Yes, you could potentially "solve the problem." Unless there is a test or a treatment that is easy and effictive enough to use empirically it is much harder to

Yes, you could potentially "solve the problem." Unless there is a test or a treatment that is easy and effictive enough to use empirically it is much harder to diagnose that person. Physicians have to balance quality of life improvement for patients along with treatment. There may be a trade-off between efficacy rate of treatments and the likelihood of adequate compliance. People may not change their diets or take their meds appropriately to improve their health for example. If you can't rely on the patient you might never find the solution. You may just be better off at improving their quality of life even if it doesn't ultimately solve their underlying condition.