- Email: [email protected]
Lipid-Lowering Therapy in Patients With Peripheral Arterial Disease: Are Guidelines Being Met?
ORIGINAL ARTICLE LIPID-LOWERING THERAPY IN PATIENTS WITH PAD
Lipid-Lowering Therapy in Patients With Peripheral Arterial Disease: Are Guidelines Being Met? DANIEL G. FEDERMAN, MD; DANA C. RANANI, MD; ROBERT S. KIRSNER, MD; AND DAWN M. BRAVATA, MD OBJECTIVES: To determine the proportion of patients with lower extremity peripheral arterial disease (PAD) who reach recommended low-density lipoprotein cholesterol (LDL-C) levels (<100 mg/dL) and to identify the patient characteristics that are independently associated with attaining the LDL-C goal (<100 mg/dL). PATIENTS AND METHODS: Eligible patients were identified from a roster of patients who had undergone testing at a nonvascular laboratory between September 1, 2001, and January 31, 2002, and were found to have evidence of PAD, defined as an anklebrachial index of 0.9 or less. We thoroughly reviewed patients’ electronic medical records. Backward elimination multivariate logistic regression modeling was used to identify factors associated with reaching the goal LDL-C level. RESULTS: Among 143 patients with PAD, 105 (73%) met the goal LDL-C level. Lipid-lowering therapy was prescribed for 109 (76%). Lower diastolic blood pressure and lower weight were independently associated with an LDL-C level of less than 100 mg/dL. CONCLUSION: We found higher rates of lipid-lowering therapy in patients with PAD than reported previously. Patients with diabetes mellitus or coronary artery disease were not more likely to meet the goal LDL-C level than those without these comorbidities. Clinical practice may be catching up to clinical guidelines.
Mayo Clin Proc. 2005;80(4):494-498 ABI = ankle-brachial index; CI = confidence interval; LDL-C = low-density lipoprotein cholesterol; PAD = peripheral arterial disease
P
eripheral arterial disease (PAD) is a common manifestation of the atherosclerotic process; the age-adjusted prevalence of PAD is 12%, and among those older than 75 years, as many as 20% to 30% may be affected.1,2 Both symptomatic and asymptomatic patients with PAD have increased mortality compared with those without PAD.3 This excess mortality attributable to PAD appears to be related in large part to an increased risk of death due to cardiovascular disease. Cardiovascular disease is the leading cause of death in patients with PAD, accounting for From the Department of Internal Medicine, West Haven Veterans Affairs Medical Center, West Haven, Conn (D.G.F., D.C.R., D.M.B.); Clinical Epidemiology Research Center, West Haven, Conn (D.M.B.); and Department of Dermatology, Miami Veterans Affairs Medical Center, Miami, Fla (R.S.K.). Dr Bravata is supported by a Career Development Award from the Department of Veterans Affairs Health Services Research and Development Service. No authors have any financial interest in any product mentioned within the text. Individual reprints of this article are not available. Address correspondence to Daniel G. Federman, MD, VA Connecticut Health Care System (111ACSL), 950 Campbell Ave, West Haven, CT 06516 (e-mail: [email protected] .gov). © 2005 Mayo Foundation for Medical Education and Research
494
Mayo Clin Proc.
•
nearly 75% of deaths.4 Compared with patients without PAD, the 10-year mortality rate for cardiovascular events is increased 3- to 6-fold in patients with PAD.5 Even in the absence of prior myocardial infarction, the risk of cardiovascular death is similar in patients with PAD compared with patients with known coronary artery or cerebrovascular disease.6 Furthermore, the rate of cardiovascular events increases with increasing PAD severity as measured by the ankle-brachial index (ABI), which is the systolic blood pressure obtained at the ankle divided by the higher brachial systolic blood pressure.7 Patients with critical leg ischemia (those with the lowest ABI values) have an annual mortality rate that approaches 25%, a higher rate than for many malignancies.8 Although current guidelines recommend that the lowdensity lipoprotein cholesterol (LDL-C) level in patients with PAD be held to the same goal as that in patients with known diabetes mellitus or coronary artery disease,9 evidence suggests that atherosclerotic risk factors are less intensively treated among patients with PAD than among those with known coronary artery disease.10 It is currently unknown whether clinicians have become more successful in the treatment of high cholesterol levels in patients with PAD within the past few years and which patient characteristics are associated with the successful attainment of goal LDL-C levels. The objectives of the current study were to determine the proportion of patients with lower extremity PAD who reach recommended LDL-C levels (<100 mg/dL) and identify the patient characteristics that are independently associated with attaining the LDL-C goal (<100 mg/dL). We hypothesized that patients with PAD and either diabetes mellitus or known coronary artery disease would more likely achieve goal LDL-C levels than patients with PAD alone. PATIENTS AND METHODS SETTING AND PATIENT IDENTIFICATION The setting for the current study was the West Haven Veterans Affairs Hospital (affiliated with Yale Medical School and part of the Veterans Affairs Connecticut Health Care System), which serves nearly 42,000 veterans. Eligible patients were identified from the roster of patients who had undergone testing at the noninvasive vascular laboratory between September 1, 2001, and January 31,
April 2005;80(4):494-498
•
www.mayoclinicproceedings.com
For personal use. Mass reproduce only with permission from Mayo Clinic Proceedings.
