The prevalence of peripheral artery disease (PAD) in industrialized countries is high, and is expected to increase as the population ages.1, 2 In Spain the population prevalence of PAD in individuals older than 55 years is 8.03%.3 Various studies have shown that patients with PAD have a high risk of coronary complications and death4 and that the ankle-brachial index may help improve the prediction of cardiovascular risk.5 Consequently, in patients with PAD this is considered as an equivalent of coronary disease6, 7 and, therefore, aggressive preventive therapy should be prescribed to reduce their high cardiovascular risk. Various clinical guidelines recommend that this population receive antiplatelet therapy and attain certain therapeutic goals to control their cardiovascular risk factors. In particular, the guidelines recommend statin therapy to reduce the low-density lipoprotein cholesterol (LDL-C) concentration to below 100mg/dL, control of blood pressure with differing targets depending on the presence of other comorbidities, and treatment with drugs that block the renin-angiotensin system.8, 9, 10, 11

Numerous studies have assessed the extent to which these treatments are prescribed and the therapeutic goals are attained in patients with established coronary disease12, 13 as well as their relationship to the decrease in cardiovascular morbidity and mortality.14 However, few studies have assessed the rates of prescription in patients with PAD,15 and even fewer have looked at the data in Spain.16

The aim of this study was to evaluate the rate of evidence-based prescription to reduce cardiovascular risk and the extent to which cholesterol and blood pressure targets are achieved in a broad sample of Spanish patients with PAD but no associated coronary or cerebrovascular disease.

PERIFERICA is a cross-sectional, observational epidemiologic study conducted in outpatient clinics staffed by primary care, internal medicine, vascular surgery, and other specializations that usually see patients with PAD, such as cardiology, endocrinology, and nephrology. Each investigator was asked to include 10 consecutive patients aged ≥45 years, with documented clinical diagnosis of PAD (ankle-brachial index<0.9 with or without clinical symptoms of intermittent claudication, lower limb amputation due to chronic ischemia, or a previous revascularization procedure at this level) and a blood workup that included lipid and lipoprotein concentrations in the previous 6 months. The study excluded patients with a history of coronary disease (angina, myocardial infarction, or revascularization procedure) and/or cerebrovascular disease (stroke, transient ischemic attack, or cognitive deterioration of possible vascular origin), subjects with uncontrolled hypothyroidism, and patients with any serious or terminal illness.

A medical history was taken from all participants, and all participants underwent a physical examination that included measurements of weight, height, waist circumference, and blood pressure measured on 3 occasions, 2min apart.

Patients were considered hypertensive if they were taking antihypertensive drugs or blood pressure values >140/90mmHg were recorded on 2 occasions more than 1 week apart, prior to study inclusion. Diabetes mellitus was diagnosed according to the American Diabetes Association recommendations.17 Heart failure and atrial fibrillation were diagnosed according to the patient's medical history. The estimated glomerular filtration rate was calculated using the simplified Modification of Diet in Renal Disease study equation.18

For this study, well-controlled blood pressure was considered to be <130/80 mmHg in patients with diabetes, acute heart failure, or glomerular filtration rate<60 mL/min/1.73 m2, and <140/90mmHg in the rest of the population. Target bloodwork was considered to be LDL-C<100mg/dL, high-density lipoprotein cholesterol (HDL-C)>40 mg/dL in men and >45mg/dL in women, and triglycerides<150mg/dL. Control of all risk factors was considered to be optimal if the therapeutic goals for blood pressure and LDL-C had been reached and the patient did not smoke.

The quantitative variables were compared by the single-factor ANOVA test in the case of normal distribution or by the Kruskall-Wallis H test if the parametric assumptions were not met. The discrete variables were compared by the χ2 test.

In order to evaluate factors associated with control of the risk factors, a forward stepwise logistic regression model was used. In this model, all factors associated with control and all clinically relevant variables were introduced in the univariate model.

Statistical significance was set at P<.05. The statistical analysis was performed using SPSS V17.0.

The study was approved by the ethics committee of Hospital Carlos III in Madrid.

From May to December 2009, 440 doctors included a total of 4321 patients, of whom 234 (5.4%) were excluded because they failed to meet at least 1 inclusion criterion, yielding a total of 4087 (94.6%) assessable subjects. Patient data were collected from all autonomous communities except the Autonomous City of Ceuta and the Autonomous City of Melilla. Andalusia (16.1%), Catalonia (15.6%), the Valencian Community (12%), and the Community of Madrid (10%) were the regions with the highest number of participants. Most participants were from primary care clinics (56.5%), followed by outpatient clinics (22% internal medicine, 9.4% vascular surgery, 7.7% cardiology, and 4.5% endocrinology or nephrology).

The prevalence of risk factors was very high (Table 1). Half the participants had diabetes, with a mean disease history of 10.6 years; the estimated glomerular filtration rate was 30-60 mL/min/1.73 m2 in 26.3% of participants and <30 mL/min/1.73 m2 in 2.4%.

