Sensitivity and Specificity of Ankle-Brachial Index for Detecting Angiographic Stenosis of Peripheral Arteries

Transcription

1 Circ J 2008; 72: Sensitivity and Specificity of Ankle-Brachial Index for Detecting Angiographic Stenosis of Peripheral Arteries Xiaoming Guo, MD ; Jue Li, MD ; Wenyue Pang, MD*; Mingzhong Zhao, MD; Yingyi Luo, MD**; Yingxian Sun, MD*; Dayi Hu, MD Background The aim of the present study was to prospectively evaluate the sensitivity, specificity, positive and negative likelihood ratios (LR+, LR ) of the ankle brachial index (ABI), using conventional digital subtraction angiography (DSA) as the reference standard, in the assessment of lower extremity arteries, and to research the threshold value of the ABI in diagnosing periphery arterial disease (PAD), as well as the relationship between the ABI value and stenosis in the artery of the lower extremity in Chinese high-risk cardiovascular patients. Methods and Results A total of 298 consecutive patients (199 men, 99 women, 64.9±11.3 years old) underwent conventional DSA and ABI measurement. Receiver operator characteristics (ROC) analysis was performed to assess possible threshold values that predict PAD in these patients. The greater the stenosis in the artery of the lower extremity, the lower the measured ABI value. DSA was used as the gold standard in defining lesions 30%, 50%, and 70% and the respective areas under the ROC curve were (95% confidence interval (CI) 0.712, 0.860), (95% CI 0.869, 0.984), and (95% CI 0.927, 0.999). Conventional DSA was the gold standard in defining 50% luminal stenosis for the diagnosis of lower extremity PAD. The 0.95 is the overall cutoff of the ABI that was associated with 91% sensitivity, 86% specificity, 6.5 LR+ and 0.1 LR for detection of hemodynamically significant stenosis (lesions 50%) in all 298 subjects (p<0.001). Conclusion The ABI value shows a decreasing tendency with increasing severity of stenosis in patients with PAD. ABI measurement is an accurate and reliable noninvasive alternative to conventional DSA in the assessment of lower extremity arteries and the cut-off of 0.95 is the threshold ABI value for detecting PAD in Chinese patients. (Circ J 2008; 72: ) Key Words: Ankle brachial index; Digital subtraction angiography; Peripheral arterial disease; Sensitivity and specificity Peripheral arterial disease (PAD) of the lower extremities is a chronic atherosclerotic occlusive disease that is highly prevalent, 1 being estimated as between 4.5% and 29% worldwide and affecting more than 20% of individuals aged over 75 years. 2 It can be accompanied by intermittent claudication or other leg pain and may result in vascular surgery or lower limb amputation. 3 In addition, both symptomatic and asymptomatic PAD are associated with a greater risk of functional decline and cardiovascular disease, including stroke morbidity and mortality. 4 8 However, PAD, which is defined as equivalent to coronary artery disease (CAD), is frequently overlooked and not appropriately treated. 9 The diagnosis of PAD is made from a typical history, physical examination and the ankle brachial index (ABI) measurement, 10 which is a simple non-invasive screening test 11 that is an inexpensive measure of subclinical athero- (Received August 8, 2007; revised manuscript received October 27, 2007; accepted November 15, 2007) Heart, Lung and Blood Vessel Center, Tongji University, Shanghai, *Department of Cardiology, The Second Affiliated of China Medical University, Shenyang and **Shanghai Institute of Health Sciences, Shanghai, PR China Xiaoming Guo and Jue Li are co-first authors. Mailing address: Dayi Hu, MD, Heart, Lung and Blood Vessel Center, Tongji University, 1239 Siping Road, Shanghai, , PR China. jueli59jp@yahoo.co.jp All rights are reserved to the Japanese Circulation Society. For permissions, please cj@j-circ.or.jp sclerosis. 12 The prevalence of PAD based on an ABI <0.9 increases from 4.3% in those aged over 40 years old to 14.5% in the population aged more than 70 years old. 13 Prospective studies have found that those with ABI-defined PAD are approximately fold more likely to have a clinical cardiovascular event than those without PAD. 7,14,15 The report from the American Heart Association Prevention Conference V concluded that the ABI provides risk information over and above that provided by traditional risk factors and suggested that the test might be a useful addition to the assessment of coronary heart disease (CHD) risk in selected populations. 