KL Sloan, VA Puget Sound Health Care System; AE Sales, VA Puget Sound Health Care System; H Sun, VA Puget Sound Health Care System; S Pineros, VA Puget Sound Health Care System; JA Rothendler, Bedford VA, HSR&D Center of Excellence

Objectives: Ischemic heart disease (IHD) is a significant public health problem for which secondary prevention is effective in reducing morbidity and mortality. Although practice guidelines exist, it is not known how well they have already been incorporated into routine clinical practice or at how best to intervene to improve concordance. The Lipid Measurement and Management System project profiled concordance with the IHD secondary prevention guidelines among twenty-five VA medical centers from the VA New England Health Care System (VISN 1), VISN 16, the Northwest Network (VISN 20) and the Cincinnati VA Medical Center.

Methods: Utilizing data from VISN-wide data warehouses, the Austin Automation Center and local facilities’ VistA systems, we identified all currently active primary care or cardiology patients with a diagnosis of IHD. Three measures were used to evaluate guideline concordance: (1) current Low Density Lipoprotein (LDL) results (within the last 15 months), (2) LDL value at goal (<=100mg/dl), and (3) lipid lowering agent (LLA) prescribed. Two facilities were extreme outliers on LDL availability because of identified data problems and were removed from the analysis.

Results: The number of patients identified with IHD was 42,638.Current LDLs ranged from 51.6% to 84.5%. LDLs at goal ranged from 34.3% to 53.2% of the current LDL group and 19.8% to 39.6% of the total sample. Guideline Concordant treatment (defined as LDL<=130mg/dl or being on a LLA) ranged from 86.9% to 96.2% of the current LDL group and 47.1% to 78.6% of the total sample. When we examine the percent of the total IHD population with LDL<=100 (At Goal/Total), facility rankings correlate only 0.69 with ranks obtained using the percent of the total IHD population with current LDLs (Current/Total) and 0.71 with ranks obtained using the percent of the IHD population with current LDLs with LDL<100 (At Goal/Current). However, the ranks of At Goal/Current and Current/Total are essentially uncorrelated (rho=0.04).

Conclusions: Although the distal outcome Percent At Goal correlates with both Current/Total and At Goal/Current, the two intermediate outcomes are not correlated with each other. Therefore, it is unwarranted to conclude that a given facility provides globally poor IHD secondary preventative care based on a single distal outcome. Intermediate outcomes should be profiled prior to planning interventions.

Impact: Clinical outcomes dependent on process performance may be misleading when profiling facilities' treatment of IHD. Administrators and managers should carefully analyze the processes underlying outcomes of interest and ensure that deficiencies in process are accounted for in the assessment of performance on outcomes.