BN Doebbeling, The University of Iowa Colleges of Medicine and Public Health; Departments of Internal Medicine and Epidemiology; TE Vaughn, Department of Health Management and Policy, University of Iowa College of Public Health; MM Ward, Department of Health Management and Policy, University of Iowa College of Public Health; WR Clarke, Department of Biostatistics University of Iowa College of Public Health; RF Woolson, Department of Biostatistics University of Iowa College of Public Health; LG Branch, The Duke University School of Medicine; E Letuchy, The University of Iowa College of Medicine; Department of Internal Medicine; BJ BootsMiller, Program in Health Services Research, Iowa City VAMC; T Tripp-Reimer, The University of Iowa College of Nursing; BA Sorofman,The University of Iowa College of Pharmacy

Objectives: The U.S. Preventive Services Task Force (USPSTF) publishes clinical preventive services (CPS) recommendations and supporting evidence. The objective of this study was to examine consistency of delivery of 11 CPS recommendations from the USPSTF within VAMCs system-wide and to identify institutional factors associated with high rankings.

Methods: The 1998 Veteran Health Survey (VHS) is a stratified national survey of 300 men and 150 women with >= 1 primary care visits in the prior year from 138 VAMCs. Eleven CPS outcomes were calculated as the weighted proportion of men and women who appropriately received the service within the recommended time interval. Receipt within the VA, outside the VA, or both counted towards facility-level adherence rates. Each outcome was standardized across facilities. All facilities were then assigned percentile ranks for each outcome measure.

Two different approaches to ranking facilities were explored. 1) A mean Z score was calculated across the set of 11 Z scores for each facility. The means and standard deviations for this set of 11 Z scores were ranked compared to all facilities. "Best practices", or "consistently good" performance, was defined as a mean Z score in the upper tertile of means and the lowest tertile of standard deviation, whereas "room for improvement" was defined as the lowest tertile Z score for means and the highest tertile for standard deviations. 2) A percentile rank, based on different a priori-established percentage levels was used to rank order each VAMC nationwide. Different cutoffs (50%, 60%, 70%, 80%, and 90%) were explored; facilities with the highest number of percentile ranks over 80% (5- 7 out of a possible 11) and the lowest number of percentile ranks less than 50% (0-4 out of a possible 11) were defined as the "best practices" group. A "quality of fit " test determined consistent.

Results: VHS data included 39,939 surveys (27,597 men, 12,342 women) and a 67% response rate; women were oversampled (unweighted=31%, weighted=4%). Sensitivity testing using different percentile cutoffs in, demonstrated that certain facilities perform consistently better than others, regardless of the cutoff. For each cutoff assessed, the "quality of fit" test p-value was <= 0.001. Twenty facilities were classified as "best practices" facilities by the Z score approach and 15 by the percentile rank. Nine were classified as "suboptimal practices" facilities by mean Z score and 13 by percentile rank. The regression model (adjusted R-square 0.28) to predict the highest mean Z score included: FTE trainees + residents (B= 0.0008, p=0.0602); special and non-special ventures (B=0.0701, p=0.095); rural location (<100K population) (B=0.385, p<0.0001), and VISN (B 0.442 to –0.439, p <0.02 to <0.001).

Conclusions: Adherence to CPS guidelines varies across facilities throughout the VA system with some facilities and VISNs performing consistently well, suggesting that differences in CPG implementation may be responsible. Alternatively, patient population characteristics including ethnicity, age, and migration varying by facilities may effect ratings.

Impact: Further study is needed to identify institutional and provider-level factors important in facilitating effective CPG implementation. Network administrators and clinicians should focus on facilitating the implementation of CPS.