12 Recent guidelines for the Detection, Evaluation and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) recommend that an ABI <0.9 should be considered a CHD risk equivalent (10-year CHD risk >20%). 16 Although the ABI generally is , there is a progressive fall in systolic blood pressure (SBP) at the ankle with an increasing degree of severity of arterial disease in the lower extremities. An ABI 0.90 or 0.95 is used to make the diagnosis of PAD in the clinical setting. Because individuals with PAD may not have typical symptoms of exertional leg discomfort, the ABI is a useful diagnostic measure to ascertain the presence of PAD. 17 Digital subtraction angiography (DSA) is considered the best method for assessing PAD severity, location and extent. 18 Traditionally, intra-arterial DSA has been used for this purpose, but it uses ionizing radiation and carries a small, but not insignificant, risk for local and systemic com-

2 606 GUO X et al. Table 1 Baseline Characteristics and Medications of 282 Study Subjects Age, years 64.93±11.32 Men, n (%) 66.8 SBP (mmhg) ±19.84 DBP (mmhg) 71.68±11.71 TC (mmol/l) 4.78±2.50 TG (mmol/l) 1.75±1.27 LDL (mmol/l) 2.88±0.93 HDL (mmol/l) 1.10±0.26 FPG (mmol/l) 6.23±2.11 Ever smoked, n (%) 45.0 CAD (%) 53.0 CI (%) DM (%) 22.8 Dyslipidemia (%) 33.9 Statins (%) 90.3 ACEI (%) 71.1 ARB (%) 12.8 Antiplatelet (%) receptor blockers (%) 66.1 CCB (%) 36.2 Diuretics (%) 11.4 OHA (%) 16.8 Nitrates (%) 77.5 Digitalis (%) 5.7 SBP, systolic blood pressure; DBP, diastolic blood pressure; TC, total cholesterol; TG, triglycerides; LDL, low-density lipoprotein; HDL, high-density lipoprotein; FPG, fasting plasma glucose; CAD, coronary artery disease; CI, cerebral infarction; 2-DM, type 2 diabetes mellitus; ACEI, angiotensinconverting enzyme inhibitor; ARB, angiotensin-receptor blockers; CCB, calcium-channel blockers; OHA, oral hypoglycemic agents. plications because of the invasive nature of the procedure and the use of nephrotoxic contrast media. Thus, it is preferable to use a noninvasive and easy method of detecting PAD. However, there are no reports on the threshold value of ABI in detecting PAD in Chinese patients. The present study aimed to compare the noninvasive method (ABI) with the invasive method (DSA) in evaluating the extent of PAD in order to determine the sensitivity, specificity and threshold value of ABI in diagnosing PAD in Chinese patients. We also compared the relationship between the value of ABI and the degree of stenosis in the artery of the lower extremity. Methods Study Subjects Subjects were recruited from the cardiology in-patient clinic at 2 university hospitals in Shanghai and Shenyang from October 2006 to January PAD on angiography was defined as the presence of 50% or more stenosis of any lower extremity artery from the aorto-iliac bifurcation to the ankle arteries. Inclusion criteria included patients with an ABI performed within 30 days prior to DSA. Participant selection criteria were: Han ethnicity older than 35 years, living in the community, and unrelated to any other participants. Exclusion criteria were: multiple organ dysfunction syndrome, pregnancy or lactation, mental disorder, serious diabetes mellitus or hypertension and their complications (ketoacidosis, hypertensive crisis etc), secondary hypertension, and type 1 diabetes. Patients with noncompressible vessels and ABI >1.40 were excluded. The study cohort included 298 consecutive participants who were referred for conventional DSA and ABI measurement. Fasting blood samples were taken from all participants to investigate serum lipid and glucose levels. Baseline data, history of illness and in-hospital medications of all patients were also recorded. This study was approved by the Ethics Committee of Tongji University and informed consent was given by the participants. Measurement of ABI The ABI value was obtained by oscillometric method, which is an effective equivalent to the value obtained by the Doppler method. 