These are some general questions to ask when evaluating a paper that reports on a diagnostic test:
- Was the study valid? Was there a clearly defined question?
- Was the presence or absence of an illness or condition confirmed with a "gold standard" test?
- Was the test administered on an appropriate spectrum of patients with the condition?
- Was the "gold standard test" applied to all patients in the study?
- Did the authors of the paper address both the sensitivity (the percentage of people with the condition who have a positive test) and specificity (the percentage of people free of the disease who test negative) of the test?
Reference
Heneghan, C., & Badenoch, Douglas. (2006). Evidence-based medicine toolkit (2nd ed.). Malden, Mass. ; Oxford: BMJ Books/Blackwell Pub.