LIPID-LOWERING THERAPY IN PATIENTS WITH PAD
2002, and were found to have evidence of PAD, which we defined as an ABI of 0.9 or less or having undergone previous lower extremity bypass procedures. We thoroughly reviewed patients’ electronic medical records and collected data pertaining to patient demographics, weight, blood pressure, comorbidity (eg, diabetes mellitus, coronary artery disease, hypertension, cerebrovascular disease, claudication), medications, smoking history, and laboratory results. To allow practitioners time to initiate or titrate lipid-lowering therapy based on the results of noninvasive vascular testing, we identified patients who had undergone noninvasive testing 12 to 18 months before our medical record review. When more than 1 blood pressure, hemoglobin A1c, or LDL-C value was obtained, we used the most recent value for analysis. Patients who died within 1 year of noninvasive testing were excluded from analysis. Two of the authors (D.G.F., D.C.R.) abstracted the medical records data using standard definitions and an extraction form developed for this study. None of the authors were involved in the clinical care of any of the study patients. Any coding uncertainties were documented, resolved by consensus by 3 of the authors (D.G.F., D.C.R., D.M.B.), and recorded in a coding dictionary. Two authors (D.G.F., D.C.R.) reviewed 10% of the medical records to assess interrater reliability. A comparison of these medical records showed complete coding agreement for all abstracted variables and confirmed that both authors used the same methods for recording questions about the medical records data. DEFINITIONS The primary outcome for this study was LDL-C level. For multivariate modeling, we defined achieving goal as an LDL-C level of less than 100 mg/dL and not achieving goal as an LDL-C level of 100 mg/dL or higher. We recorded both left and right lower extremity ABIs. However, for some analyses, we used the minimum ABI, which we define as the lower value of both extremities. STATISTICAL ANALYSES Univariate descriptive statistics were used to summarize baseline characteristics (eg, proportions for binary and ordinal variables and mean ± SDs for continuous variables). The baseline characteristics associated with achieving goal LDL-C level were identified using t test for continuous variables and Fisher exact and χ2 tests for binary variables. P<.05 was used to identify statistical significance, and exact binomial confidence intervals (CIs) were used where noted. We performed backward elimination multivariate logistic regression modeling using the dependent variable of LDL-C level less than 100 mg/dL and included indepenMayo Clin Proc.
•
TABLE 1. Baseline Characteristics of 143 Patients With Peripheral Arterial Disease* Mean ± SD age (y) (range) Race White African American Hispanic Claudication Previous peripheral vascular procedures Diabetes mellitus Known coronary artery disease History of hypertension Stroke or carotid stenosis >50% Tobacco use (current or former smoker) Prescribed cholesterol-lowering medications Mean ± SD blood pressure (mm Hg) (range) Systolic Diastolic Mean ± SD hemoglobin A1c in diabetic patients (%) (range) Mean ± SD total cholesterol (mg/dL) (range) Mean ± SD triglycerides (mg/dL) (range) Mean ± SD HDL-C (mg/dL) (range) Mean ± SD LDL-C (mg/dL) (range)
69.6
±10.6 (47-87) 121 (85) 16 (11) 4 (3) 93 (65) 72 (50) 62 (43) 102 (71) 134 (94) 58 (41) 134 (94) 109 (76)
134±18 (96-192) 71±12 (25-95) 7.2 ±1.4 (4.8-11.4) 164±33 (84-270) 153 ±76 (39-449) 42 ±13 (13-95) 90 ±27 (35-191)
*Data are number (percentage) of patients unless indicated otherwise. HDL-C = high-density lipoprotein cholesterol; LDL-C = low-density lipoprotein cholesterol.
dent variables on the basis of a priori clinical judgment (eg, history of diabetes mellitus or myocardial infarction) and independent variables identified as being associated with having an LDL-C level at goal from the bivariate analysis (P<.05). We maintained an events per variable ratio of more than 10:1 for multivariate modeling.11,12 The outcome rate (LDL-C level <100 mg/dL) exceeded 10%; therefore, odds ratios should not be interpreted as relative risks.13 The sample size for this study was designed to provide 80% power to detect a 20% absolute difference in the rates of patients having goal LDL-C levels for patients with PAD alone vs patients with PAD and diabetes mellitus or myocardial infarction (with a 2-tailed P=.05). All calculations were performed using the software program PC-SAS 8.0 (SAS Institute Inc, Cary, NC). The study protocol was approved by the institutional review board at the West Haven Veterans Affairs Medical Center in West Haven, Conn. RESULTS A total of 151 patients met our inclusion criteria. Among these 151 patients, 147 had total cholesterol levels measured, and 143 had undergone testing for full lipoprotein analysis within the study period. These 143 patients formed the cohort for the current study, and the characteristics of these patients are provided in Table 1. Our cohort was mostly older (mean age, 69.6 years) and white (121 [85%] of 143). Claudication was a common complaint, occurring
April 2005;80(4):494-498
•
www.mayoclinicproceedings.com
For personal use. Mass reproduce only with permission from Mayo Clinic Proceedings.
495
LIPID-LOWERING THERAPY IN PATIENTS WITH PAD
TABLE 2. Characteristics Associated With Meeting Goal LDL-C Level (<100 mg/dL)* LDL-C (mg/dL) <100 (n=105)
Characteristic Mean ± SD age (y) Race† White African American Hispanic Diabetes mellitus Known coronary artery disease Either diabetes or coronary artery disease Previous peripheral vascular procedures Claudication Claudication relief medications Minimum ABI ≤0.4 >0.4 to ≥0.7 >0.7 to ≥0.9 >0.9 Cholesterol-lowering medications Hypertension Stroke or carotid stenosis >50% Current smoker Mean ± SD smoking (pack-years) Mean ± SD blood pressure (mm Hg) Systolic Diastolic Mean ± SD weight (lb) Mean ± SD minimum ABI Mean ± SD total cholesterol (mg/dL) Mean ± SD triglycerides (mg/dL) Mean ± SD HDL-C (mg/dL) Mean ± SD hemoglobin A1c (%)
71.3 ±9.9
≥100 (n=38) 66.1±11.9
86 (82) 13 (12) 4 (4) 47 (45) 79 (75) 88 (84) 51 (49) 65 (62) 22 (21)
35 (92) 3 (8) 0 (0) 15 (39) 23 (61) 29 (76) 21 (55) 28 (74) 11 (29)
9 (9) 39 (37) 34 (32) 23 (22) 84 (80) 100 (95) 43 (41) 30 (29) 54.0 ±39.7
2 (5) 15 (39) 11 (29) 10 (26) 25 (66) 34 (89) 15 (39) 19 (50) 63.3±34.9
131.9±17.6 69.8±10.8 175.0 ±28.9 0.76 ±0.23 149.9±22.4 151.6 ±79.5 41.8 ±13.3 7.4 ±1.5
138.1±19.7 76.2±14.3 191.7±37.5 0.74±0.25 201.3±27.0 156.4±67.2 44.1±10.3 6.8±1.1
P value .02 .32
.57 .09 .30 .48 .38 .32 .85
.08 .21 .87 .02 .21 .08 .02 .02 .71 <.001 .74 .34 .21
*Data are number (percentage) of patients unless indicated otherwise. ABI = ankle-brachial index; HDL-C = high-density lipoprotein cholesterol; LDL-C = low-density lipoprotein cholesterol. †Race could not be determined for 2 patients with an LDL-C level less than 100 mg/dL.