22,23 The Vascular Profiler-1000 (Model BP203RPE II, Form PWV/ABI, OMRON Colin Medical Instruments, Tokyo, Japan), which has been approved by the US Food and Drug Administration, was used to obtain blood pressure (BP) in all 4 extremities, ABI (anklebrachial index of BP), and ankle-brachial pulse wave velocity simultaneously. The device uses waveform analysis and vascular evaluation technology to measure arterial compliance in central (large) arteries as well as in the peripheral arteries. The lowest ABI of both legs was the index leg used. 18 DSA DSA, which is the gold standard in defining 50% luminal stenosis for the diagnosis of lower extremity PAD, was performed in Tongji Hospital of Tongji University. Catheterization was performed via the transfemoral or transbrachial approach. Appropriate anteroposterior sequential views of the lower abdomen, pelvis and lower extremities were obtained. Oblique views were obtained for the iliac and proximal femoral arteries. DSA images were reviewed by 2 experienced angiographers. Diameter stenosis of the arteries was made by visual estimates with comparison of adjacent normal arterial segments. PAD was defined when there was 50% or more stenosis of any of the vessels from the aorto-iliac bifurcation to the distal ends of the anterior and posterior tibial and peroneal arteries. Occlusion less than 50% of the lumen, mild atherosclerosis and luminal irregularities were not considered as PAD. Image Analysis Two reviewers independently reviewed the DSA images in a randomized order and without knowledge of the patient s clinical data. In case of disagreement about the degree of vessel stenosis, a final consensus interpretation was performed. Segments were classified as patent or occluded. Patent segments were further classified as having 50% or 50% stenosis. In cases of multiple sites of disease, only the most severely affected site was scored. Stenosis of the arterial segment was graded with a 6-point scale and electronic calipers. Grade 1 indicated a normal vessel; grade 2 indicated mild vessel irregularities (<30% luminal narrowing); grade 3 indicated moderate arterial stenosis (30 49% luminal narrowing); grade 4 indicated severe arterial stenosis (50 69% luminal narrowing); grade 5 indicated severe arterial stenosis (70 89% luminal narrowing); grade 6 indicated arterial stenosis ( 90%, or occlusion). Arterial stenosis with a grade of 1, 2 or 3 (<50% luminal narrowing) was considered to be hemodynamically insignificant, whereas grades 4, 5 or 6 (50 100% luminal narrowing) were considered hemodynamically significant.

3 Sensitivity and Specificity of ABI in Diagnosing PAD 607 Fig 1. Ankle-brachial index (ABI) of the severity of stenosis of the artery of the lower extremity. Statistical Analysis Data were analyzed using the software program SPSS13.0 (Chicago, IL, USA). Continuous variables are expressed as the mean ± SD, and categorical variables as a percentage. A p-value of 0.05 or less was considered to indicate statistically significant difference. LR+ was defined as sensitivity/ 1-specificity and LR as specificity/1-sensitivity. Receiver operator characteristic (ROC) 20,21 curves were calculated to identify potential ABI cut-off values for diagnosing PAD. Results There are 298 subjects underwent both conventional DSA and ABI measurement (mean age 64.9±11.3 years; 199 men, 99 women). According to our criterion, DSA was used as the gold standard in defining 50% luminal stenosis for the diagnosis of lower extremity PAD, so in this cohort the morbidity of PAD was 7.09%. When DSA was used as the gold standard in defining 30% luminal stenosis for the diagnosis of lower extremity PAD, the morbidity was 19.8%. Baseline Characteristics and Medications of the Study Group As shown in Table 1, 66.8% of the study group were men with an average age of 64.9 years. The average SBP, diastolic BP (DBP), total cholesterol, triglycerides and low-density lipoprotein-cholesterol (LDL-C) were ± mmhg, 71.68±11.71 mmhg, 4.78±2.50 mmol/l, 1.75±1.27 mmol/l and 2.88±0.93 mmol/l, respectively. The average LDL-C of patients with CAD was 2.58± 0.75 mmol/l, and that of patients without CAD was 3.06± 0.68 mmol/l. The following in-hospital medications were the most common among all participants: statins (90.3%), angiotensin-converting enzyme inhibitors (71.1.