Physical Examination and Chronic Lower-Extremity Ischemia a Critical Review

ORIGINAL INVESTIGATION Physical Examination and Chronic Lower-Extremity Ischemia A Critical Review

Steven R. McGee, MD; Edward J. Boyko, MD, MPH

Objective: To determine the clinical utility of physical cians diagnose the presence of peripheral arterial dis- examination in patients with suspected chronic ische- ease: abnormal pedal pulses, a unilaterally cool extrem- mia of the lower extremities. ity, a prolonged venous filling time, and a femoral bruit. Other physical signs help determine the extent and dis- Data Sources: MEDLINE search (January 1966 to tribution of vascular disease, including an abnormal fem- January 1997), personal files, and bibliographies of oral pulse, lower-extremity bruits, warm knees, and the textbooks on physical diagnosis, surgery, and vascular Buerger test. The capillary refill test and the findings of surgery. foot discoloration, atrophic skin, and hairless extremities are unhelpful in diagnostic decisions. Mathematical Study Selection: Both authors independently graded formulas, derived from 2 studies using multivariate analy- the studies as level 1, 2, or 3, according to predeter- sis, allow clinicians to estimate the probability of periph- mined criteria. Criteria deemed essential for analysis of eral arterial disease in their patients. sensitivity, specificity, and likelihood ratios were (1) clear definition of study population, (2) clear definition of Conclusion: Certain aspects of the physical examination physical examination maneuver, and (3) use of an ac- help clinicians make accurate judgments about the pres- ceptable criterion standard test for comparison. ence of peripheral arterial disease and its distribution.

Results: The following positive findings help clini- Arch Intern Med. 1998;158:1357-1364

N HIS book The Circulatory Dis- helpful, especially capillary refill and de- turbances of the Extremities, pub- pendent rubor.2,3 lished in 1924, Buerger intro- Chronic arterial disease usually af- duced to clinicians several new fects 3 distinct segments in the lower limbs: physical signs of peripheral vas- the aortoiliac segment (especially the in- Icular disease. Whereas earlier writers on frarenal abdominal aorta and common iliac vascular disease had confined them- arteries), the femoropopliteal segment (es- selves to descriptions of gangrene, Buerger pecially the superficial femoral artery in stressed the importance of certain “pro- the adductor canal, a tunnel amid muscle dromal” findings, including absent pulses, groups of the middle third of the thigh), toe and foot ulcers, impaired nail growth, and the peroneotibial segment (below the soft tissue atrophy, poor capillary refill (the knee). Most patients have femoropop- “expression test”), foot pallor with eleva- liteal disease, aortoiliac disease, or both.4 tion, and a deep-red color of the limb in Significant peroneotibial disease is un- the lowered position (“erythromelia”).1 common unless the patient has diabetes Modern clinicians have accepted Buer- or thromboangiitis obliterans. According ger’s findings as characteristic of chronic to traditional teachings, disease in each of arteriosclerosis, a somewhat surprising in- these segments causes specific claudi- ference because Buerger considered them cation symptoms and pulse abnormali- most characteristic of thromboangiitis ties, and requires different surgical ap- obliterans, a rare form of small-vessel vas- proaches (Table 1). From the Department of culitis now known as Buerger disease. The ankle-to-arm systolic pressure in- Medicine, Veterans Affairs Moreover, other prominent physicians, dex (AAI), the most common criterion Puget Sound Health Care writing soon after the appearance of Buer- standard used in studies of physical diag- System, University of ger’s book, argued that some of Buerger’s nosis, is obtained by measuring the high- Washington, Seattle. findings were unreliable and clinically un- est systolic blood pressure at the ankle

