Talk to the Veterans Crisis Line now
An official website of the United States government
IIR 09-082
Implementation and Impact of VA Patient-Centered Medical Home
Elizabeth M Yano, PhD MSPH VA Greater Los Angeles Healthcare System, Sepulveda, CA Sepulveda, CA Funding Period: October 2010 - March 2013 Portfolio Assignment: Systems Modeling, Design, and Delivery |
|
BACKGROUND/RATIONALE:
The Veterans Health Administration (VA) launched its national patient-centered medical home (PCMH) initiative, called Patient Aligned Care Teams (PACT), in 2010, building on growing evidence that the PCMH model of care holds promise for improving quality, patient experience, and provider/staff satisfaction, while lowering unnecessary utilization and costs of care. While evidence outside the VA documents PCMH effects, VA policymakers and managers lack needed information on the relationships between specific VA PCMH features and outcomes among Veterans, and the facility and practice characteristics that promote their adoption. This study capitalized on unique national primary care (PC) organizational survey data that predated PACT implementation, in the context of interdisciplinary consensus development techniques and secondary data on patient experiences and quality of care metrics, for arriving at a VA-focused model of medical home implementation and impact for use in VA policy and planning. OBJECTIVE(S): Our aims have been to evaluate the impacts of existing VA patient-centered The main goal of this study was to examine the state of implementation of VA medical homes prior to PACT implementation, to characterize their key features, success factors and performance, and ultimately to inform evidence-based practice and policy recommendations for ongoing improvement. We used national VA PC organizational and performance data (2007-2010) to: 1) Evaluate the impacts of existing VA PCMH models on quality of care, including which elements are most effective; 2) Determine factors related to successful site adoption of PCMH models and features in VA; 3) Use expert panel methods to integrate and apply findings from the first two aims to develop practice and policy recommendations on evidence-based PC delivery models. METHODS: This mixed methods study incorporated secondary analyses of existing multilevel data (area, organizational and patient-level measures), cross-case analyses to examine patterns of PCMH feature adoption and quality metrics, and expert panel methods to come to consensus on definitional elements and success factors related to a VA-focused framework for PCMH measurement. Secondary data sources included (1) the VA Clinical Practice Organizational Survey (CPOS) (2006-07); (2) the VA National Primary Care Survey (NPCS) (2008-09); (3) the American College of Physicians' Medical Home Builder Survey (ACP-MHB) (2009); (4) VA External Peer Review Program (EPRP) chart-based quality indicators (2007-10); (5) the VA Survey of Healthcare Experiences of Patients (SHEP) data (2007-10); and (6) Area Resource File data (2008). CPOS and NPCS included VAs serving 4000+ PC patients and delivered 20,000+ PC visits (n=250); ACP-MHB surveys included all VA PC practices (n>850), including VA-staffed and contract community-based outpatient clinics (CBOCs). Patient-level EPRP (e.g., diabetes process measures, preventive practices) and SHEP (e.g., patient ratings of care) data were mapped to their respective home sites using substation identifiers. Facility SHEP scores of accessibility, continuity, coordination were also used as PCMH definitional features in selected analyses. We conducted two expert panels using modified Delphi techniques. The first panel identified and came to consensus on core definitional PCMH features vs. (enabling) success factors. We mapped CPOS and NPCS measures to the resulting PCMH features and success factors, and tested for associations with PCMH outcomes. We received supplemental funding to integrate ACP-MHB and later quality data (2009-10) to cross-validate findings and examine alternate specifications of the definitional VA PCMH features. We convened a second panel meeting to generate hypothesized relationships without data, and then underwent structured review and discussion of study results. FINDINGS/RESULTS: The expert panel arrived at seven definitional elements of a VA PCMH (e.g., PC provider supported by a team; practice ensures continuous care by one provider; appropriate use of technology; performance monitoring to improve quality; integrated mental health; ongoing assessment of patients' communication needs; practice ensures accessibility). They also came to consensus on five success factors (e.g., adequate PC provider support staffing; adequate pharmacist staffing; care coordination in the PCMH "neighborhood"; strong quality improvement culture; incentive use to improve quality). Mapping of expert panel derived definitional elements and success factors to existing data sources varied in completeness of construct coverage. Pre-PACT (2007), 6% of VA practices had all seven definitional PCMH features. Overall, VAs with more PCMH features did not have significant performance advantages over those with fewer features, though very uneven implementation of individual features may have limited predictive power. Nonetheless, as hypothesized, many discrete definitional features of PCMH had significant relationships to patient experiences and quality of care metrics. For example, having more explicit or technological mechanisms for notifying PC providers about healthcare events that occurred outside of the practice (e.g., hospitalizations, emergency visits, specialty consults) was associated with better patient-reported continuity, coordination and overall quality. Having mechanisms in place to address patient communication, self-management needs and patient preferences was associated with better patient experiences (e.g., access, coordination and/or overall quality). In contrast, no PCMH features were consistently and positively associated with a preponderance of quality metrics. Instead, some quality metrics appeared more sensitive to PCMH than others (e.g., substance use disorder screening positively associated with PCMH-ness as a whole and with four of seven PCMH features). Organizational factors outside of PC (e.g., academic affiliation and size) and area factors (e.g., region) appeared to be important quality determinants and were, themselves, associated with implementation of different PCMH features. For example, Veterans seen in VAMCs were more likely to report that they received care quickly compared to CBOCs, but rated CBOCs more highly in doctor-nurse communication and overall health care quality. We also noted decrements in patient ratings of access and coordination for every 1,000 more patients seen in VA settings, consistent with larger facility size and greater organizational complexity. IMPACT: Despite historical gains in PC delivery during VA's quality transformation of the mid-1990s, very few VA facilities retained the range of PC features needed to meet PCMH requirements, demonstrating the importance of the PACT initiative. In the absence of full PCMH models, we found evidence of key linkages between individual PCMH features and patient experiences of VA care and VA quality of care metrics that may inform ongoing refinement of the VA PACT model and insights into opportunities for focused improvements in implementation priorities. Contextual factors at the facility and area levels also appear to be critical both to local PCMH implementation and the quality VA PACTs can achieve, and will likely require greater attention. External Links for this ProjectNIH ReporterGrant Number: I01HX000171-01A1Link: https://reporter.nih.gov/project-details/7867808 Dimensions for VA Dimensions for VA is a web-based tool available to VA staff that enables detailed searches of published research and research projects. Learn more about Dimensions for VA. VA staff not currently on the VA network can access Dimensions by registering for an account using their VA email address. Search Dimensions for this project PUBLICATIONS:Journal Articles
DRA:
Health Systems Science
DRE: Treatment - Comparative Effectiveness Keywords: none MeSH Terms: none |
