Systematic Review Search Strategies Are Poorly Reported and Not Reproducible

Authors

Melissa L. Rethlefsen, University of New Mexico Health Sciences Library and Informatics Center, Albuquerque New MexicoFollow

Publication Date

2-2-2024

Abstract

In this seminar, Prof. Rethlefsen will describe the process and outcomes of a research project examining the reproducibility of 100 randomly selected systematic reviews indexed in MEDLINE in November 2021. Published systematic reviews were analyzed for their compliance with six basic reporting items from PRISMA-S, an extension to the Preferred Reporting Items for Systematic reviews and Meta-Analyses statement focused on literature searches. All database searches in the 100 systematic reviews were reproduced, and the number of results compared between the original searches and the reproduced searches. Only one systematic review was fully reproducible, which indicates that published systematic reviews may have hidden risk of bias. Lessons regarding the complexities of systematic review searches and reporting will be discussed.

Learning Objectives:

1. Participants will be able to describe reporting elements that impact systematic review search reproducibility
2. Participants will be able to understand how poor-quality search reporting and lack of reproducibility impact the potential for risk of bias in systematic reviews

Document Type

Video

Publisher

Biomedical Informatics Seminar Series (BioMISS)

Language

English

Keywords

systematic review

Description

If you would like a copy of the presentation slides, they are located here.

Recommended Citation

Rethlefsen, Melissa L.. "Systematic Review Search Strategies Are Poorly Reported and Not Reproducible." (2024). https://digitalrepository.unm.edu/bmi/26