%), anti-platelet drugs (96.3%), and -adrenergic blocking drugs (66.1%). With regard to the history of illness, 53% had had CAD, 13.4% cerebral infarction, 22.8% had type 2 diabetes mellitus and 33.9% had dyslipidemia. ABI of Differing Levels of Stenosis of Lower Extremity Arteries As shown in Fig 1, the greater the stenosis of the artery of the lower extremity, the lower the ABI value. Table 2 Sensitivity (True-Positive Ratio), Specificity (True-Negative Ratio), and Positive and Negative LR of the ABI at Different Cut-Off Points ABI cut-off point Sensitivity(%) Specificity(%) LR+ LR Digital subtraction angiography used as gold standard for defining 50% luminal stenosis as diagnosis of lower extremity peripheral arterial disease. LR, likelihood ratios; ABI, ankle-brachial index. Sensitivity, Specificity, Positive and Negative LR of ABI The sensitivity, specificity, positive and negative LR ratios of the ABI screening test at different cut-off points for detecting stenoses 50% compared with angiography are shown in Table 2. When DSA was used as the gold standard in defining 50% luminal stenoses for the diagnosis of lower extremity PAD, the sensitivity (true-positive ratio), specificity (truenegative ratio), positive and negative LR ratios of different cut-off points of the ABI were as follows: at a cut-off point of 1.12, sensitivity 100, specificity 40, LR+ 1.67, LR 0; at a cut-off point of 0.95, sensitivity 91, specificity 86, LR+ 6.7, LR 0.1; at a cut-off point of 0.90, sensitivity 76, specificity 90, LR+ 7.6, LR 0.27; at a cut-off point of 0.53, sensitivity 14.3, specificity 100, LR+ 0.14, LR ROC Curve Fig 2 shows conventional DSA used as the gold standard in defining 30%, 50% and 70% stenosis and the respective areas under the ROC curves were (95%confidence interval (CI): 0.712, 0.860), (95% CI: 0.869, 0.984), and (95%CI 0.927, 0.999). Discussion This study provided information about the diagnostic value of the ABI for assessing PAD in Chinese patients. We conclude that the ABI value shows a decreasing tendency with increasing severity of stenosis in patients with PAD (Fig 1). Most of the recent studies have used a cut-off point of ABI < Carter considered that an ABI <0.90 had 95% sensitivity and 90% specific for angiographically documented PAD. 27 That cut-off point has been also accepted by the Strong Heart Study in which the upper limit of normal

4 608 GUO X et al. Fig 2. The ankle-brachial index (ABI) for different degreea of stenoses of the artery of the lower extremity. (A) Receiver operator characteristic (ROC) curve of ABI value in diagnosing peripheral arterial disease (PAD) (conventional digital subtraction angiography (DSA) was used as the gold standard for defining 30% luminal stenoses of the artery of the lower extremity). The area under the ROC curve is (95% confidence interval (CI) 0.712, 0.860). (B) ROC curve of the ABI value for diagnosing PAD (conventional DSA was used as the gold standard for defining 50% luminal stenoses in the artery of the lower extremity). The area under the ROC curve is (95%CI 0.869, 0.984). (C) ROC curve of the ABI value for diagnosing PAD (conventional DSA was used as the gold standard for defining 70% luminal stenoses of the artery of the lower extremity). The area under the ROC curve is (95%CI 0.927, 0.999). ABI values did not exceed More recently, the American College of Cardiology and the American Heart Association (ACC/AHA) guidelines recommended an ABI 0.90 as the criterion for the diagnosis of PAD. 17 However, this cut-off point was based on research in European and American populations and there has not been a report about whether the threshold value differs between races. The current study shows that when conventional DSA was used as the gold standard in defining 30%, 50% and 70% luminal stenosis in the artery of the lower extremity, the area under the ROC curve was (95% CI 0.712, 0.849), (95% CI 0.869, 0.984) and (95% CI 0.927, 0.999), respectively (Fig 1). The area under the ROC curve is a measure of the performance of a test and is used for comparing diagnostic tests independent of the threshold values used. ROC curves are used to compare the accuracy of 2 or more tests over a range of cut-off points and the ROC curve for a given test is constructed by plotting sensitivity (true-positive rate) against 1 specificity (false-positive rate). The points on a curve represent pairs of sensitivity and specificity at the