Statistics from Altmetric.com
Editor,—Evidence based and guideline based medicine is justifiably emphasised in current accident and emergency (A&E) medicine practice. At St James's University Hospital in Leeds, the A&E trainees participate in regular evidence based critical appraisal sessions as part of education development, and such skills are assessed in our Faculty of Accident and Emergency Medicine exit examination.
One source of valued literature is Sackett et al,1 particularly their appraisal cards on the validity, importance, and applicability of a particular type of study. Lacking a photographic memory, I put forward simple acronyms that have helped me to facilitate timely and efficient appraisal for everyday use when selectively scanning relevant journals. Preceding these specific acronyms is a “standards” acronym that follows Crombie's suggestion that standard questions should be used as a filter for all papers.2
I hope they are of use to fellow practitioners of evidence based medicine, and further suggestions will be gratefully received.
Tests and measures appropriate?
Arithmetic (do the numbers add up?)
Null findings (were they considered?)
Relevance to your practice?
Different results from previous reports?
Sample size/power adequate?
Diagnostic test needed?
Independent blind comparison?
Gold standard used regardless of test result?
Numerogram (2×2 table) constructable?
Odds (pre-test and post-test)/likelihood ratios important?
Sensitivity and specificity important?
Safe, cheap, and helpful?
Objective and blinded outcome criteria?
Groups adjusted for prognostic factors?
Numbers recruited and followed up adequate?
Study findings precise?
Similar patients to your own?
Identified risks and benefits?
Decision options clear?
Evidence based decisions?
Large variations in current practice?
Trial ethically approved?
High recruitment and follow up?
Equal treatment and assessment in each group?
Randomised and how?
Potential benefits for patients?
End points applicable?
Absolute risk reduction/number needed to treat?
Systematic search strategy?
Randomised and relevant trials?
Each trial assessed?
Valid results in all trials?
Inconsistent populations or results?
Evidence of benefit via odds ratios/number needed to treat?
Was hypothesis satisfied by the review results?
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.