in 93 patients (65%). Of the 93 patients with claudication, 35 (38%) were currently prescribed cilostazol or pentoxifylline. Prior peripheral vascular procedures were relatively common in our cohort: 72 patients (50%) had previously undergone vascular surgery. A history of known coronary artery disease was found in 102 patients (71%), of whom 79 (77%) were currently prescribed β-blockers. Sixty-two patients (43%) had a history of diabetes mellitus. LIPID LEVELS The mean total serum cholesterol level was 164 mg/dL, the mean serum triglyceride level was 153 mg/dL, the mean high-density lipoprotein cholesterol level was 42 mg/dL, and the LDL-C level was 90 mg/dL (Table 1). The use of cholesterol-lowering medications was common: 109 patients (76%) were taking statins, fibrates, niacin, or bile acid sequestrants. Among patients with either diabetes or coronary artery disease, 92 (79%) of 117 received a lipid496
Mayo Clin Proc.
•
lowering medication compared with 17 (65%) of 26 patients with neither diabetes nor coronary artery disease (P=.15). A total of 105 patients had LDL-C levels of less than 100 mg/dL. Therefore, in 38 patients (27%; 95% CI, 19%35%), the LDL-C level was above the recommended goal. BIVARIATE ANALYSIS: FACTORS ASSOCIATED WITH MEETING LDL-C GOAL We found that older age, being a current smoker, lower diastolic blood pressure, lower body weight, and lower total cholesterol level were associated with having an LDLC level of less than 100 mg/dL (Table 2). Patients with known coronary artery disease were more likely to achieve goal LDL-C levels (79 [77%] of 102) than patients without known coronary artery disease (26 [63%] of 41); however, this difference was not statistically significant (P=.09). Having a history of diabetes mellitus was not associated with meeting the LDL-C goal (P=.57). There was no differ-
THE
April 2005;80(4):494-498
•
www.mayoclinicproceedings.com
For personal use. Mass reproduce only with permission from Mayo Clinic Proceedings.
LIPID-LOWERING THERAPY IN PATIENTS WITH PAD
ence in mean ABI levels between patients who met LDL-C goals and those who did not (data not shown). MULTIVARIATE ANALYSIS: FACTORS INDEPENDENTLY ASSOCIATED WITH MEETING THE LDL-C GOAL In a logistic regression model that predicted having an LDL-C level less than 100 mg/dL and that included age, weight, diabetes mellitus, minimum ABI, known coronary artery disease, current tobacco use, and diastolic blood pressure, only lower diastolic blood pressure (adjusted odds ratio for diastolic blood pressure ≤90 mm Hg, 3.99; 95% CI, 11.41-13.93) and lower weight (adjusted odds ratio, 9.38 for each weight ≤220 lb; 95% CI, 1.7051.63) were independently associated with an LDL-C level of <100 mg/dL. DISCUSSION It is well established that dyslipidemia is a strong independent risk factor for atherosclerotic complications and that its treatment is beneficial in patients with PAD.14 In patients with claudication, lipid-lowering therapy reduces disease progression and claudication.15-17 In the large Heart Protection Study,14 high-risk patients randomized to lipid-lowering therapy had an approximately 25% reduction in cardiovascular events compared with those randomized to placebo; similar results were found in the subset of patients enrolled with PAD and no known cardiovascular disease. Our findings indicate that most veterans (113 [75%] of 151) with PAD have achieved an LDL-C level of less than 100 mg/dL. This is surprising given the fact that previous studies have shown low rates of treatment of lipid disorders and other vascular risk factors.10,18-20 For example, although we found that 76% of our patients were receiving lipidlowering therapy, only 46% of patients with PAD and hypercholesterolemia received cholesterol-lowering drug therapy in a study by McDermott et al10 conducted at an academic medical center in the United States. This finding compares with rates of lipid-lowering therapy in patients with PAD of only 16% and 9% in studies from Canada18 and Scotland,19 respectively. Our study differs from that of McDermott et al in that not only did we determine the use of cholesterol medication but also we ascertained how frequently patients met LDL-C goals. Because our study was conducted in a Veterans Affairs hospital, our findings of a high rate of lipid-lowering medication use and goal attainment may be attributable in part to quality improvement initiatives implemented throughout the Veterans Affairs system in the late 1990s.21 For example, health care practitioners receive prompts when using the electronic medical record to check cholesterol Mayo Clin Proc.
