I want to compare the accuracy of two diagnostic modalities, which research design is appropriate to accomplish this task?
Hi Fahmi,
The diagnostic platform that is accepted as the gold standard varies by application; you have not provided enough information. In some cases, an experimental design can be utilized that uses a combination of reference tests when there is not a perfect gold standard in order to minimize biases of a suboptimal "gold" standard. Calculation of of sensitivity, specificity, positive- and negative predictive values are straightforward when 1) there is an accepted gold standard and 2) the gold standard performs as well as or better than the test method. However, innovative diagnostic tests often perform better than the accepted "gold standard." There are a number of publications that address the appropriate research design when there is or is not an accepted gold standard. Here are some publications that you may find useful:
Article Evaluation of diagnostic test when there is no gold standard...
Article Diagnostic tests for pneumonia in ventilated patients: Prosp...
Best regards,
Marcus
Dear Fahmi Khan. I agree with Marcus Crim and could add that you should be aware that if you compare your test with the gold standard in a case-control design, you can not show that your new test perform better than the gold standard because the gold standard defines the truth. If none of the tests you like to compare are gold standard method you can use the gold standard as reference in a case-control design. Otherwise a cohort study would be the appropriate way to show that your test predicts outcome better than another test. Additionally, you can compare other relevant aspects than the clinical performance. That is parameters like analytical performance and cost-effectiveness.
- Badges
- Science topic
- Similar topics
- Biological Science
- Physiology
- Physiological Processes
- Stress