CS Rosen, HSR&D Center for Health Care Evaluation, VAPAHCS and Stanford University; HC Chow, HSR&D Center for Health Care Evaluation, VAPAHCS; MM Greenbaum, HSR&D Center for Health Care Evaluation, VAPAHCS; JF Finney, HSR&D Center for Health Care Evaluation, VAPAHCS and Stanford University; RH Moos, HSR&D Center for Health Care Evaluation, VAPAHCS and Stanford University; JA Yesavage, VISN 21 Mental Illness Research Education and Clinical Center (MIRECC), VAPAHCS and Stanford University

Objectives: To assess adherence to dementia practice guidelines among providers in VISN 21. These data are used to identify targets for quality improvement efforts by the Sierra-Pacific Network Mental Illness Research, Education and Clinical Center (MIRECC) and to provide a baseline for evaluating the impact of the MIRECC on clinical practice patterns.

Methods: Survey of clinicians. Setting: Six United States Department of Veterans Affairs medical centers. Participants: 200 registered nurses, primary care physicians, social workers, neurologists, psychologists, nurse practitioners, and psychiatrists treating patients with dementia (85% response rate). Measurements: Self-report survey assessing the proportion of patients for whom clinicians provided assessment, education, support, and treatment practices recommended in the California Workgroup Guidelines for Alzheimer's Disease Management.

Results: Recommended practices that were performed most consistently were neurological exams, obtaining reports from caregivers, informing the family of the diagnosis, addressing durable power of attorney, and mandated reporting of drivers with dementia. Practices that were often consistent with guidelines but could be further improved included standardized cognitive assessment, discussion of patient care and decision-making issues, and reporting of elder abuse. Relatively few providers routinely performed recommended caregiver support practices, with the possible exception of support group referrals. Providers prescribed donepezil or other anti-dementia medications for one-quarter of their dementia patients. Choices of medications for mood and for behavior problems were generally consistent with practice guidelines.

Conclusions: These findings highlight areas of convergence and divergence with practice guidelines and suggest priority areas for VISN 21 MIRECC quality improvement efforts. Further research is needed to understand reasons for specific gaps between guidelines and practice, particularly in the area of caregiver support.

Impact: These findings are an example of using practice data to inform TQI efforts. This study highlights areas of strength and potential opportunities for improving care for veterans with dementia, provides baseline data for evaluating the effect of quality-improvement efforts over time, and suggest important issues to be explored in future translational research.