To understand how current systems work, you need to understand where current medical records come from, so I'd recommend you read that part first, if you haven't already.
Current electronic healthcare record systems (EHR) vary a bit in the features they have, but except for a few frills here and there, they largely automate the sequential recording of the patient's history as I described in the history of the medical record. This design leads to an unproductive, work intensive, and faintly ridiculous workflow which I describe here.
As the patient enters the surgery, the first thing the doctor needs to know is why he is there. If he enters horizontally, leaving blood tracks all along the corridor, and clearly has a knife sticking out his back, this isn't to hard to figure out, but in general, it's more subtle than that. Most EHR systems have no trouble showing a one-line description of the stated cause of the visit. Sometimes it's even correct.
The next thing the doctor needs to figure out is which confirmed diseases the patient has, and which potential diseases we're still working on. Now, here we run into a wall. Most EHR systems have no list of diseases of any kind. Hard to believe, I know, but check for yourself if you don't believe me.
In order to figure out what diseases the patient has, the doctor now needs to read through the entire recorded medical journal. If it contains just a few visits or documents, this can be done, but since most patient records contain hundreds of notes, each ten or more lines of text, this is impossible. In most of these cases, one or more diagnoses are simply missed, or if one is lucky, one can ask the patient which problems he has and get a much more reliable answer.
In short, what the doctor is doing during the first few minutes of a visit is to run through the entire process of taking over a patient. That process used to take a number of hours spread out over two years in the olden days and just an hour or two in the less olden days, but now we're expected to do the same in just a few minutes. Actually, it's a lot worse, since with the interconnected systems we have today, the amount of information in the medical record has grown many times larger than it ever were before. Not necessarily containing more real information, but the information is now buried in masses of repetitive and often useless prose.
By now, the doctor has largely read through the last two, maybe three encounters and at least gathered what's been going on lately. He has absolutely no guarantee he didn't miss some really important diagnoses that weren't touched upon during those last two or three encounters. To increase the chances of finding other important diagnoses, he then scans the medication list to see if he finds anything there that can clue him in to something important, but this is a very indirect and unreliable way to learn about the patient, since there is no one-to-one correspondence between pharmacological treatments and diseases. But it's all we have, so we chance it.
Finally, we may scan the list of ICD-10 diagnoses, but in most systems this list is utterly useless. It usually contains only a fraction of the relevant diagnoses, and those it contains it usually contains a large number of repeats of. For entire classes of problems, it doesn't contain anything at all. Don't forget, the ICD-10 diagnoses aren't even intended to encode the actual relevant diagnoses of the patient, they're there for budgetary reasons. That is entirely something else.
The last step is then to discreetly check with the patient himself to see if he can tell us something more, something we missed in the records, and that is usually the case. And, of course, we look him over discreetly to see if we find any obvious surgical scars and such, and sometimes we do.
This whole process has one difficulty and one danger. The real danger is that we miss an important diagnosis that is there in the records, buried somewhere in a swamp of excessive information. The difficulty is to remain detached and looking entirely at ease while trying to make the patient tell us all those things we have no chance of finding out from our medical records within the alloted time.
So far, we've come just a couple of minutes into the visit. Now we're going to ask the patient how he feels, do the physical exam, and decide what to do next.