Systematically assessing methods used by librarians to teach evidence-based practice: What works best? Assako N. Holyoke, PhD, MSLIS, Saint Louis University Medical Center Library; Stephanie M. Swanberg, MSI, AHIP, Oakland University William Beaumont School of Medicine; Genevieve Gore, Schulich Library of Science and Engineering; Virginia Pannabecker, Arizona State University Libraries Health Sciences Library; Carolyn C. Dennison, MA, MLIS, AHIP, University of Hawaii at Manoa Library; Christine F. Marton, PhD, University of Toronto; Alison D. Farrell, BA, MLIS, Memorial University of Newfoundland Health Sciences Library; Mindy Thuna, BSc, MSc, MISt, University of Toronto Mississauga Library; Kelly K. O'Brien, MLIS, Crawford Library of the Health Sciences; Viola Machel, St. Catharines Public Library.
Objective: To conduct a systematic review of the literature on methods of teaching evidence-based practice (EBP) and assess which methods are most effective.
Background: This study attempts to address one of the fifteen research questions generated from a delphi study by Eldredge JD, which was part of a Medical Library Association Research Section’s Research Agenda project.
Methods: A team of 10 librarians with a common interest in EBP participated in this systematic review project. The research protocol was developed according to PRISMA standards. The research question was defined, the inclusion and exclusion criteria determined, and the base literature search strategy developed and tested on PubMed. Fourteen databases were searched through November 2013, including MEDLINE, EMBASE, LISA, and other relevant databases to the study. The retrieved citations were divided into 5 blocks after duplicates were eliminated and two librarians were allocated to each block for selection, data extraction, and critical appraisal of studies identified as research papers and final data analyzed.
Result: A total of 45,457 citations were retrieved and over 20,000 duplicates were excluded. A total of 178 studies met the inclusion/exclusion criteria based on title and abstract. Further full-text analysis identified 19 studies for data extraction, critical appraisal, and final data analysis. Most of these 19 papers concerned EBP educational efforts addressed to medical students or residents and nursing and allied health students. Six were identified as randomized controlled trial studies. Of the 19 studies, 68% concerned educational efforts using the traditional face-to-face instruction method, most of them (85%) combined with hands-on computer-assisted teaching. Over 90% assessed learning through test scores and literature searches, based on before and after intervention. Statistical analysis used to assess learning outcome included paired T-test, ANCOVA, ANOVA, Mann-Whitney U test, and Wilcoxon paired test. Although most of the studies claimed success in improving students’ skill, we did not find enough studies to allow us to conduct valid statistical analysis of different teaching methods.
Conclusion: Studies comparing different teaching methods are rare, which indicates the need for future research in this area. EBP courses taught by librarians are in general integrated in medical school or other health sciences school curriculum. Such courses generally address the ‘acquire’ step of EBM as part of or incorporated in the existing EBM curriculum. Traditional lecture-type instruction is most common. Lecture-type instruction is generally provided in combination with hands-on computer lab instruction. Web-based courses are the next most common form of instruction. Web-based courses frequently take the form of online tutorials using tools such as WebCT/Blackboard. The pre-post format is frequently used to assess learning by using tests and literature search exercises, although validated or non-validated questionnaires, such as Berlin Questionnaire or Fresno Test have also been addressed.