•
levels semiannually. Other factors that might be partly responsible for our findings include the fact that the Veterans Affairs electronic medical record ensures that all departments (primary care and specialty services) have access to a patient’s medical record and that the staff has academic affiliations. However, because the study population was limited to one particular Veterans Affairs facility, our findings may not be applicable to other Veterans Affairs facilities. Also, whether other health care systems within the United States currently have similar success rates has yet to be determined. Besides finding a high rate of successful goal attainment with respect to treatment of high LDL-C levels, we found that lower weight and lower diastolic blood pressures were independent predictors of achieving the goal. Health care practitioners who treat dyslipidemia aggressively might also be more aggressive in the treatment of these other vascular risk factors, such as higher weight and high blood pressure. Similarly, patients not adherent or unwilling to take cholesterol-lowering medications might be unlikely to adhere to a regimen of diet, exercise, or antihypertensive therapy. Alternatively, elevated cholesterol levels, higher weight, and higher blood pressure may coexist as part of the metabolic syndrome and be more refractory to intervention. In either case, these patients are at increased risk and represent an opportunity for intervention. Another possible explanation is that patients with chronic illness and wasting may have lower total cholesterol and LDL-C levels not necessarily due to aggressive treatment by their health care practitioners. We did not find that patients with coronary artery disease or diabetes mellitus were more likely to achieve the goal with respect to LDL-C levels than those without these comorbidities. Possibly, widespread dissemination of the National Cholesterol Education Program/Adult Treatment Panel III guidelines has successfully educated clinicians that goals of therapy for patients with PAD are the same as those for patients with established coronary artery disease or diabetes mellitus.9 Alternatively, because a much higher percentage of patients with PAD achieved the goal with respect to their LDL-C level than we anticipated, our study might be underpowered to detect differences among these subgroups. Additional, larger studies might answer these questions. We acknowledge several other possible limitations of our study. We did not systematically collect data regarding the reason for referral to the noninvasive vascular laboratory. Patients undergoing vascular testing might be more adherent to therapy and therefore more likely to achieve goal LDL-C levels than those not referred for testing. Likewise, because almost half of the patients had undergone previous vascular procedures, it is possible that this cohort
April 2005;80(4):494-498
•
www.mayoclinicproceedings.com
For personal use. Mass reproduce only with permission from Mayo Clinic Proceedings.
497
LIPID-LOWERING THERAPY IN PATIENTS WITH PAD
was a more actively managed cohort than patients with PAD who were not referred for testing. CONCLUSION In this study, most veterans with PAD have achieved their goal LDL-C levels. Moreover, patients with PAD and either diabetes mellitus or myocardial infarction are not more likely to have an LDL-C level at goal than patients with PAD without these comorbidities. We hope that these higher rates of attaining cholesterol goals than described previously will translate into improved outcomes.
REFERENCES 1. Criqui MH, Fronek A, Barrett-Connor E, Klauber MR, Gabriel S, Goodman D. The prevalence of peripheral arterial disease in a defined population. Circulation. 1985;71:510-515. 2. Fowkes FG. Epidemiology of peripheral vascular disease. Atherosclerosis. 1997;131(suppl):S29-S31. 3. Rockson SG, Cooke JP. Peripheral arterial insufficiency: mechanisms, natural history, and therapeutic options. Adv Intern Med. 1998;43:253-277. 4. Kannel WB. The demographics of claudication and the aging of the American population. Vasc Med. 1996;1:60-64. 5. Criqui MH, Langer RD, Fronek A, et al. Mortality over a period of 10 years in patients with peripheral arterial disease. N Engl J Med. 1992;326:381386. 6. Newman AB, Shemanski L, Manolio TA, et al, Cardiovascular Health Study Group. Ankle-arm index as a predictor of cardiovascular disease and mortality in the Cardiovascular Health Study. Arterioscler Thromb Vasc Biol. 1999;19:538-545. 7. McKenna M, Wolfson S, Kuller L. The ratio of ankle and arm arterial pressure as an independent predictor of mortality. Atherosclerosis. 1991;87: 119-128. 8. Dormandy J, Heeck L, Vig S. The fate of patients with critical leg ischemia. Semin Vasc Surg. 1999;12:142-147.
498
Mayo Clin Proc.
•
9. Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. Executive summary of the Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). JAMA. 2001;285:2486-2497. 10. McDermott MM, Mehta S, Ahn H, Greenland P. Atherosclerotic risk factors are less intensively treated in patients with peripheral arterial disease than in patients with coronary artery disease. J Gen Intern Med. 1997;12:209215. 11. Concato J, Peduzzi P, Holford TR, Feinstein AR. Importance of events per independent variable in proportional hazards analysis, I: background, goals, and general strategy. J Clin Epidemiol. 1995;48:1495-1501. 12. Peduzzi P, Concato J, Feinstein AR, Holford TR. Importance of events per independent variable in proportional hazards regression analysis, II: accuracy and precision of regression estimates. J Clin Epidemiol. 1995;48:15031510. 13. Zhang J, Yu KF. What’s the relative risk? a method of correcting the odds ratio in cohort studies of common outcomes. JAMA. 1998;280:16901691. 14. Heart Protection Study Collaborative Group. MRC/BHF Heart Protection Study of cholesterol lowering with simvastatin in 20,536 high-risk individuals: a randomised placebo-controlled trial. Lancet. 2002;360:7-22. 15. Leng GC, Price JF, Jepson RG. Lipid-lowering for lower limb atherosclerosis. Cochrane Database Syst Rev. 2000;2:CD000123. 16. Blankenhorn DH, Azen SP, Crawford DW, et al. Effects of colestipolniacin therapy on human femoral atherosclerosis. Circulation. 1991;83:438447. 17. Lewis B. Randomised controlled trial of the treatment of hyperlipidaemia on progression of atherosclerosis. Acta Med Scand Suppl. 1985; 701:53-57. 18. Anand SS, Kundi A, Eikelboom J, Yusuf S. Low rates of preventive practices in patients with peripheral vascular disease. Can J Cardiol. 1999; 15:1259-1263. 19. Clark AL, Byrne JC, Nasser A, McGroarty E, Kennedy JA. Cholesterol in peripheral vascular disease—a suitable case for treatment? QJM. 1999;92: 219-222. 20. Hirsch AT, Criqui MH, Treat-Jacobson D, et al. Peripheral arterial disease detection, awareness, and treatment in primary care. JAMA. 2001;286: 1317-1324. 21. Jha AK, Perlin JB, Kizer KW, Dudley RA. Effect of the transformation of the Veterans Affairs Health Care System on the quality of care. N Engl J Med. 2003;348:2218-2227.
