There are many things you can ask but here is a minimalistic list of informative data:
1. How many visits to a psychiatric ER?
“Too many to count“: usually indicates a severe Axis I disorder (psychotic or affective), personality disorder, moderate (or more) panic disorder, somatization disorder, or secondary gain (which can also be part of a personality disorder). Very few or none (esp. in an older patient) usually rule out Axis II.
2. How many of the above (ER visits) were followed by hospitalization?
“Most of them“. Balance tilts toward Axis I. Most of them were not: balance tilts toward Axis II.
3. How many psychiatric hospitalizations total?
Too many to count: reserved prognosis (due to chronicity, dual diagnosis or Axis II comorbidity). Consider non-compliance and dual diagnosis. Just a few/none: good prognosis.
4. What was the chief complaint/reason for most of the hospitalizations/ER visits?
Great screening question. Voices or paranoia versus depression with suicidal ideation will greatly focus further inquiry.
5. What was the longest hospitalization?
I started to appreciate this question after one patient told me his been hospitalized only twice before. I was puzzled and had a difficult time understanding the mismatch between his level of dysfunction (which was pretty significant) and the very low number of hospitalizations. It turned out that his second hospitalization was 12 years long, in one of the state forensic facilities. So, while the total number of hospitalizations is informative, what really matters (when it comes to prognosis and understanding dysfunction is the total duration of the time one spend inpatient).
6. What medication(s) worked the best/was the worst?
Good question to further narrow the differential. Also, it will save you time to cross off the list ineffective/poorly tolerated medications. Futher, usually there is no need to reinvent the wheel: if a specific intervention worked once, change is that it will work again (the exceptions here is a medication tachyphylaxis reaction, more commonly known as “poop out” phenomenon, where a medication that is effective stops working)
7. Is there a history of noncompliance? Why?
Two benefits: 1. Gives a good measure of the patient’s insight. 2. Places the patient on the stages of change continuum and thus allows you to tailor your intervention accordingly.
8. Were you using drugs around the times when you got hospitalized?
If the answer is”yes, time and again” you should re-asses what’s primary versus secondary.
9. Tell me what’s the most violent you’d ever been.
Open ended increases the chance to get a valid answer.
10. Did you ever find yourself wishing to be dead?
Suicidality is a sensitive topic. Go straight for a past history of attempts you might put the patient off. Also, your range is too narrow. Start with a broad question and follow up with more specific questions (use a concentric approach) and you will get more honest and comprehensive (thus also more valid) answers.