Evaluation of the SAMHSA Primary and Behavioral Health Care Integration (PBHCI) Grant Program: Final Report (Task 13)
Apr 21, 2014
Serious mental illness (SMI) — which includes schizophrenia, bipolar disorder, and major depression — affects more than 4 percent of adults in the United States. Persons with SMI often carry the added of burden of physical health problems. Compared with adults without mental illness, people with SMI have higher rates of chronic disease, such as hypertension, diabetes, and obesity, and are more than twice as likely to die prematurely from these conditions, in part because they experience significant barriers to accessing primary and preventive medical care.
Behavioral health care settings are a trusted site in which people with SMI can seek care. Thus, integrating primary care with behavioral health care offers a promising strategy for improving this population's access to a fuller range of health care services. To promote better integration of behavioral health care and physical health care, the Substance Abuse and Mental Health Services Administration (SAMHSA) developed the Primary and Behavioral Health Care Integration (PBHCI) program, which is intended to offer primary care to adults with SMI in community mental health centers and other community-based behavioral health settings. Grant recipients received up to $500,000 annually to develop integrated services that include four core features: (1) screening and referral for physical health care, (2) a tracking system for consumers' physical health needs and outcomes, (3) care management, and (4) prevention and wellness services.
SAMHSA commissioned RAND to evaluate the program's success in integrating services and improving consumer health. Over the course of three years, the RAND team examined 56 PBHCI programs, using site visits, a web-based survey of providers, and physical health screening of consumers. RAND's evaluation addressed three questions:
The evaluation assessed the integration of behavioral health and primary care services along four dimensions:
These four scores were averaged into a score for overall integration.
The results showed considerable variation in the extent to which programs succeeded in integrating primary care and behavioral health care services. On average, programs were most successful in co-locating services and in creating shared structures and systems, and they struggled most with creating an integrated practice culture. Overall, the majority of programs were judged to be moderately integrated across all measures. Barriers to integration reported by many programs included the challenge of creating integrated health records; long-term financial sustainability, particularly for non-billable services (such as wellness programs); recruiting and retaining qualified staff; and engaging consumers in integrated services over time.
Integrated programs reported early enrollment success. More than half of consumers used basic integrated services in the first year, including screening for physical health problems, primary care, and case management. However, rates of service use were not different for consumers with and without an identified physical health care need, suggesting that programs were not necessarily successful at targeting consumers likely to benefit most from integrated care. Consumer enrollment after the first year was often lower than expected.
To understand the effects of integrated care on consumer health, the RAND team compared changes in physical and behavioral health indicators among consumers at three PBHCI sites and three comparable control sites. The results showed that consumers treated at PBHCI clinics experienced greater improvement in some indicators related to diabetes, dyslipidemia (cholesterol), and hypertension compared with consumers treated at control sites (see table). There was no similar improvement in indicators related to obesity and smoking. The smoking result provided evidence that smoking-cessation programs at PBHCI clinics needed improvement; steps have already been taken in this direction.
|Chronic Condition||Indicator: PBHCI Sites Compared with Control|
|Diabetes||Hemoglobin A1c (a measure of blood sugar)|
|Plasma glucose (a measure of blood sugar)|
|Hypertension||Diastolic blood pressure|
|Systolic blood pressure|
|Obesity||Body mass index (BMI)|
|Smoking||Self-reported smoking status|
NOTE: Green shows greater reductions in risk indicators among consumers at PBHCI clinics; yellow shows no difference; red shows a greater reduction in risk indicators among consumers at control clinics. HDL = high-density lipoprotein; LDL = low-density lipoprotein.
The analysis found no clear connection between integrated care and most behavioral health outcomes. This result was somewhat expected because PBHCI did not target behavioral health issues and because most consumers were already in behavioral health care prior to enrolling in PBHCI.
Indicators for select chronic conditions — diabetes, high blood pressure, and high cholesterol — improved when consumers received integrated care.
Consumers were more likely to access integrated services when...
The final phase of the evaluation examined whether particular program features were associated with improved consumer access to services or improvements in health. Three of the dimensions of integrated services described earlier — co-location of services, integrated provider practices, and staff perceptions of belonging to a team — were associated with greater consumer access to care. Shared structures and systems, such as electronic health records, were associated with lower access to care, a counterintuitive result that may reflect reduced consumer "face time" with providers or program focus on infrastructure development at the expense (at least early in the program) of clinical care. In addition, consumers demonstrated greater access to care if programs offered more days per week of primary care and held more regularly scheduled meetings between integrated staff teams.
The PBHCI programs succeeded on several fronts, most notably in building integrated, multidisciplinary teams and practices that offered an array of services to a clientele whose health outcomes improved in several areas. The programs encountered difficulties in other areas, most notably maintaining consumer enrollment and engagement over multiple years, implementing wellness programs and shared information systems, and planning for the financial sustainability of the program beyond the grant period.
To help programs build on their success and address these challenges, the RAND team recommended several steps for improvement:
RAND is conducting additional evaluation activities to quantify program costs and service use and to compare the effectiveness of alternative models of integrated care. This work should further clarify the value of integrated care for improving the physical health of adults with SMI.