Premium
Comparison of a full systematic review versus rapid review approaches to assess a newborn screening test for tyrosinemia type 1
Author(s) -
TaylorPhillips Sian,
Geppert Julia,
Stinton Chris,
Freeman Karoline,
Johnson Samantha,
Fraser Hannah,
Sutcliffe Paul,
Clarke Aileen
Publication year - 2017
Publication title -
research synthesis methods
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 3.376
H-Index - 35
eISSN - 1759-2887
pISSN - 1759-2879
DOI - 10.1002/jrsm.1255
Subject(s) - systematic review , tyrosinemia , data extraction , medline , medical physics , test (biology) , medicine , evidence based medicine , computer science , pathology , alternative medicine , political science , paleontology , nuclear magnetic resonance , tyrosine , biology , physics , law
Background Rapid reviews are increasingly used to replace/complement systematic reviews to support evidence‐based decision‐making. Little is known about how this expedited process affects results. Objectives To assess differences between rapid and systematic review approaches for a case study of test accuracy of succinylacetone for detecting tyrosinemia type 1. Methods Two reviewers conducted an “enhanced” rapid review then a systematic review. The enhanced rapid review involved narrower searches, a single reviewer checking 20% of titles/abstracts and data extraction, and quality assessment using an unadjusted QUADAS‐2. Two reviewers performed the systematic review with a tailored QUADAS‐2. Post hoc analysis examined rapid reviewing with a single reviewer (basic rapid review). Results Ten papers were included. Basic rapid reviews would have missed 1 or 4 of these (dependent on which reviewer). Enhanced rapid and systematic reviews identified all 10 papers; one paper was only identified in the rapid review through reference checking. Two thousand one hundred seventy‐six fewer title/abstracts and 129 fewer full texts were screened during the enhanced rapid review than the systematic review. The unadjusted QUADAS‐2 generated more “unclear” ratings than the adjusted QUADAS‐2 [29/70 (41.4%) versus 16/70 (22.9%)], and fewer “high” ratings [22/70 (31.4%) versus 42/70 (60.0%)]. Basic rapid reviews contained important inaccuracies in data extraction, which were detected by a second reviewer in the enhanced rapid and systematic reviews. Conclusions Enhanced rapid reviews with 20% checking by a second reviewer may be an appropriate tool for policymakers to expeditiously assess evidence. Basic rapid reviews (single reviewer) have higher risks of important inaccuracies and omissions.