April 2005;80(4):494-498
•
www.mayoclinicproceedings.com
For personal use. Mass reproduce only with permission from Mayo Clinic Proceedings.
Lipid-Lowering Therapy in Patients With Peripheral Arterial Disease: Are Guidelines Being Met? DANIEL G. FEDERMAN, MD; DANA C. RANANI, MD; ROBERT S. KIRSNER, MD; AND DAWN M. BRAVATA, MD OBJECTIVES: To determine the proportion of patients with lower extremity peripheral arterial disease (PAD) who reach recommended low-density lipoprotein cholesterol (LDL-C) levels (<100 mg/dL) and to identify the patient characteristics that are independently associated with attaining the LDL-C goal (<100 mg/dL). PATIENTS AND METHODS: Eligible patients were identified from a roster of patients who had undergone testing at a nonvascular laboratory between September 1, 2001, and January 31, 2002, and were found to have evidence of PAD, defined as an anklebrachial index of 0.9 or less. We thoroughly reviewed patients’ electronic medical records. Backward elimination multivariate logistic regression modeling was used to identify factors associated with reaching the goal LDL-C level. RESULTS: Among 143 patients with PAD, 105 (73%) met the goal LDL-C level. Lipid-lowering therapy was prescribed for 109 (76%). Lower diastolic blood pressure and lower weight were independently associated with an LDL-C level of less than 100 mg/dL. CONCLUSION: We found higher rates of lipid-lowering therapy in patients with PAD than reported previously. Patients with diabetes mellitus or coronary artery disease were not more likely to meet the goal LDL-C level than those without these comorbidities. Clinical practice may be catching up to clinical guidelines.
Mayo Clin Proc. 2005;80(4):494-498 ABI = ankle-brachial index; CI = confidence interval; LDL-C = low-density lipoprotein cholesterol; PAD = peripheral arterial disease
P
eripheral arterial disease (PAD) is a common manifestation of the atherosclerotic process; the age-adjusted prevalence of PAD is 12%, and among those older than 75 years, as many as 20% to 30% may be affected.1,2 Both symptomatic and asymptomatic patients with PAD have increased mortality compared with those without PAD.3 This excess mortality attributable to PAD appears to be related in large part to an increased risk of death due to cardiovascular disease. Cardiovascular disease is the leading cause of death in patients with PAD, accounting for From the Department of Internal Medicine, West Haven Veterans Affairs Medical Center, West Haven, Conn (D.G.F., D.C.R., D.M.B.); Clinical Epidemiology Research Center, West Haven, Conn (D.M.B.); and Department of Dermatology, Miami Veterans Affairs Medical Center, Miami, Fla (R.S.K.). Dr Bravata is supported by a Career Development Award from the Department of Veterans Affairs Health Services Research and Development Service. No authors have any financial interest in any product mentioned within the text. Individual reprints of this article are not available. Address correspondence to Daniel G. Federman, MD, VA Connecticut Health Care System (111ACSL), 950 Campbell Ave, West Haven, CT 06516 (e-mail: [email protected] .gov). © 2005 Mayo Foundation for Medical Education and Research
494
Mayo Clin Proc.
•
nearly 75% of deaths.4 Compared with patients without PAD, the 10-year mortality rate for cardiovascular events is increased 3- to 6-fold in patients with PAD.5 Even in the absence of prior myocardial infarction, the risk of cardiovascular death is similar in patients with PAD compared with patients with known coronary artery or cerebrovascular disease.6 Furthermore, the rate of cardiovascular events increases with increasing PAD severity as measured by the ankle-brachial index (ABI), which is the systolic blood pressure obtained at the ankle divided by the higher brachial systolic blood pressure.7 Patients with critical leg ischemia (those with the lowest ABI values) have an annual mortality rate that approaches 25%, a higher rate than for many malignancies.8 Although current guidelines recommend that the lowdensity lipoprotein cholesterol (LDL-C) level in patients with PAD be held to the same goal as that in patients with known diabetes mellitus or coronary artery disease,9 evidence suggests that atherosclerotic risk factors are less intensively treated among patients with PAD than among those with known coronary artery disease.10 It is currently unknown whether clinicians have become more successful in the treatment of high cholesterol levels in patients with PAD within the past few years and which patient characteristics are associated with the successful attainment of goal LDL-C levels. The objectives of the current study were to determine the proportion of patients with lower extremity PAD who reach recommended LDL-C levels (<100 mg/dL) and identify the patient characteristics that are independently associated with attaining the LDL-C goal (<100 mg/dL). We hypothesized that patients with PAD and either diabetes mellitus or known coronary artery disease would more likely achieve goal LDL-C levels than patients with PAD alone. PATIENTS AND METHODS SETTING AND PATIENT IDENTIFICATION The setting for the current study was the West Haven Veterans Affairs Hospital (affiliated with Yale Medical School and part of the Veterans Affairs Connecticut Health Care System), which serves nearly 42,000 veterans. Eligible patients were identified from the roster of patients who had undergone testing at the noninvasive vascular laboratory between September 1, 2001, and January 31,
April 2005;80(4):494-498
•
www.mayoclinicproceedings.com
For personal use. Mass reproduce only with permission from Mayo Clinic Proceedings.
