Study-Specific Questions

Clinical Trial

Were patients randomly assigned to the comparison groups?
Was the study blinded? Did the patients or doctors know which group the patient was in?
Was the randomization effective in creating two groups which were similar with respect to age, gender, race, and other potentially confounding variables?
Did patients adhere to the treatment? Did patients drop out?
Were appropriate statistical tests used to compare the groups?
Were the groups analyzed based on their randomized assignment, i.e. a so-called "intention to treat analysis"
Was the sample size large enough to detect a meaningful difference if it had existed?

How did they select the subjects in the comparison groups? Were the groups comparable with respect to other factors?
How did they ascertain risk factor status? Was the data accurate?
Could there have been bias?
How complete was the follow up data?
Was the statistical analysis appropriate?
Did they control for possible confounding variables?
Was the sample size adequate to detect clinically important differences if they existed?

What was the source population? How were cases and controls defined?
Was there selection bias? Was the 'would' criterion met?
How was information collected? Was it accurate? Was it collected in a comparable way in both groups?
Could there have been recall bias?
Interviewer bias?
Was the statistical analysis appropriate?
Did they control for possible confounding variables?
Was the sample size adequate to detect clinically important differences if they existed?

If so, did they have an independent blind comparison with a reference diagnostic technique, i.e., a "gold standard"?
Was the diagnostic test evaluated in an appropriate group of patients, similar to those you would find in your practice?
Did they address the ability of the test to discriminate between normal and abnormal? How was abnormality defined? Did they calculate sensitivity and specificity or likelihood ratios or report their data in such a way that you could calculate them?