various cut-off levels selected for a given positive diagnosis. The greater the area under the curve, the more accurate the test. An area equal to 1.0 indicates perfect performance of the test, whereas an area of 0.5 or <0.5 indicates no discriminatory power. In the present study, although the greatest area under the ROC curve was when 70% luminal stenosis of the artery of the lower extremity was defined as the gold standard, the number of patients was too small (only 1 subject), compared with 21 subjects in the group with 50% luminal stenosis., which had an area under the ROC curve of 0.927, close to Thus, 50% luminal stenosis of the artery of lower extremity was used as the gold standard in our study and we subsequently analyzed the sensitivity, specificity, positive and negative LR of different ABI cut-off points (Table 2). Sensitivity (true-positive ratio) is defined as the proportion of patients with the disease who have a positive test. 29 Specificity (true-negative ratio) is defined as the proportion of patients without the disease who have a negative test. 17 Sensitivity and specificity convey information about the usefulness of a test in making a diagnosis. Once the test has been performed, the sensitivity and specificity do not indicate whether a positive result truly means the presence of disease; that information is given by the predictive values. The LR are an alternative, newer method of judging the accuracy of a test. Some authors argue that LR are more useful than sensitivity and specificity because the LR compare the proportion of patients with the disease that has a positive (or negative) test result with the proportion of patients without the disease that has a positive (or negative) test result. The LR is the ratio of these 2 proportions or likelihoods. 30 The higher the sensitivity, the bigger the LR+ and LR+ >10 and LR <0.1 are credible proof of the inclusion and exclusion of a certain disease. Even when using a cut-off point of 0.95, that value was associated the highest possible levels of sensitivity (91%), specificity (86%) and LR+ (6.50) and the lowest LR (0.10). A LR of 0.10 is likely to result in a post test probability that excludes PAD with a very high degree of certainty. Thus, we can conclude that an ABI value of 0.95 is the threshold value in diagnosing PAD in Chinese patients, which is somewhat inconsistent with the ACC/AHA guideline suggesting that ethnic differences should be considered in the diagnosis.

5 Sensitivity and Specificity of ABI in Diagnosing PAD Our study also demonstrates that when DSA was used as the gold standard for defining 50% luminal stenosis for the diagnosis of lower extremity PAD the prevalence of lower extremity PAD in the total study population was 6.5%, and the average age of the study group was 64.9 years. This prevalence is accordance with the result of a recently reported study, in which PAD was defined as an ABI <0.9 in either leg, and the prevalence of PAD ranged from 7% in persons aged years to 23% in persons aged 80 years and older. 31 However, it is difficult to compare our PAD estimates with previously reported estimates from other studies. Two previous population-based studies 7,32 used the criterion of ABI <0.9 to determine PAD. The Cardiovascular Health Study reported an overall 12.4% prevalence of PAD, higher than the overall prevalence of 7.09% in the present study. 7 The Rotterdam Study reported an overall PAD prevalence of 19%, ranging from approximately 12% for ages years to more than 55% for individuals aged 85 years and older, the latter also being a higher rate of PAD than our estimates. 32 Two recent population-based studies in China also used the criterion of ABI <0.9 to define PAD, and the prevalence of PAD in the total patients with high cardiovascular risk was 25.4%. 33,34 However, when DSA was used as the gold standard in defining 30% luminal stenosis, the prevalence of lower extremity PAD in all study population was 19.8%, which is consistent with the previous report. 