LIPID-LOWERING THERAPY IN PATIENTS WITH PAD
2002, and were found to have evidence of PAD, which we defined as an ABI of 0.9 or less or having undergone previous lower extremity bypass procedures. We thoroughly reviewed patients’ electronic medical records and collected data pertaining to patient demographics, weight, blood pressure, comorbidity (eg, diabetes mellitus, coronary artery disease, hypertension, cerebrovascular disease, claudication), medications, smoking history, and laboratory results. To allow practitioners time to initiate or titrate lipid-lowering therapy based on the results of noninvasive vascular testing, we identified patients who had undergone noninvasive testing 12 to 18 months before our medical record review. When more than 1 blood pressure, hemoglobin A1c, or LDL-C value was obtained, we used the most recent value for analysis. Patients who died within 1 year of noninvasive testing were excluded from analysis. Two of the authors (D.G.F., D.C.R.) abstracted the medical records data using standard definitions and an extraction form developed for this study. None of the authors were involved in the clinical care of any of the study patients. Any coding uncertainties were documented, resolved by consensus by 3 of the authors (D.G.F., D.C.R., D.M.B.), and recorded in a coding dictionary. Two authors (D.G.F., D.C.R.) reviewed 10% of the medical records to assess interrater reliability. A comparison of these medical records showed complete coding agreement for all abstracted variables and confirmed that both authors used the same methods for recording questions about the medical records data. DEFINITIONS The primary outcome for this study was LDL-C level. For multivariate modeling, we defined achieving goal as an LDL-C level of less than 100 mg/dL and not achieving goal as an LDL-C level of 100 mg/dL or higher. We recorded both left and right lower extremity ABIs. However, for some analyses, we used the minimum ABI, which we define as the lower value of both extremities. STATISTICAL ANALYSES Univariate descriptive statistics were used to summarize baseline characteristics (eg, proportions for binary and ordinal variables and mean ± SDs for continuous variables). The baseline characteristics associated with achieving goal LDL-C level were identified using t test for continuous variables and Fisher exact and χ2 tests for binary variables. P<.05 was used to identify statistical significance, and exact binomial confidence intervals (CIs) were used where noted. We performed backward elimination multivariate logistic regression modeling using the dependent variable of LDL-C level less than 100 mg/dL and included indepenMayo Clin Proc.
•
TABLE 1. Baseline Characteristics of 143 Patients With Peripheral Arterial Disease* Mean ± SD age (y) (range) Race White African American Hispanic Claudication Previous peripheral vascular procedures Diabetes mellitus Known coronary artery disease History of hypertension Stroke or carotid stenosis >50% Tobacco use (current or former smoker) Prescribed cholesterol-lowering medications Mean ± SD blood pressure (mm Hg) (range) Systolic Diastolic Mean ± SD hemoglobin A1c in diabetic patients (%) (range) Mean ± SD total cholesterol (mg/dL) (range) Mean ± SD triglycerides (mg/dL) (range) Mean ± SD HDL-C (mg/dL) (range) Mean ± SD LDL-C (mg/dL) (range)
69.6
±10.6 (47-87) 121 (85) 16 (11) 4 (3) 93 (65) 72 (50) 62 (43) 102 (71) 134 (94) 58 (41) 134 (94) 109 (76)
134±18 (96-192) 71±12 (25-95) 7.2 ±1.4 (4.8-11.4) 164±33 (84-270) 153 ±76 (39-449) 42 ±13 (13-95) 90 ±27 (35-191)
*Data are number (percentage) of patients unless indicated otherwise. HDL-C = high-density lipoprotein cholesterol; LDL-C = low-density lipoprotein cholesterol.
dent variables on the basis of a priori clinical judgment (eg, history of diabetes mellitus or myocardial infarction) and independent variables identified as being associated with having an LDL-C level at goal from the bivariate analysis (P<.05). We maintained an events per variable ratio of more than 10:1 for multivariate modeling.11,12 The outcome rate (LDL-C level <100 mg/dL) exceeded 10%; therefore, odds ratios should not be interpreted as relative risks.13 The sample size for this study was designed to provide 80% power to detect a 20% absolute difference in the rates of patients having goal LDL-C levels for patients with PAD alone vs patients with PAD and diabetes mellitus or myocardial infarction (with a 2-tailed P=.05). All calculations were performed using the software program PC-SAS 8.0 (SAS Institute Inc, Cary, NC). The study protocol was approved by the institutional review board at the West Haven Veterans Affairs Medical Center in West Haven, Conn. RESULTS A total of 151 patients met our inclusion criteria. Among these 151 patients, 147 had total cholesterol levels measured, and 143 had undergone testing for full lipoprotein analysis within the study period. These 143 patients formed the cohort for the current study, and the characteristics of these patients are provided in Table 1. Our cohort was mostly older (mean age, 69.6 years) and white (121 [85%] of 143). Claudication was a common complaint, occurring
April 2005;80(4):494-498
•
www.mayoclinicproceedings.com
For personal use. Mass reproduce only with permission from Mayo Clinic Proceedings.
495
LIPID-LOWERING THERAPY IN PATIENTS WITH PAD
TABLE 2. Characteristics Associated With Meeting Goal LDL-C Level (<100 mg/dL)* LDL-C (mg/dL) <100 (n=105)
Characteristic Mean ± SD age (y) Race† White African American Hispanic Diabetes mellitus Known coronary artery disease Either diabetes or coronary artery disease Previous peripheral vascular procedures Claudication Claudication relief medications Minimum ABI ≤0.4 >0.4 to ≥0.7 >0.7 to ≥0.9 >0.9 Cholesterol-lowering medications Hypertension Stroke or carotid stenosis >50% Current smoker Mean ± SD smoking (pack-years) Mean ± SD blood pressure (mm Hg) Systolic Diastolic Mean ± SD weight (lb) Mean ± SD minimum ABI Mean ± SD total cholesterol (mg/dL) Mean ± SD triglycerides (mg/dL) Mean ± SD HDL-C (mg/dL) Mean ± SD hemoglobin A1c (%)
71.3 ±9.9
≥100 (n=38) 66.1±11.9
86 (82) 13 (12) 4 (4) 47 (45) 79 (75) 88 (84) 51 (49) 65 (62) 22 (21)
35 (92) 3 (8) 0 (0) 15 (39) 23 (61) 29 (76) 21 (55) 28 (74) 11 (29)
9 (9) 39 (37) 34 (32) 23 (22) 84 (80) 100 (95) 43 (41) 30 (29) 54.0 ±39.7
2 (5) 15 (39) 11 (29) 10 (26) 25 (66) 34 (89) 15 (39) 19 (50) 63.3±34.9
131.9±17.6 69.8±10.8 175.0 ±28.9 0.76 ±0.23 149.9±22.4 151.6 ±79.5 41.8 ±13.3 7.4 ±1.5
138.1±19.7 76.2±14.3 191.7±37.5 0.74±0.25 201.3±27.0 156.4±67.2 44.1±10.3 6.8±1.1
P value .02 .32
.57 .09 .30 .48 .38 .32 .85
.08 .21 .87 .02 .21 .08 .02 .02 .71 <.001 .74 .34 .21
*Data are number (percentage) of patients unless indicated otherwise. ABI = ankle-brachial index; HDL-C = high-density lipoprotein cholesterol; LDL-C = low-density lipoprotein cholesterol. †Race could not be determined for 2 patients with an LDL-C level less than 100 mg/dL.
