EDU 08-414
Interprofessional Training for Improving Diabetes Care
Susan R Kirsh, MD Louis Stokes VA Medical Center, Cleveland, OH Cleveland, OH Funding Period: October 2009 - March 2014 Portfolio Assignment: Education |
BACKGROUND/RATIONALE:
The persistent gap between actual care and best practice for complex chronic diseases such as diabetes is a major challenge for the healthcare system. Despite recognition that quality care for such diseases is best provided in an interdisciplinary manner, the acute care-oriented healthcare system fosters separate silos decision making. Training models for interprofessional practice are limited, as is understanding of linkages between interprofessional training, actual practice, and patient outcomes. This study assessed the impact of an educational intervention to orient primary care providers towards interprofessional practice. Shared Medical Appointments (SMAs) constitute a delivery system redesign. Groups of patients (10-20) are seen together by a 2-6 person interprofessional team for an approximately 90 minute appointment. The "ReSPECT" (Role modeling in Shared medical appointments to Promote Establishing Collaborative Teams) intervention modeled interprofessional practice using demonstration SMAs. We assessed the intervention's effect on primary care providers and diabetic (DM) patients at Ohio CBOCs by cluster-randomized clinical trial. OBJECTIVE(S): The trial was designed to assess the educational intervention's impacts on i) clinicians attitudes towards interprofessional practice, diabetes treatment self-efficacy, perceptions of their interprofessional practice, utilization of other professionals, and clinical inertia, and ii) patients' glycemic control and perceptions of, and satisfaction with, care coordination, as well as iii) how such impacts occur. METHODS: a. Design and Setting. ReSPECT was designed as a matched cluster-randomized trial at 31 Ohio CBOCs, with CBOC-associated clusters of providers and patients randomized to ReSPECT or a telemedicine control intervention within each of 14 CBOC-strata defined by regional network (Chillicothe, Cincinnati, Cleveland, Columbus, Dayton) and diabetic patient load. Primary analysis uses mixed statistical models. b. Participants. All clinicians in participating CBOCs were eligible for the study. DM patients were identified through the VA Diabetes Registry. Patients were considered eligible if living, without a guardian flag on the medical record; had made two or more visits to a primary care stop code in the most recent 13 months; had a 250 or 249 ICD-9-CM diagnosis code or had a diabetes-specific medication dispensed twice since October 2001; and had a last recorded hbA1c 6.0. c. Intervention. The intervention consisted of 1-3 on-site SMAs, conducted by an experienced inter-professional SMA team from the Cleveland VAMC with providers and DM patients from each intervention CBOC primary care provider's (PCP) patient panel. These were followed three months after the last SMA, by provision of an SMA manual developed to reinforce the interprofessional practice skills the SMAs modeled. We hypothesized this would improve interprofessional practices and overall quality of care. Control CBOCs received no practice intervention. d. Study Measures. Quantitative Clinician Scales Attitudes: Team Skills, Attitudes Toward Health Care Teams. Self-Efficacy: Provider Abilities (subscale from the MW Clinicians' Network). Perceptual: Team Climate Inventory, Physician Comfort with Delegation. Clinical Behaviors Engagements of non-MD professionals. Insulin starts. Patients on maximal oral hypoglycemic therapy with A1c>cut points of 8 and 9. Patients Perceptual: Patient Assessment of Chronic Illness Care as modified for Diabetes (PACIC-DM): overall score and 11 subscales, especially follow-up/coordination. Biomarker: Most recent hbA1c. Changes in hbA1c. Qualitative Pre- and post-intervention structured interviews with physician and patient subsamples. FINDINGS/RESULTS: Quantitative Data are still accumulating (see Section 5), especially on clinical results. Thus, results below are preliminary and confined to selected attitudinal and perceptual outcomes from questionnaires and interviews. Pre-intervention patient surveys were conducted at each CBOC. For administrative reasons, Chillicothe and Columbus-affiliated CBOCs did not receive the intervention, nor were post-intervention surveys conducted for these or Dayton-affiliated CBOCs (see Comments for Central Office Review). PACICs with over half their items omitted (<5% of questionnaires) were excluded from analysis, and missing items otherwise singly-imputed from completed items using fully conditional specification (SAS PROC MI; multiple imputation will be used for final analyses). Forty-eight percent of the 2129 pre-intervention PACICs analyzed from all CBOCs, as well as of the 2880 post-intervention PACICs from all Cleveland and Cincinnati-affiliated CBOCs and of the subset of these from whom pre-intervention questionnaires were also available, were from patients of intervention CBOCs. Respondents from the two treatment arms were similar demographically and similarly distributed among CBOC parent facilities. PACIC total and follow-up/coordination scales in the pre-intervention sample were very similar, and other PACIC and 5A subscales relatively similar, to those of male subjects in the initial PACIC primary care validation sample of diabetics. Mean ( SD) pre-intervention PACIC total and follow-up/coordination scales of 3.11 1.06 and 2.83 1.12 were virtually identical for intervention and control site patients (total: 3.12 vs. 3.10; follow-up/coordination: 2.84 vs. 2.81), but somewhat favored intervention CBOCs among those later surveyed post-intervention (total: 3.22 vs. 3.10; follow-up/coordination 2.91 vs. 2.83). Mean post-intervention total and follow-up/coordination scores in the Cleveland and Cincinnati CBOCs were 3.22 vs. 3.16 for intervention and control total and 2.94 vs. 2.89 for the follow-up/coordination scale. Preliminary mixed linear model analysis (SAS PROC MIXED) does not show statistical significance. Mean pre- to post-intervention changes for respondents at both times were small and favored controls, suggesting ReSPECT improves these PACIC-DM scores only modestly at best. Qualitative Most patients reported that communication and trust significantly impacted satisfaction with the care they received. Most clinicians cited time constraints as a major challenge impacting ability to provide patient education, as well as communication between physicians. Post interviews revealed that the implementation of PACT helped improve communication between clinicians and positively impacted staff cohesion. Clinicians at intervention CBOCs described SMAs as being educational for staff as well as patients. IMPACT: Our early results with SMAs have contributed to the promotion of their use by Patient Care Services Office of Primary Care as part of VHA Patient Aligned Care Team implementation. This project has provided some external validation of the experience at PI's home site (and the SMA Manual created by the home site team that has been made available to all) related to patient and provider acceptance, In addition, it informed the design of a VISN-wide collaborative to implement SMAs, particularly for chronic disease. External Links for this ProjectDimensions for VADimensions 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, Diabetes and Other Endocrine Conditions
DRE: Diagnosis, Treatment - Observational, Treatment - Comparative Effectiveness Keywords: Diabetes, Education Research, Outpatient MeSH Terms: none |