32 There were some reasons that could account for the lower rate of PAD in the present study. Risk factors for atherosclerosis, such as cigarette smoking, diabetes, dyslipidemia, hypertension, and hyperhomocysteinemia, increase the likelihood of developing lower extremity PAD, as they do for other manifestations of atherosclerosis. The risk factors of PAD were relatively effectively controlled in our study population, so the prevalence of PAD was lower than that reported by previous studies. The effectively controlled risk factors could also account for the different morbidity when DSA was used as a gold standard in defining 30% and 50% luminal stenoses. To our surprise, in the present study the prevalence of PAD was 16.2% and 19.32%, respectively, when the threshold value of ABI was defined as 0.9 and 0.95 for detecting PAD. The reasons for the difference in the incidence of PAD when different methods are adopted will be discussed in our future research. Study Limitations First, there was a strong selection bias because patients in this investigation only included those who had high cardiovascular risk. Another is the use of angiography as the gold standard for evaluating a noninvasive test. Angiography gives morphologic information about the lower extremity arteries, in contrast to the noninvasive test (ABI), which only provides hemodynamic information. Finally, we did not evaluate the interobserver variability of performing the ABI because it was performed in different cities; however, the ABI is a highly reproducible test with minimal interobserver variability being reported. 17 In conclusion, our results demonstrate that ABI measurement is a simple, feasible, noninvasive and reliable test for assessing arteriosclerosis in the lower extremity arteries. Moreover, because of its noninvasive nature, the ABI is an alternative to conventional DSA, with high sensitivity and specificity in Chinese patients with PAD. The cut-off point is the threshold ABI value for diagnosing PAD in Chinese high-risk cardiovascular patients and in clinical practice, the physician should pay much attention to the measurement of ABI value in such patients. Acknowledgement This study was financially supported by Omron (China) Co Ltd and Beijing Century Trade Corp. References 1. Hiatt WR. Medical treatment of peripheral arterial disease and claudication. N Engl J Med 2001; 344: Meijer WT, Hoes AW, Rutgers D, Bots ML, Hofman A, Grobbee D. Peripheral arterial disease in the elderly: The Rotterdam study. Arterioscler Thromb Vasc Biol 1998; 18: Bowlin SJ, Medalie JH, Flocke SA, Zyzanski SJ, Goldbourt U. Epidemiology of intermittent claudication in middle-aged men. Am J Epidemiol 1994; 140: McDermott MM, Liu K, Greenland P, Guralnik JM, Criqui MH, Chan C, et al. Functional decline in peripheral arterial disease: Associations with the ankle brachial index and leg symptoms. JAMA 2004; 292: Murabito JM, Evans JC, Larson MG, Nieto K, Levv D, Wilson PW. The ankle-brachial index in the elderly and risk of stroke, coronary disease, and death: The Framingham Study. Arch Intern Med 2003; 163: Newman AB, Siscovick DS, Manolio TA, Polak J, Fried LP, Borhani NO, et al. Ankle-arm index as a marker of atherosclerosis in the Cardiovascular Health Study. Circulation 1993; 88: Leng GC, Fowkes FG, Lee AJ, Dunbar J, Housley E, Ruckley CV. Use of ankle brachial pressure index to predict cardiovascular events and death: A cohort study. BMJ 1996; 313: Li J, Luo YY, Xu YW, Yang JG, Zheng LQ, Buaijiaer H, et al. Risk factors of peripheral arterial disease and relationship between low ankle brachial index and mortality from all-cause and cardiovascular disease in Chinese patients with type 2 diabetes. Circ J 2007; 71: Rehring TF, Sandhoff BG, Stolcpart RS, Merenich JA, Hollis HW Jr. Atherosclerotic risk factor control in patients with peripheral arterial disease. J Vasc Surg 2005; 41: Stoffers HE, Kester AD, Kaiser V, Rinkens PE, Knottnerus JA. Diagnostic value of signs and symptoms associated with peripheral arterial occlusive disease seen in general practice: A multivariable approach. Med Decis Making 1997; 17: Winsor T. Influence of arterial disease on the systolic blood pressure gradients of the extremity. Am J Med Sci 1950; 220: Greenland P, Abrams J, Aurigemma GP, Bond MG, Clark LT, Criqui MH, et al. Prevention Conference V: Beyond secondary prevention: Identifying the high-risk patient for primary prevention: Noninvasive tests of atherosclerotic burden. Circulation 2000, 101: e16 e Selvin E, Erlinger T. Prevalence of and risk factors for peripheral arterial disease in the United States: Results from the National Health and Nutrition Examination Survey, Circulation 2004; 110: Ogren M, Hedblad B, Isacsson SO, Janzon L, Jungquist G, Lindell SE. Non-invasively detected carotid stenosis and ischaemic heart disease in men with leg arteriosclerosis. Lancet 1993, 342: Newman AB, Shemanski L, Manolio TA, Cushman M, Mittelmark M, Polak JF, et al. Ankle-arm index as a predictor of cardiovascular disease and mortality in the Cardiovascular Health Study: The Cardiovascular Health Study Group. Arterioscler Thromb Vasc Biol 1999; 19: National Cholesterol Education Program. 