in 93 patients (65%). Of the 93 patients with claudication, 35 (38%) were currently prescribed cilostazol or pentoxifylline. Prior peripheral vascular procedures were relatively common in our cohort: 72 patients (50%) had previously undergone vascular surgery. A history of known coronary artery disease was found in 102 patients (71%), of whom 79 (77%) were currently prescribed β-blockers. Sixty-two patients (43%) had a history of diabetes mellitus. LIPID LEVELS The mean total serum cholesterol level was 164 mg/dL, the mean serum triglyceride level was 153 mg/dL, the mean high-density lipoprotein cholesterol level was 42 mg/dL, and the LDL-C level was 90 mg/dL (Table 1). The use of cholesterol-lowering medications was common: 109 patients (76%) were taking statins, fibrates, niacin, or bile acid sequestrants. Among patients with either diabetes or coronary artery disease, 92 (79%) of 117 received a lipid496
Mayo Clin Proc.
•
lowering medication compared with 17 (65%) of 26 patients with neither diabetes nor coronary artery disease (P=.15). A total of 105 patients had LDL-C levels of less than 100 mg/dL. Therefore, in 38 patients (27%; 95% CI, 19%35%), the LDL-C level was above the recommended goal. BIVARIATE ANALYSIS: FACTORS ASSOCIATED WITH MEETING LDL-C GOAL We found that older age, being a current smoker, lower diastolic blood pressure, lower body weight, and lower total cholesterol level were associated with having an LDLC level of less than 100 mg/dL (Table 2). Patients with known coronary artery disease were more likely to achieve goal LDL-C levels (79 [77%] of 102) than patients without known coronary artery disease (26 [63%] of 41); however, this difference was not statistically significant (P=.09). Having a history of diabetes mellitus was not associated with meeting the LDL-C goal (P=.57). There was no differ-
THE
April 2005;80(4):494-498
•
www.mayoclinicproceedings.com
For personal use. Mass reproduce only with permission from Mayo Clinic Proceedings.
LIPID-LOWERING THERAPY IN PATIENTS WITH PAD
ence in mean ABI levels between patients who met LDL-C goals and those who did not (data not shown). MULTIVARIATE ANALYSIS: FACTORS INDEPENDENTLY ASSOCIATED WITH MEETING THE LDL-C GOAL In a logistic regression model that predicted having an LDL-C level less than 100 mg/dL and that included age, weight, diabetes mellitus, minimum ABI, known coronary artery disease, current tobacco use, and diastolic blood pressure, only lower diastolic blood pressure (adjusted odds ratio for diastolic blood pressure ≤90 mm Hg, 3.99; 95% CI, 11.41-13.93) and lower weight (adjusted odds ratio, 9.38 for each weight ≤220 lb; 95% CI, 1.7051.63) were independently associated with an LDL-C level of <100 mg/dL. DISCUSSION It is well established that dyslipidemia is a strong independent risk factor for atherosclerotic complications and that its treatment is beneficial in patients with PAD.14 In patients with claudication, lipid-lowering therapy reduces disease progression and claudication.15-17 In the large Heart Protection Study,14 high-risk patients randomized to lipid-lowering therapy had an approximately 25% reduction in cardiovascular events compared with those randomized to placebo; similar results were found in the subset of patients enrolled with PAD and no known cardiovascular disease. Our findings indicate that most veterans (113 [75%] of 151) with PAD have achieved an LDL-C level of less than 100 mg/dL. This is surprising given the fact that previous studies have shown low rates of treatment of lipid disorders and other vascular risk factors.10,18-20 For example, although we found that 76% of our patients were receiving lipidlowering therapy, only 46% of patients with PAD and hypercholesterolemia received cholesterol-lowering drug therapy in a study by McDermott et al10 conducted at an academic medical center in the United States. This finding compares with rates of lipid-lowering therapy in patients with PAD of only 16% and 9% in studies from Canada18 and Scotland,19 respectively. Our study differs from that of McDermott et al in that not only did we determine the use of cholesterol medication but also we ascertained how frequently patients met LDL-C goals. Because our study was conducted in a Veterans Affairs hospital, our findings of a high rate of lipid-lowering medication use and goal attainment may be attributable in part to quality improvement initiatives implemented throughout the Veterans Affairs system in the late 1990s.21 For example, health care practitioners receive prompts when using the electronic medical record to check cholesterol Mayo Clin Proc.
