Article Text

Download PDFPDF

076 Enhancing the Acceptance and Implementation of GRADE Summary Tables for Evidence about Diagnostic Tests
  1. R Mustafa1,2,
  2. W Wiercioch1,
  3. J Brozek1,
  4. M Lelgemann3,
  5. D Buehler3,
  6. A Garg1,4,
  7. P Bossuyt5,
  8. H Schünemann1
  1. 1McMaster University, Hamilton, Canada
  2. 2University of Missouri-Kansas City, Kansas City, USA
  3. 3GKV, Berlin, Germany
  4. 4University of Western Ontario, London, Canada
  5. 5University of Amsterdam, Amsterdam, The Netherlands


Background The GRADE Working Group developed Summary tables adapted to summarise and present evidence from diagnostic test accuracy (DTA) systematic reviews.

Objective To develop guidance on what information to include in these summary tables and to determine the best method(s) for presentation for different end users, including healthcare providers, systematic reviewers and guideline developers.

Methods We presented a number of alternative summary tables to participants. We conducted questionnaires and one-on-one user testing interviews with target end users. We presented printed copies of summary tables and asked open-ended and 7-point Likert-scale questions to obtain information about users’ understanding and preferences.

Results All participants (n = 60) agreed that using summary tables to present results of DTA reviews is helpful. Presentation of several disease prevalence values was identified as a source of confusion. There was an overall preference for placement of sensitivity and specificity values inside summary tables to allow making a link to individual test results (TP, FN, TN, FP). A third of the participants read explanatory content in table footnotes. Two thirds of the participants noted that additional data, including adverse effects, costs, and treatment consequences, would be helpful for making appropriate conclusions and decisions about diagnostic tests.

Discussion As results of DTA reviews are conceptually complicated, presenting the data in a clear, comprehensive, comprehensible way that is tailored to different end users is critical.

Implications for Guideline Developers/Users We are developing a 3-layer approach, with varied content in summary tables of each layer tailored to the needs of different end users.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.