This is a common question. I think the answer basically depends, as you've seen, on the patient's pathology, how many patients you have, and how much time you have.
I can speak to what I do personally.
When I come in to pre-round in the AM, I basically do a focused, informal interview that usually lasts no longer than 5-15 mins. Taking any more time than this won't allow me to get to rounds on time without at least having eyeballed my patients.
To break it down, I usually start with a greeting, and ask how they slept that night. Sleep is very important and will seriously hinder improvement if i'ts not going well. Oftentimes, the patients are sleepy or not even awake. Obviously, I'll return to see them later. Assuming you're doing your interviews when they're awake (i.e. after breakfast) or during the middle of the day, when they're theoretically awake, you could go on to ask about the acute problem that brought them into the hospital. For a schizophrenic, you could let them talk for a couple minutes to get an idea of their thought process, organization, thought content, and the like. You could then ask pointed questions about salient aspects of their disease, such as A/V hallucinations, affective symptoms, etc. To be thorough, you should always ask about medications, and their side effects, and conduct a truncated physical if necessary. For example, if you recently increased Haldol from 10-15mg daily, or if the person received a prolixin decanoate shot 3 days prior, it is good to ask about stiffness, dystonia, akathesia, and if necessary, assess for these things physically.
For a depressed patient, one would ask about their mood, suicidal ideation, if they have plans for the future (forward thinking assessment is useful) and how they feel they are doing themselves.
Your example of the demented patient might garner varying responses. It might depend on why the patient was admitted, and what your proposed end point would be. If the person is in end-stage Alzheimer's, and was sent from thh nursing home for intractible screaming, then you'll know right away when your medication regimen change has kicked in, or if a physical problem you have discovered is improving, both of which will effect the presenting problem.
I personally do not do a MMSE on the same patient every day. At other hospitals in which I rotate, they do MMSEs on all types of patients - even patients for which it was not intended, all the time (which sort of drove me crazy). There is a practice effect for these things, so there is also a diminishing return on the end score meaningfullness of the MMSE performed every day. If you're tracking a change in mental status which is acute, a daily MMSE isn't a bad idea necessarily, but I'll say that many docs will perform a less-standardized, more informal version of this when they see the patient. I've known plenty of patients with MMSEs of 22 who have lived independently for years with stable scores, save for the occasional bout of depression, for example.
Hope this helps.