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). Final report. Bethesda, MD: National Heart, Lung, and Blood Institute, National Institutes of Health, Hirsch AT, Criqui MH, Treat-Jacobson D. Peripheral arterial disease detection, awareness, and treatment in primary care. JAMA 2001; 286: Oser RF, Picus D, Hicks ME, Darcy MD, Hovsepian DM. Accuracy of DSA in the evaluation of patency of infrapopliteal vessels. J Vasc Interv Radiol 1995; 6: Beckman JA, Higgins CO, Gerhard-Herman M. Automated oscillometric determination of the ankle brachial index provides accuracy

6 610 GUO X et al. necessary for office practice. Hypertension 2006; 47: Pan CR, Staessen JA, Li Y, Wang JG. Comparison of three measures of the ankle-brachial blood pressure index in a general population. Hypertens Res 2007; 30: Chu K. An introduction to sensitivity, specificity, predictive values and likelihood ratios. Emerg Med 1999; 11: Sasse EA. Objective evaluation of data in screening for disease. Clin Chim Acta 2002; 315: Resnick HE, Lindsay RS, McDermott MM, Devereux RB, Jones KL, Fabsitz RR, et al. Relationship of high and low ankle brachial index to all-cause and cardiovascular disease mortality: The Strong Heart Study. Circulation 2004; 109: Hiatt WR, Marshall JA, Baxter J, Sandoval R, Hildebrandt W, Kahn LR, et al. Diagnostic methods for peripheral arterial disease in the San Luis Valley Diabetes Study. J Clin Epidemiol 1990; 43: Tseng CH, Tseng CP, Tai TY, Chong CK. Effect of angiotensin blockade on the association between albuminuria and peripheral arterial disease in elderly Taiwanese patients with type 2 diabetes mellitus. Circ J 2005; 69: Carter SA. The role of pressure measurements in vascular disease. In: Bernstein EF, editor. Non-invasive diagnostic techniques in vascular disease. St Louis: Mosby; 1985; Fabsitz RR, Sidawy AN, Go O, Lee ET, Welty TK, Devereux RB, et al. Prevalence of peripheral arterial disease and associates risk factors in American Indians: The Strong Heart Study. Am J Epidemiol 1999; 149: Hirsch AT, Haskal ZJ, Hertzer NR, Bakal CW, Rosenfield KA, Creager MA, et al. ACC/AHA 2005 guidelines for the management of patients with peripheral arterial disease (lower extremity, renal, mesenteric, and abdominal aortic): A collaborative report from the American Association for Vascular Surgery/Society for Vascular Surgery, Society for Cardiovascular Angiography and Interventions, Society for Vascular Medicine and Biology, Society of Interventional Radiology, and the ACC/AHA Task Force on Practice Guidelines (Writing Committee to Develop Guidelines for the Management of Patients with Peripheral Arterial Disease). J Am Coll Cardiol 2006; 47: e1 e Jaeschke R, Guyatt GH, Sackett DL for the Evidence-Based Medicine Working Group. Users guides to the medical literature. III: How to use an article about a diagnostic test. A: Are the results of the study valid? JAMA 1994; 271: Woolson RF. Statistical methods for the analysis of biomedical data. New York: Wiley & Sons; Ostchega Y, Paulose-Ram R, Dillon CF, Gu Q, Huqhes JP. Prevalence of peripheral arterial disease and risk factors in persons aged 60 and older: Data from the national health and nutrition examination survey Am J Geriatr Soc 2007; 55: Meijer WT, Grobbee DE, Hunink MG, Hofman A, Hoes AW. Determinants of peripheral arterial disease in the elderly: The Rotterdam Study. Arch Intern Med 2000; 160: Hasimu B, Li J, Nakayama T, Yu JM, Yang JG, Li XK, et al. Ankle brachial index as a marker of atherosclerosis in Chinese patients with high cardiovascular risk. Hypertens Res 2006; 29: Hasimu B, Li J, Ma Y, Zhao MZ, Nakayama T, Ma W, et al. Evaluation of medical treatment for peripheral arterial disease in Chinese high-risk patients. Circ J 2007; 71: