To date this is the first study to comprehensively assess quality of care for CVD-related risk factors in both depression and SMI compared to a non-psychiatric control group. In contrast, prior studies did not control for the effects of race or comorbidity, or primarily focused on care for patients with SMI only13
and did not compare quality of care to those with depression, one of the most common diagnoses seen in primary care settings.
Overall, we found that the majority of VA patients in the EPRP FY 2005 sample received adequate quality of care. Our overall findings reflect prior studies,29,30
which found that nationally, the VA performed better than non-VA healthcare organizations on performance measures for hypertension (78% vs. 64%), hyperlipidemia (64% vs. 53%), and diabetes (70% vs. 57%) than did non-VA care settings.30
Differences in quality of care in the VA have been attributed to organizational attributes not universally found in non-VA settings, including electronic medical record use, single payer structure, and widespread application of performance measures to benchmark quality of care nationally.30
Quality of care for intermediate clinical outcomes (blood pressure, low-density lipoprotein, and HbA1C levels) was similar among patients with SMI, depression, or no psychiatric diagnosis, reflecting findings from previous VA studies.12
Nonetheless, we found that patients with SMI were the least likely to receive adequate processes of care for CVD-related conditions compared to those with depression or no psychiatric diagnosis, notably for hyperlipidemia screening and routine follow-up assessments recommended for diabetes (foot exam, retinal exam, and renal testing). These results remained robust even after controlling for race, the interaction between race and diagnosis, and comorbidity.
These observed gaps in processes of care for patients with SMI were also clinically significant, as these patients were 32–42% less likely to receive care for these conditions. With the exception of renal testing, effect sizes for differences in quality between depressed and non-psychiatric patients were smaller, indicating that they were less clinically significant than differences observed for patients with SMI.
For patients with SMI, inadequate processes of care (screening and follow-up care for CVD-related conditions) over time can be problematic because lack of medical attention could lead to adverse medical consequences (e.g., vision problems, amputations)8
Moreover, the percentage of patients with SMI and poor intermediate CVD risk-related clinical outcomes is likely an underestimate, given that widespread use of atypical antipsychotic medications did not occur until after 2005.
Our observed gaps in processes of care for CVD-related conditions among patients with SMI could be attributable to patient, provider, and/or system-level factors, including location of care, lack of coordination and stigma.29,30
Potential disparities in processes of care may be greater for patients with SMI than for those with depression because they are more likely to be exclusively treated in mental health settings. Our findings regarding processes of diabetes care contrast with earlier EPRP-based research suggesting that quality of care for diabetes was comparable among VA patients with and without SMI.11
In contrast to our study, prior studies were based on past EPRP cohorts (e.g., 2001–03), which only included patients receiving care from outpatient general medical clinics and not from mental health outpatient clinics. In FY2005, the EPRP sampling frame was expanded to include mental health outpatients, where the majority of patients with SMI receive care.15
Many patients with SMI consider the mental health clinic their “home” site for care because of the need for more intensive psychiatric services compared to unipolar depression.15
The availability of general medical services may be limited in mental health settings because mental health providers may be less able to perform screenings, provide integrated care in mental health specialty settings, or assist in following up on referrals to medical specialty care.32
Moreover, patients with SMI may have trouble with navigating care across different providers, particularly for general medical services and follow-up care.32
In addition, providers in general medical or specialty medical care (e.g., endocrinology, ophthalmology) may not have the experience, training, or time to accommodate patients with serious mental illness.33
Consistent with our findings, prior evidence suggests that quality of care for certain conditions that require follow-up with specialists or additional procedures was worse for SMI compared to non-SMI veterans, notably for procedures and medications related to myocardial infarction or coronary artery disease.14,31
The stigma of mental illness, particularly SMI, may also preclude patients from seeking care outside the comfort zone that exists in the mental health clinic.
Despite our use of a national sample on quality of care for patients with and without mental disorders, there are limitations to this study that warrant consideration. First, the cross-sectional nature of the analyses preclude us from examining trends in quality of care over time or causal effects of diagnosis on quality of care. Second, EPRP may have excluded some patients with SMI, as those who were exclusively seen within the VA homeless outreach or mental health work therapy programs were not included in the EPRP sample. Nonetheless, excluding these individuals may have led to an underestimation in the gaps in quality of care for VA patients with SMI. In addition, we were unable to control for clinical factors that might influence care-seeking in persons with SMI, including current psychiatric symptoms, or factors that might influence access to medical services, such as the organizational features of general medical services in mental health clinics. The exact location of service for which quality of care was ascertained (e.g., within mental health versus general medical clinics) was unavailable in the EPRP dataset, and hence, we were unable to determine whether quality varied by treatment setting. Nonetheless, we feel that for quality to improve for veterans with SMI, an overall facility-level quality measure is necessary in order to hold all providers at that facility accountable for improving quality of care for veterans with mental disorders, regardless of the location of their care. Finally, the focus on VA patients may potentially limit the generalizability of our findings.
Overall, our results suggest that the quality of care for CVD-related chronic conditions is suboptimal among patients with SMI, particularly for care involving contacts with medical specialists. Providers and healthcare leaders should consider efforts to improve access and continuity of medical treatment for persons with SMI. Promising treatment models that have been found to be effective within the VA setting include co-location of general medical providers in mental health clinics34
and the Chronic Care Model adapted to address medical care self-management for persons with chronic mental illness.35
The VA’s national infrastructure and electronic medical record system can facilitate the further dissemination of these treatment models across a wider range of VA facilities. Outside the VA, additional efforts are needed to address the administrative and financial separation of mental health and physical health care in order to improve integration of care for persons with SMI. State-level efforts (i.e., Medicaid) to allow for reimbursement of medical services in mental healthcare settings and behavioral management of CVD risk show promise in facilitating access to medical care for vulnerable persons with mental disorders. Still, the stigma of mental illness is a common barrier across VA and non-VA settings, and can especially lead to delays in medical care-seeking and follow-up for persons with SMI. Efforts to address the stigma barrier include improved education of general medical providers in managing patients with SMI, coupled with additional education for patients on managing CVD risk.