Examining the Validity of a Model of Integrated Healthcare Knowledge

Background and Purpose 

The U.S. health care delivery system fails to provide high quality care for individuals with comorbid physical and mental health conditions.  With approximately 17% of the U.S. adult population having comorbid mental and physical health conditions within any 12 month period (Druss & Walker, 2011), an uncoordinated system of separated mental and physical health care results in poor outcomes, higher utilization of health related services, and increased cost (Petterson et al., 2008; Simon, Ormel, VonKorff, & Barlow, 1995). In order to improve quality of care, the U.S. healthcare delivery system is transitioning to an integrated delivery system where physical and mental health providers are co-located, services are coordinated, and screening measures and proactive care strategies are focused on early identification and treatment of comorbid conditions (Berkman, 1996; Epping-Jordan, 2005). 

Workforce barriers impact the transition to integrated care.  Simply transitioning specialty care behavioral health clinicians to integrated settings is ineffective due to lack of knowledge essential to practice in integrated settings and poor skill fit of specialty care practices to primary care settings (Blount & Miller, 2009; Richardson et al., 2001).   Further complicating this transition is the limited evidence identifying the essential skills needed to deliver care given the variation in current integrated models and settings (Aitken & Curtis, 2004).  This study aims to fill a gap by testing the extent to which domains of knowledge identified in, and the relationships between the domains of a model of integrated healthcare knowledge are validated by behavioral health providers working in integrated healthcare settings.   

Methods

This study used a survey design to examine the extent to which specific domains of knowledge related to medical and psychiatric diagnoses, screening measures, and interventions are validated by 154 behavioral health providers working in integrated healthcare settings using confirmatory factor analysis (CFA) models. 

Results

Provider agreement for medical diagnosis items ranged from 81.8% for metabolic syndrome to 96.9% for obesity, while agreement for psychiatric diagnostic categories were either 99.4% or 100%.   Agreement with screening measure items ranged from 87% for cholesterol and hemoglobin A1c to 98.7% for mood disorder, anxiety disorder and substance use measures.  Interventions agreement ranged from 90.3% for medical medications to 100% for brief therapeutic interventions. The individual knowledge domains were confirmed by CFA models in the sample.  The final CFA model to examine the relationships between the total scores on the subscales and the extent to which these scores loaded on the latent construct,  “integrated healthcare knowledge” suggested a close fit: X²(1) = 0.85, p > .05; CFI = 1.0, TLI = 1.0; RMSEA < .001, 90% CI (< .001, .21), PCLOSE = .44; and SRMR = .06. 

Conclusions and Implications

The study findings provide further support that behavioral health providers working in adult integrated healthcare settings require knowledge of specific medical diagnoses, psychiatric diagnoses, medical and mental health screening measures, medications, levels of care, lifestyle interventions, and brief substance abuse and therapeutic interventions.