A large proportion of patients with coronary heart disease (CHD) do not have a current measurement of their low-density lipoprotein (LDL) levels.1–7 The patient level determinants of not getting a LDL measurement and the outcomes of patients who do not have a current LDL level are unknown, as most studies have not focused on this group of patients. Our objectives were to identify patients at risk for not undergoing a lipid measurement and to determine whether patients without a LDL measurement had higher risk-adjusted morbidity and mortality compared with patients with a lipid measurement. • • • The Veterans Affairs (VA) Ischemic Heart Disease Quality Enhancement Research Initiative (IHDQUERI) is a national initiative to improve outcomes in veterans with CHD by improving concordance with national treatment guidelines. As part of IHD-QUERI, we extracted data on all active primary care and cardiology patients with CHD from 8 VA hospitals in the Pacific Northwest from an existing relational database. This database contains data from the clinical information systems at each of the 8 facilities, including patient demographics, outpatient and inpatient diagnoses, pharmacy records, and laboratory data. Any lipid level measured within 15 months before the October 1, 1998 index date was included in the analyses as a “current” lipid measurement. Although the National Cholesterol Education Program, Adult Treatment Panel III guidelines recommend yearly LDL cholesterol measurements, the 15-month cutoff was chosen to account for the practicalities of clinical care.8 The 15 months before October 1, 1998 were defined as the baseline period. Prescription of any lipid-lowering medication was determined using VA pharmacy prescription data. Patients with a current lipid-lowering medication prescription were defined as those who were dispensed a 3-hydroxy-3-methylglutaryl coenzyme-A reductase inhibitor (statins), fibric acid derivatives, or bile acid sequestrants within the 12 months before October 1, 1998. A current prescription for -adrenergic blocking agents was defined in a similar fashion. Patients were included in the analyses if they were active patients in primary care or cardiology clinics in the Northwest VA hospital facilities and if they had known CHD. The VA facilities were located in Alaska (Anchorage), Idaho (Boise), Oregon (Portland, Roseburg, and White City), and Washington (Puget Sound, Spokane, and Walla Walla). Active patients were defined as being alive on October 1, 1998 and having 1 primary care or cardiology clinic visit per year in each of the 2 previous years. Patients were defined as having CHD if they met 1 of the following criteria: (1) percutaneous transluminal coronary angioplasty or coronary artery bypass graft surgery performed at any VA facility; (2) a hospital discharge diagnosis of myocardial infarction or unstable angina (International Classification of Diseases, 9th Revision (ICD-9) codes 410 to 411); (3) a hospital discharge diagnosis of previous myocardial infarction, stable angina, or other chronic CHD (ICD-9 codes 412 to 414); (4) 1 outpatient visit at a Northwest Network VA facility with a CHD diagnosis (ICD-9 code(s) 410 to 414) in the 12 months before October 1, 1998 and 3 prescriptions filled with a nitrate antianginal medication(s); or (5) a recorded history of percutaneous transluminal coronary angioplasty or coronary artery bypass grafting (ICD-9 codes V45.81, V45.82). We reviewed inpatient discharge diagnoses and procedures from Septemer 3, 1986 to October 1, 1998 and outpatient diagnoses for the 12 months before October 1, 1998. This algorithm was then validated with chart review of 306 patients. Overall, a large proportion of patients met criteria for CHD on 1 level of the identification algorithm. The degree of corroboration for the diagnosis of CHD from chart review ranged from 98% for criteria 4 to 68% for criteria 5, which is not surprising because the purpose of the V-codes is to document procedures occurring outside the treating facility. Using these criteria, 12,135 patients were determined to have CHD and comprised the study population. Patient follow-up occurred through December 31, 2000. Co-morbidities were defined by the following ICD-9 codes from the current year outpatient diagnoses: diabetes (250), hypertension (401), chronic obstructive pulmonary disease (490 to 496, 500 to 505), peripheral vascular disease (440.2, 441 to 443, V434), cerebrovascular disease (433 to 437), congestive heart failure (428), and depression (311). Distance to medical center was calculated using the From the Cardiology and Health Services Research, VAMC, and the University of Colorado Health Sciences Center, Denver, Colorado; and The Ischemic Heart Disease Quality Enhancement Research Initiative (IHD-QUERI), Health Services Research and Development, VA Puget Sound Health Care System, Seattle, Washington. This research was supported by the Department of Veterans Affairs, Veterans Health Administration, Health Services Research and Development Service, Washington, DC, and by Merck & Co., Inc., Whitehouse Station, New Jersey. Dr. Ho’s address is: 1055 Clermont St. (111B), Denver, Colorado 80220. E-mail: Michael.ho@med.va.gov. Manuscript received September 30, 2002; revised manuscript received and accepted December 18, 2002.