Rapid reviews: Do they really save time and resources?

Introduction: Conducting full systematic reviews or health technology assessments (HTA) is time-consuming and costly. It usually requires between six months and one year to complete a systematic review, and at least one year to complete a HTA. Caused by the urgency of some questions, information is needed within a shorter timeframe. There are different possibilities to shorten the review process mostly comprising methodological restrictions. Our aim was to test the feasibility of AMSTAR, a tool to assess the quality of systematic reviews, for rapid reviews. Furthermore, we investigated the time needed for a shortened review process. Methods: A systematic literature search for rapid reviews dealing with surgical interventions was conducted. We analyzed publications using AMSTAR, and the results of the rapid reviews were compared to those of an overview of systematic reviews in which AMSTAR was used as well. The duration of the rapid review processes was estimated by calculating the time between the last search date and the approximate or reported review publication date. Results: We identified 17 rapid reviews. Mean AMSTAR score was 5.1 of 11, compared to 5.8 for systematic reviews. Items fulfilled more frequently in rapid reviews were: search for grey literature (65% vs. 33%), listing of excluded studies (59% vs. 37%), and provision of study characteristics (77% vs. 44%). Items fulfilled less frequently were: consideration of study quality in formulating conclusions (47% vs. 70%), conduct of meta-analysis (14% vs. 70%), and statement of conflicts of interest (6% vs. 56%). Median time between search and publication was eight months, range 1 – 27. Conclusion: Apart from differences regarding individual items, mean quality of rapid and systematic reviews was comparable. Thus, the use of AMSTAR for rapid reviews is feasible and makes sense with some adjustments. Strikingly, only 14,3% of rapid reviews were published within three months.

Referenzen beim Verfasser