Carl Counsell, MRCP(UK)
SYSTEMATIC REVIEW SERIES
Series Editors: Cynthia Mulrow, MD, MSc; Deborah Cook, MD, MSc
Much time and effort are spent on designing primary research studies.Similar care must be given to planning systematic reviews. The review should be based on an important, well-focused question that is relevant to patient care. By formulating the question properly, the criteria that primary studies must meet to be included in the review become clear. These criteria, which comprise the types of persons involved, exposure, control group, outcomes, and study designs of interest, can then be refined so that they are clinically relevant, sensible, and workable. Inclusion criteria that are too narrow will limit the amount of data in the review, thereby increasing the risk for chance results and making the review less useful for the reader. Reviews should include studies whose designs offer the least biased answer to the question being asked. To maximize available data and reduce the risk for bias, as many relevant studies as possible need to be identified, regardless of publication status or language. Multiple overlapping search strategies should therefore be used and must be carefully planned. Strategies include searching the many electronic databases available (after careful consideration of which terms to enter), manually searching journals and conference proceedings, searching bibliographies of articles, searching existing registers of studies, and contacting companies or researchers. The time taken to formulate the question adequately and develop appropriate searches will increase the chance of producing a high-quality, meaningful review.
Carl Counsell. Formulating Questions and Locating Primary Studies for Inclusion in Systematic Reviews. Ann Intern Med. 1997;127:380–387. doi: 10.7326/0003-4819-127-5-199709010-00008
Download citation file:
Published: Ann Intern Med. 1997;127(5):380-387.
Emergency Medicine, Hematology/Oncology, Neurology.
Results provided by:
Copyright © 2017 American College of Physicians. All Rights Reserved.
Print ISSN: 0003-4819 | Online ISSN: 1539-3704
Conditions of Use