Table 1. Clinical Characteristics and Analytical Results in the 4087 Assessable Study Participants.

HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol.
Unless otherwise specified, the data are expressed as mean (standard deviation) or median [interquartile range].

The mean age at the time of PAD diagnosis was 63.2 years; 15.4% had a history of revascularization (surgical in two thirds and endovascular in one third) and 8.4% had undergone amputation of some kind (minor in 5.4%; major in 3%).

A total of 71% of patients reported some kind of daily physical activity, 38.1% for less than 30min and 33% for more than 30min.

Lipid-lowering drugs were prescribed to 79.1%; of these, 76.2% were receiving statins and 11.4% other lipid-lowering drugs. A total of 83.4% were receiving antithrombotics (75.4% antiplatelets and 10.8% anticoagulants). Antihypertensive drugs were used by 85.5%; 66% were receiving a renin-angiotensin blocker, 36.8% diuretics, 33.5% calcium antagonists, 9.9% beta blockers, and 6.6% other antihypertensives. A total of 36.2% were taking oral hypoglycemic drugs and 18.7% were receiving insulin. Up to 45% were taking some drug for claudication.

Blood pressure was controlled in 29.5% of participants, and bloodwork targets had been attained for LDL-C by 30.4%, for HDL-C by 69.6%, and for triglycerides by 54.2%. Only 13% of participants had adequate control of all 3 lipid variables.

Among patients with diabetes mellitus, 24% had glycohemoglobin<6.5% and 43.2% had a level <7%.

Because 25.5% were active smokers, only 7.9% of all participants had optimal control of all risk factors (ie, blood pressure and LDL-C on target) and were not actively smoking.

Analyses were performed to evaluate the factors associated with control of each risk factors alone and control of the risk factors as a whole.

The factors associated with adequate LDL-C control (Table 2) were diabetes, Fontaine stage I, and care from a medical specialist. Conversely, LDL-C was more poorly controlled in women and active smokers. The percentage of patients with LDL-C on target did not depend on lipid-lowering prescription (LDL-C<100mg/dL in 30.5% of patients with prescription and 30.1% without prescription; P=.796).

Table 2. Factors Associated With Poor Low-density Lipoprotein Cholesterol Control. Multivariate Analysis.

95%CI, 95% confidence interval; AOR, adjusted odds ratio.
The variables included in the model as clinically significant in the univariate analysis were age, sex, medical specialization of the attending physician, smoking, diabetes mellitus, history of revascularization surgery or amputation, Fontaine stage, and glomerular filtration rate. The variables included as clinically relevant were educational level and lipid-lowering and antihypertensive therapy.

Blood pressure control was better (Table 3) in the oldest patients and in patients with high school or university studies. In contrast, control was worse in patients with acute heart failure or low glomerular filtration rates, groups in which the targets for good control were stricter. The percentage of patients with good control was higher among subjects not receiving antihypertensives (37.7% vs 28.1%; P<.001).

Table 3. Factors Associated With Poor Blood Pressure Control. Multivariate Analysis.

95%CI: 95% confidence interval; AOR: adjusted odds ratio; GFR: glomerular filtration rate.
The variables included in the model as clinically significant in the univariate analysis were age, sex, medical specialization of the physician attending the patient, educational level, physical activity, diabetes mellitus, heart failure, and glomerular filtration rate. The variables included in the model as clinically relevant were hypertension, atrial fibrillation, Fontaine stage, and antihypertensive therapy.

The factors associated with optimal control of all risk factors (blood pressure, LDL-C, and no smoking) were advanced age, patients under the care of specialists other than general practitioners (usually an endocrinologist or nephrologist), higher educational level, atrial fibrillation, and daily physical exercise (Table 4). On the other hand, a history of hypertension lowered the possibilities of optimal control.

Table 4. Factors Associated With Not Achieving Optimal Control of Cardiovascular Risk Factors (Blood Pressure, Low-density Lipoprotein Cholesterol, and Nonsmoker). Multivariate Analysis.

95%CI, 95% confidence interval; AOR, adjusted odds ratio.
The variables included in the model as clinically significant in the univariate analysis were age, sex, medical specialization of the physician attending the patient, educational level, diabetes mellitus, hypertension, atrial fibrillation, history of revascularization surgery or amputation, Fontaine stage, physical activity, and glomerular filtration rate. The variables included in the model as clinically relevant were heart failure and lipid-lowering and antihypertensive therapy.