•
levels semiannually. Other factors that might be partly responsible for our findings include the fact that the Veterans Affairs electronic medical record ensures that all departments (primary care and specialty services) have access to a patient’s medical record and that the staff has academic affiliations. However, because the study population was limited to one particular Veterans Affairs facility, our findings may not be applicable to other Veterans Affairs facilities. Also, whether other health care systems within the United States currently have similar success rates has yet to be determined. Besides finding a high rate of successful goal attainment with respect to treatment of high LDL-C levels, we found that lower weight and lower diastolic blood pressures were independent predictors of achieving the goal. Health care practitioners who treat dyslipidemia aggressively might also be more aggressive in the treatment of these other vascular risk factors, such as higher weight and high blood pressure. Similarly, patients not adherent or unwilling to take cholesterol-lowering medications might be unlikely to adhere to a regimen of diet, exercise, or antihypertensive therapy. Alternatively, elevated cholesterol levels, higher weight, and higher blood pressure may coexist as part of the metabolic syndrome and be more refractory to intervention. In either case, these patients are at increased risk and represent an opportunity for intervention. Another possible explanation is that patients with chronic illness and wasting may have lower total cholesterol and LDL-C levels not necessarily due to aggressive treatment by their health care practitioners. We did not find that patients with coronary artery disease or diabetes mellitus were more likely to achieve the goal with respect to LDL-C levels than those without these comorbidities. Possibly, widespread dissemination of the National Cholesterol Education Program/Adult Treatment Panel III guidelines has successfully educated clinicians that goals of therapy for patients with PAD are the same as those for patients with established coronary artery disease or diabetes mellitus.9 Alternatively, because a much higher percentage of patients with PAD achieved the goal with respect to their LDL-C level than we anticipated, our study might be underpowered to detect differences among these subgroups. Additional, larger studies might answer these questions. We acknowledge several other possible limitations of our study. We did not systematically collect data regarding the reason for referral to the noninvasive vascular laboratory. Patients undergoing vascular testing might be more adherent to therapy and therefore more likely to achieve goal LDL-C levels than those not referred for testing. Likewise, because almost half of the patients had undergone previous vascular procedures, it is possible that this cohort
April 2005;80(4):494-498
•
www.mayoclinicproceedings.com
For personal use. Mass reproduce only with permission from Mayo Clinic Proceedings.
497
LIPID-LOWERING THERAPY IN PATIENTS WITH PAD
was a more actively managed cohort than patients with PAD who were not referred for testing. CONCLUSION In this study, most veterans with PAD have achieved their goal LDL-C levels. Moreover, patients with PAD and either diabetes mellitus or myocardial infarction are not more likely to have an LDL-C level at goal than patients with PAD without these comorbidities. We hope that these higher rates of attaining cholesterol goals than described previously will translate into improved outcomes.
REFERENCES 1. Criqui MH, Fronek A, Barrett-Connor E, Klauber MR, Gabriel S, Goodman D. The prevalence of peripheral arterial disease in a defined population. Circulation. 1985;71:510-515. 2. Fowkes FG. Epidemiology of peripheral vascular disease. Atherosclerosis. 1997;131(suppl):S29-S31. 3. Rockson SG, Cooke JP. Peripheral arterial insufficiency: mechanisms, natural history, and therapeutic options. Adv Intern Med. 1998;43:253-277. 4. Kannel WB. The demographics of claudication and the aging of the American population. Vasc Med. 1996;1:60-64. 5. Criqui MH, Langer RD, Fronek A, et al. Mortality over a period of 10 years in patients with peripheral arterial disease. N Engl J Med. 1992;326:381386. 6. Newman AB, Shemanski L, Manolio TA, et al, Cardiovascular Health Study Group. Ankle-arm index as a predictor of cardiovascular disease and mortality in the Cardiovascular Health Study. Arterioscler Thromb Vasc Biol. 1999;19:538-545. 7. McKenna M, Wolfson S, Kuller L. The ratio of ankle and arm arterial pressure as an independent predictor of mortality. Atherosclerosis. 1991;87: 119-128. 8. Dormandy J, Heeck L, Vig S. The fate of patients with critical leg ischemia. Semin Vasc Surg. 1999;12:142-147.
498
Mayo Clin Proc.
•
9. Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. Executive summary of the Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). JAMA. 2001;285:2486-2497. 10. McDermott MM, Mehta S, Ahn H, Greenland P. Atherosclerotic risk factors are less intensively treated in patients with peripheral arterial disease than in patients with coronary artery disease. J Gen Intern Med. 1997;12:209215. 11. Concato J, Peduzzi P, Holford TR, Feinstein AR. Importance of events per independent variable in proportional hazards analysis, I: background, goals, and general strategy. J Clin Epidemiol. 1995;48:1495-1501. 12. Peduzzi P, Concato J, Feinstein AR, Holford TR. Importance of events per independent variable in proportional hazards regression analysis, II: accuracy and precision of regression estimates. J Clin Epidemiol. 1995;48:15031510. 13. Zhang J, Yu KF. What’s the relative risk? a method of correcting the odds ratio in cohort studies of common outcomes. JAMA. 1998;280:16901691. 14. Heart Protection Study Collaborative Group. MRC/BHF Heart Protection Study of cholesterol lowering with simvastatin in 20,536 high-risk individuals: a randomised placebo-controlled trial. Lancet. 2002;360:7-22. 15. Leng GC, Price JF, Jepson RG. Lipid-lowering for lower limb atherosclerosis. Cochrane Database Syst Rev. 2000;2:CD000123. 16. Blankenhorn DH, Azen SP, Crawford DW, et al. Effects of colestipolniacin therapy on human femoral atherosclerosis. Circulation. 1991;83:438447. 17. Lewis B. Randomised controlled trial of the treatment of hyperlipidaemia on progression of atherosclerosis. Acta Med Scand Suppl. 1985; 701:53-57. 18. Anand SS, Kundi A, Eikelboom J, Yusuf S. Low rates of preventive practices in patients with peripheral vascular disease. Can J Cardiol. 1999; 15:1259-1263. 19. Clark AL, Byrne JC, Nasser A, McGroarty E, Kennedy JA. Cholesterol in peripheral vascular disease—a suitable case for treatment? QJM. 1999;92: 219-222. 20. Hirsch AT, Criqui MH, Treat-Jacobson D, et al. Peripheral arterial disease detection, awareness, and treatment in primary care. JAMA. 2001;286: 1317-1324. 21. Jha AK, Perlin JB, Kizer KW, Dudley RA. Effect of the transformation of the Veterans Affairs Health Care System on the quality of care. N Engl J Med. 2003;348:2218-2227.
April 2005;80(4):494-498
•
www.mayoclinicproceedings.com
For personal use. Mass reproduce only with permission from Mayo Clinic Proceedings.