta-analyses when collaborating with a librarian, which aligns with what has been published in other health sciences disciplines. To better understand if this trend of increased quality through librarian collaboration is consistent in otolaryngology literature, faculty and research fellows approached librarians with an opportunity to lead this research project.

Methods: Clarivate's Journal Citation Reports was utilized to identify 33 otolaryngology journals which were then searched in PubMed using a systematic review and meta-analysis search hedge. To identify trends over time, studies from selected date ranges beginning in 2010 were included. Librarians and otolaryngology research fellows extracted data from 505 systematic reviews and meta-analyses. Two research fellows used a data extraction form in Covidence to independently extract the following data elements: journal title, publication type, level of librarian involvement, reporting guideline followed, number of databases searched, dates of database searches, database limits/filters, search peer review, flow chart, grey literature searched, citation searching performed, and times cited. Disagreements were resolved through consensus. Two librarians used a Google form, based on a modified PRESS checklist, to independently evaluate search reproducibility and quality.

Results: These results are preliminary and final results are forthcoming pending data analysis by an otolaryngology statistician. All four types of librarian involvement (no acknowledgement, mentioned in text, acknowledgement, and co-authorship) have increased over time. In particular, librarian co-authorship increased from 0% in 2010, to 1.48% in 2015, to 10% in 2021. We anticipate this trend to continue in future years. Similar to the increase in librarian involvement over time, an increase in the overall usage of reporting guidelines was observed. 48% of articles without librarian involvement did not include a reproducible search strategy, while 57% of articles with librarian involvement included a reproducible search strategy for more than one database. The average search quality was higher in searches which had librarian involvement versus searches that had no librarian involvement.

Conclusions: Many organizations recommend including librarians on systematic reviews and meta-analyses. We hope that this study adds to the body of evidence linking librarian collaboration with improved search quality, reproducibility, and reporting, as well as demonstrates the value of providing authorship to librarians on systematic reviews and meta-analyses.