Subjects with PAD have a high rate of cardiovascular complications and death.4 In these patients, the use of antiplatelet drugs, statins, and renin-angiotensin blockers has been associated with better prognosis,19, 20 even though the rate of prescription is lower than recommended.15 Most studies that evaluated their use, however, were carried out many years ago, at a time when PAD was not so widely accepted as a predictor of coronary risk as at present and the rate of evidence-based prescription for cardiovascular prevention was scant, even in patients with established coronary disease.12 In addition, all previous studies included patients regardless of whether they presented clinical manifestations in another vascular territory or not. This could explain why preventive treatment was prescribed for coronary or cerebral arteriosclerosis rather than peripheral arteriosclerosis. Our data show that during 2009 the prescription of antithrombotic drugs, lipid-lowering drugs, and renin-angiotensin blockers in patients with PAD in Spain was high, but could have been better. The use of statins was higher than that observed in participants with PAD included in the REACH study,16 a fact of particular relevance because our study included no patients with known vascular disease in any other territory, in order to better estimate the role attributable to PAD in the use of preventive resources. The use of these drugs in 2009, however, did not differ noticeably from that observed in coronary patients in 2005.12

The high prescription rate for lipid-lowering drugs is in contrast to the low percentage of patients with optimal LDL-C control. This percentage is similar to the one observed in other studies that have assessed the degree of control in patients with coronary disease,12 and somewhat lower than that seen in recent studies conducted in high-risk patients.21 Several reasons could explain this low rate of patients who meet the therapeutic goals. In addition to the 21% of patients not receiving lipid-lowering drugs, the use of statins with little potency or at suboptimal doses could produce a lower decrease than necessary. Another factor that can make it difficult to reach the goals is poor therapeutic adherence. Numerous studies have shown that a very significant percentage of patients discontinue statin therapy rapidly. This occurs more often in patients in primary prevention than in subjects who have already had a coronary complication,22 probably due to greater awareness among coronary patients of the importance of controlling risk factors to avoid a new event. The data suggest that subjects with PAD are not aware of the severity of the diagnosis. In any case, the problem does not appear to be a lack of therapeutic alternatives, as 70% failed to meet the targets but only 3% were taking 2 or more lipid-lowering drugs.

The factors associated with worse LDL-C control were female sex and active smoking, whereas diabetes and care from a specialist were associated with better control. Women were just as likely as men to use lipid-lowering drugs (data not shown), although women had a higher mean LDL-C, indicating that lower doses of lipid-lowering drugs were administered or the baseline cholesterol concentrations were higher. The latter is likely, if we consider that women have a lower prevalence of arteriosclerosis, and a greater burden of risk factors (eg, hypercholesterolemia) is needed for this disease to develop. Smoking is probably a marker of therapeutic noncompliance or medical inertia in patients who fail to cooperate in the control of their risk factors. Diabetics tend to have a normal LDL-C concentration, which is no different than that of the general population, and therefore it is presumably easier to lower their LDL-C to below 100mg/dL with lipid-lowering therapy.

The percentage of patients with adequate blood pressure control was also low, very similar to that found in other studies performed among populations with PAD.16 The percentage of well-controlled patients was higher in those who were older and had higher educational levels. Subjects with acute heart failure and renal failure required stricter blood pressure targets, and this was probably the reason for worse control among these patients. The study had conflicting results concerning the relationship between age and blood pressure control. In general, therapeutic compliance tends to be better at older ages, although some studies have shown that blood pressure control is better in older persons even after correcting for adherence to treatment.23

Optimal control of all cardiovascular risk factors was only seen in 7.9% of subjects. Patients who were older, had university studies, and did physical exercise (all markers associated with better treatment adherence), and those with atrial fibrillation had better control. Conversely, patients with hypertension or under the care of a general practitioner had worse control. It has recently been shown that most of the decrease in the coronary mortality rate observed in the past few years in Spain is attributable to better control of risk factors and the evidence-based prescription for cardiovascular prevention.24 Therefore, it is important to identify patient characteristics associated with optimal control of risk factors.

The present study had several limitations. First of all, the choice of physicians was not randomized. Hence, it may not be possible to extrapolate patient control or drug type to other physicians in Spain. Our data were insufficient to determine the reasons why the goals were not usually attained, despite a high level of prescription of preventive drugs. The study did not consider if this could be due to low adherence or to suboptimal use of therapeutic resources. However, no studies have evaluated such a large sample of patients with PAD and without vascular disease in other territories.

We concluded that the percentage of patients with PAD but no vascular manifestations in any other area who are receiving preventive treatment is high. However, the percentage of patients who reach the therapeutic goals is low. New studies are needed to determine if the low rate of attainment of therapeutic goals is due to low adherence or suboptimal use of treatment.

The PERIFERICA study was funded by Pfizer S.L., Spain.

Dr. J.M. Mostaza, Dr. E. Puras, and Dr. M. Blasco state that they have received fees for consulting and for forming part of the PERIFERICA study coordinating committee.

Statisticians from the pharmaceutical company analyzed the data but did not participate in the discussion of the results or the writing of the present paper.

Received 22 November 2011
Accepted 17 February 2012

Corresponding author: Unidad de Arteriosclerosis y Riesgo Vascular, Servicio de Medicina Interna, Hospital Carlos III, Sinesio Delgado 10, 28029 Madrid, Spain. jmostaza.hciii@salud.madrid.org