I've read about SPECT tests used in the diagnosis of depression, besides the mental examination of course. How reliable are they? Why aren’t they widely used yet? And how can I trust a diagnosis that is solely based on a conversation, no matter how long and/or extensive it is and even after ruling out other causes through a physical examination?
Although SPECT is a very useful research tool, it is not helpful in making a diagnosis of depression. (In fact, the only psychiatric diagnosis for which it may be helpful might be Alzheimer's disease).
The problem with SPECT, which is also true for virtually every other biological test ever proposed for diagnostic use, is its poor diagnostic sensitivity and specificity, which refers to its ability to detect true disorder vs. absence of disorder.
The problem is that many individuals with depression lack any SPECT findings and that many individuals with mental disorders other than depression can have positive SPECT findings. This is not such a big problem when a researcher is using SPECT to demonstrate that depressed patients, as a group, have a particular SPECT finding; the problem is trying to use it in the diagnosis of an individual patient.
On an individual patient level, the presence of the SPECT finding doesn't establish the diagnosis of depression (because it could be due to another disorder) and the absence of the finding does not rule-out a diagnosis of depression because some patients with depression lack the finding.
We are hopeful that in the next 10 years technology might advance to the point where tests such as SPECT might be diagnostically useful but we are far from that point today.
Regarding how one can trust a diagnosis based on a clinical interview which is essentially a conversation between the clinician and the patient, clinicians are trained to get careful histories and make observations of the patients appearance and behavior during the interview so that the clinician can feel confident aobut making a psychiatric diagnosis.
In those cases in which the patients in not a reliable historian, the clinician would know to speak to others who know the person well to provide this collateral information.