The Treatment of Dyslipidemia in Actual Practice Settings That Use Similar Inclusion Criteria to the ACCORD Lipid Trial

Abstract

Synopsis: Although clinical trials are a crucial source of information, it is important to understand the applicability of the results within the context of an actual practice setting in a non-randomized patient population. Purpose: To describe the case-mix characteristics of patients selected for treatment in an actual practice setting and compare with patients randomized to treatment in the ACCORD lipid trial. Methods: A retrospective claims database analysis was conducted by the use of i3 InVision™ Data Mart. Patients with a lipid panel between January 2004 and March 2008 were selected and followed from the first occurrence (index date) of the following: LDL-C 60–180 mg/dL; HDL-C <55 mg/dL for women; HDL-C <50 mg/dL for men; and TG <750 mg/dL. Additional inclusion criteria were age 40–64 years, enrollment >179 days before and after index date, more than two diabetes medical claims or more than 89 days' supply of a prescription for diabetes treatment during the 6 months pre-index, and no lipid-altering therapy during the 6 months pre-index. Patients' baseline lipid values were compared with the reported baseline lipid values from the ACCORD lipid trial. Patients were classified into four groups on the basis of treatment during the 6-month post-index period: monotherapy only, statin and fenofibrate combination, other combination (other than statin and fenofibrate combination), and no therapy. Results: A total of 141,595 patients met study criteria. The majority of the patients did not receive any treatment in the 6-month post-index period (n = 124,464); 16,559 received monotherapy, and 572 received combination therapy (181 statin and fenofibrate combinations, 391 other combinations). The mean (SD) baseline LDL-C for patients receiving no therapy, monotherapy, statin and fenofibrate combination therapy, and other combination therapy were 118.8 mg/dL (26.8), 133.5 mg/dL (29.7), 122.7 mg/dL (31.7), and 126.9 mg/dL (31.7), respectively. The median baseline TG levels for patients receiving no therapy, monotherapy, statin and fenofibrate combination, and other combination therapy were 138.0 mg/dL, 177.0 mg/dL, 253.0 mg/dL, and 207.0 mg/dL, respectively. Comparatively, patients randomized to statin monotherapy or statin and fenofibrate treatment in the ACCORD lipid trial had a baseline LDL-C (mean) and TG (median) level of 100.6 mg/dL and 162.0 mg/dL, respectively.1 Conclusions: Although randomized clinical trials are important to demonstrate safety and efficacy, their results may not be directly applicable to the actual practice setting. In actual practice, combination therapies are prescribed with a selection bias in patients with increased LDL-C and TG values compared with patients randomized to therapy in the ACCORD lipid trial.Reference1.The ACCORD Study Group. Action to Control Cardiovascular Risk in Diabetes (ACCORD) Trial: Design and Methods. Am J Cardiol. 2007;99(12A):21i-33i.