Christina,
This is a great project. You have honed in on a high risk population that is traditionally underserved/underscreened for metabolic conditions. Couple other factors at play (you can decide where they fit in the behavioral model):
There is an attitude that the psychological issues are so much more important than the diabetes risk, that this risk is inadequately discussed with patients and caregivers at the onset of treatment. Although many may ultimately decide the risk is worth the treatment, few are actually provided the implications of what it means to develop diabetes, or this is heavily downplayed compared to the benefit of the drugs. I have heard from some mental health advocates that they have felt condemned to diabetes without much choice.
The other factor is that management of diabetes is itself, a behavioral health challenge. There are lifestyle choices and other self-management that is made even more difficult by co-morbid behavioral health issues. Some have suggested bouncing back screening and management to the PCP. I would guess PCPs would really dislike getting a bunch of people with behavioral health issues AND diabetes thrown their way (can you imagine in a health center? That's like a quadruple slot appointment with no real uptick in reimbursement). Finding ways to empower psychiatrists to take on the screening, communication with PCPs regarding appropriate treatment if they identify diabetes, but then playing an integral part in behavior modification for their patients to help manage the disease. A solid relationship with behavioral health-focused diabetes educators would also be a good idea for the psychiatrists. CDEs are missing from your community list. I would definitely include them.