Quadas-3: A Revised Tool for the Quality Assessment of Diagnostic Test Accuracy Studies

Assessing risk of bias and applicability is critical in diagnostic test accuracy reviews to ensure valid, relevant findings that can inform decision making and policy. This article introduces QUADAS-3, an updated version of the widely used QUADAS-2 tool that was developed in response to the latest evidence, feedback from users of QUADAS-2, and comprehensive piloting. The QUADAS-3 tool comprises 6 phases: 1) define the systematic review synthesis question, 2) define the ideal test accuracy trial for each synthesis question, 3) draw a flow diagram, 4) identify estimates of accuracy for assessment, 5) assess risk of bias and applicability, and 6) make an overall judgment. Risk of bias is assessed across 4 domains: Participants, Index Test, Target Condition, and Analysis. The first 3 domains are also assessed in terms of concerns regarding applicability to the synthesis question. Signaling questions are included to help judge risk of bias. The QUADAS-3 tool incorporates several changes compared with the previous version (QUADAS-2) which will ideally improve its validity, usability, and usefulness. Key changes from QUADAS-2 include introduction of the concept of an ideal test accuracy trial, assessment at the estimate level rather than the study level, replacement of the "Flow and Timing" domain with a new "Analysis" domain, introduction of sections on the rationale for judgments, and formal introduction of overall judgment of risk of bias and concerns regarding applicability. Some changes to domains and signaling questions have also been made.