I have the following problem. I want to compare a self-diagnostics tool (questionnaire based) with a gold standard diagnostic tool. However, I have the following issues:
1- The gold standard is not actually gold because it is somewhat subjective. The physician can either consider the case positive or borderline, borderline or negative based on their own judgment. So, for example what I can consider a positive case, another physician may still consider it not that of a positive and it s/he can call it a borderline case. Both will still be considered fine.
2- The test is repeated every two weeks over a course of 30 weeks and of treatment and 8 weeks after treatment. Therefore, the patient can develop the condition at any stage of the treatment.
3- The questionnaire is applied at the same time of diagnostic using the 'gold standard'.
4- The patient may self assess to be positive or borderline or negative while the gold standard can give a different result.
Considering the above. What is the most suitable design and analysis?! Since, I am interested in the stage where the patient can develop the condition using both diagnostic tools. I feel that the best design is time-to-event survival analysis. What do you think?