In this issue of Diabetes Care, Massing et al. (1) present the results of their review of Medicare claims from 13,660 diabetic patients who received regular outpatient care from a primary care physician (n = 1,749). During a 24-month period, 31% received no lipid profile, 24% received only one lipid profile, and 45% of the diabetic patients received two or more lipid profiles. Further analysis revealed that Caucasians compared with African Americans were 1.6 times more likely to receive a lipid panel, and patients with stroke or heart failure were also less likely to receive a lipid profile.
There are clinical decisions pertaining to lipid screening made by the primary care physician that are not captured with Medicare claims data and may partially account for an under-representation of the adherence rates. One situation is a patient’s refusal to obtain a lipid profile. Another situation, which is rare, is the individual with a total cholesterol <100 mg/dl. In this case, the primary care physician may not request a full fasting lipid profile because the calculated LDL cholesterol goal will be <100 mg/dl and the additional information would not affect a clinical decision. The final clinical situation that has a practical application for diabetic patients is utilizing a direct LDL cholesterol measurement (2). The direct LDL cholesterol’s Current Procedural Terminology (CPT) code is 83721, and this code was not used in the analysis by Massing et al. (1). The practical application of a direct LDL cholesterol measurement is that the sample can be obtained in a nonfasting state, unlike the calculated LDL cholesterol, which requires at least a 10-h fast. This is one solution to the difficult situation in which a diabetic patient treated with insulin has an office visit at 4:00 p.m. and must maintain a prolonged fast to obtain an accurate calculated LDL cholesterol measurement. Another indication for the direct LDL cholesterol measurement is if the serum triglyceride level exceeds 400 mg/dl. Even if these three situations were incorporated in the Medicare claims data analysis, I suspect the adherence rate would still be low.
The finding of Massing et al. of a low rate of adherence to the American Diabetes Association’s recommendations for lipid testing in diabetic patients is concerning and underscores the need for improvement. Improvement is important because coronary heart disease is a major cause of mortality and morbidity in diabetic populations (3). Furthermore, trials have confirmed the proven benefit of lipid treatment (4,5). A screening lipid test is only the initial step in cardiovascular risk modification and lipid treatment. The finding (1) that 31% of the Medicare population with diabetes did not receive a screening lipid profile in a 2-year period is unfortunate because these individuals were effectively excluded from the opportunity to benefit from the advances in cardiovascular risk modification and lipid treatment. Thus, it is imperative that the adherence rate for lipid screening be improved.
Ironically, the rate of lipid screening in this audit (1) could suggest improvement, since earlier surveys have recorded even lower rates of lipid screening (6–8). Studies have documented an improving trend for awareness and adherence to lipid testing (9,10), which is the result of several factors. One factor is the contribution of various organizations to educate and reinforce the relevance of lipid screening and treatment (10). The continued reporting in the medical literature of adherence rates drives further improvement because the published rates call upon health care providers to exceed these published rates in their own practice. These publications are also the stimulus to develop solutions to exceed these rates. There are several systems that have been developed and utilized to improve the adherence rate for lipid screening (11,12). Incorporation of these systems into practice of large groups of practitioners and the busy practitioner who sees patients every 8–12 min, regardless of the patient’s medical complexity, may result in further improvements in the rates of lipid screening.
Another area that needs improvement is the racial disparity for lipid testing. Massing et al. observed that Caucasians compared with African Americans were 1.6 times more likely to receive a lipid panel. The racial disparity persisted after controlling the variable, access to health care. The racial disparity in cholesterol screening has been reported in other surveys (13). This finding is paradoxical because the group that may be the “most vulnerable” has the one of the lowest rates of lipid testing. A racial disparity has been demonstrated in other health care services, such as influenza vaccination (14), diabetic end-stage real disease (15), coronary artery revascularization procedures (16), and early detection of cancer (17). The continued documentation of this racial disparity in the quality of health care for preventive services suggests that these observations are not aberrant findings. I believe further research and investigation is warranted to determine the factors that contribute to racial disparity for lipid testing.
References
Address correspondence to Richard W. Bergstrom, MD, Portland Diabetes and Endocrinology Center, 1130 NW 22nd, Portland, OR 97210. E-mail: [email protected].