ARCH INTERN MED/ VOL 158, JUNE 22, 1998 1357

Downloaded From: https://jamanetwork.com/ on 09/30/2021 0.97 identify patients with angiographically proven occlusions or significant stenoses with a sensitivity of 96% MATERIALS AND METHODS to 97% and specificity of 94% to 100%.6,8 Most patients with claudication have AAIs between 0.5 and 0.8 and dis- One of us (S.R.M.) performed a MEDLINE search of ease in a single segment; those with rest pain and gangrene have AAIs less than 0.5 and disease in at least 2 articles published between January 1966 and Janu- 6-8 ary 1997 to retrieve all relevant publications on the segments. bedside diagnosis of peripheral vascular disease (the This review addresses the diagnostic accuracy of specific strategy was arterial occlusive diseases/di AND many of Buerger’s prodromal findings, compared with physical examination OR peripheral vascular dis- the AAI and other criterion standard measures, and ease/di OR intermittent claudication/di OR pulse/), whether certain aspects of the physical examination ac- all limited to the English language and human sub- tually help clinicians accurately determine the location jects. Based on review of titles and abstracts, rel- of the patient’s arterial occlusions and stenoses, as sug- evant publications were retrieved, including all ar- gested in Table 1. ticles that determined interobserver variability and all that compared physical signs with an objective criterion standard. To complete the search, the bibli- RESULTS ographies of these articles and those of textbooks on physical diagnosis, surgery, and vascular surgery were reviewed. DIAGNOSIS OF PERIPHERAL ARTERIAL DISEASE Both of us independently examined the se- lected studies for the following criteria: (1) clear defi- Examination of the Pedal Pulses nition of the study population, (2) clear definition of the physical examination maneuver, (3) use of an To best understand the diagnostic value of the pulse ex- acceptable criterion standard test (calculation of AAI, direct pressure measurement, angiography), (4) un- amination, it is important to first address interobserver biased selection of patients (ie, consecutive patients variation and the normal variations in the vascular per- or population-based sample), (5) predefined abnor- fusion of the foot. When deciding whether a particular mal thresholds for the criterion standard, and (6) pulse is present or absent, clinicians demonstrate fair to blinded comparison of test and criterion standard re- almost perfect agreement (␬=0.20-0.92) (Table 2), al- sults. Studies satisfying all 6 criteria were graded level though attempts to distinguish the normal and reduced 1; those with fewer than all 6 but satisfying at least pulse cause considerable dissension (␬=0.01-0.15) (Table 1, 2, and 3 were graded level 2; and all others were 2). Table 3 reveals that, in studies of large numbers of graded level 3. We discussed and resolved any dis- healthy individuals, the dorsalis pedis, posterior tibial, agreement in classification by consensus. and femoral pulses are not palpable 8.1%, 2.9%, and 0% Of the 17 studies identified, 7 were graded level 37 9-15 16-20 of the time, respectively (Table 3), findings that are con- 1, 5 were graded level 2, and 5 were graded level 35 3.21-25 Four studies10,11,16,21 were excluded from the sistent across all age groups. Even when a physician can- analysis of the pulse examination because each de- not palpate one of the pedal pulses of a healthy indi- fined as abnormal the absence of only 1 of the 2 pedal vidual, however, it is usually audible by a Doppler pulses, a common finding in healthy individuals (see flowmeter, the dorsalis pedis artery being congenitally below) and a biologically implausible sign of dis- absent in only 2% of extremities and the posterior tibial ease (lower-extremity ischemia, in any of the upper in 0.1%.39,40 Anatomists dissecting normal feet have found 2 segments, should cause both pedal pulses to dis- that when the dorsalis pedis artery is small or absent, the appear). Level 3 studies and other citations may ap- posterior tibial artery is prominent; when the posterior pear in the discussion if they provide original de- tibial artery is small, the dorsalis pedis is prominent.41 scriptions of physical findings, unique insights, or data that support findings from level 1 and 2 studies. How- These normal patterns of collateral circulation not only ever, we used only level 1 and 2 studies for calcula- help explain why clinicians use whichever pedal pulse tions of sensitivity, specificity, and likelihood ratios has the highest blood pressure to calculate the AAI, but (LRs). also why only 0.7% of normal extremities are missing both Data were expressed as sensitivity, specificity, pedal pulses (Table 3). positive and negative LRs, and the ␬ statistic accord- Consistent with these anatomic studies and inves- ing to traditional definitions.26 If a particular LR, posi- tigations in healthy subjects, the absence of both pedal tive or negative, has a value close to 1, that outcome pulses has been shown to be a powerful predictor for the of the test is unhelpful in making diagnostic deci- presence of vascular disease (AAIϽ0.9) in the extremity sions at the bedside.27 examined (LR+=9.0-44.6; Table 4). For definitions of disease that include only the most severe cases (AAIϽ0.5), however, the positive LRs are lower (LR+=3.0-3.8), prob- ably because of diminished specificity (ie, many pa- (dorsalis pedis and posterior tibial arteries) with a hand- tients with AAIs between 0.5 and 0.9 also lack pedal held Doppler flowmeter and dividing it by the blood pres- pulses). sure in the brachial artery. In healthy individuals, the AAI Nevertheless, up to about one third of patients with is greater than 0.97 (this cutoff represents the lower 2.5% disease, sometimes at limb-threatening levels (ie, of measurements from large numbers of young, non- AAIϽ0.5),12 have at least 1 palpable pedal pulse (sensi- smoking, asymptomatic individuals).6-9 Values less than tivity, 0.63-0.95) (Table 4). Although some correlation

ARCH INTERN MED/ VOL 158, JUNE 22, 1998 1358

Downloaded From: https://jamanetwork.com/ on 09/30/2021 Table 1. Diagnosis and Management of Chronic Arterial Disease of the Lower Extremities: Classic Approach

Pulse Examination

Anatomic Segment Location of Claudication Femoral* Popliteal Pedal Usual Surgical Procedure Aortoiliac Buttock, thigh, calf† Absent Absent Absent Aortofemoral bypass with bifurcated prosthetic graft‡ Femorofemoral bypass (if patient cannot tolerate abdominal procedure and other iliac widely patent) Axillobifemoral bypass (if patient cannot tolerate abdominal procedure and both iliac arteries diseased) Femoropopliteal* Calf Present Absent Absent Femoropopliteal bypass with saphenous vein Peroneotibial None or foot§ Present Present Absent Femorotibial or femoroperoneal bypass with saphenous vein (if disease limited to proximal-mid calf and good distal runoff)

*“Femoro” of “femoropopliteal” indicates superficial femoral artery and of “femoral pulse,” common femoral artery. †May cause erectile dysfunction if internal iliac arteries are involved. ‡In the aortoiliac segment, percutaneous transluminal angioplasty is preferred for stenoses (not occlusions) of short segments (Ͻ10 cm).5 §Disease in this segment usually causes no claudication in patients with diabetes but causes foot pain in those with thromboangiitis obliterans.

Table 2. Palpation of Pulse: Interobserver Agreement

No. of No. of Study* Patients Examiners Pulses Examined Findings† Lawson et al28 63 3 Posterior tibial, dorsalis pedis (graded 0-4) Weighted ␬: 0.20-0.72 (before training); 0.58-0.87 (after training) Myers et al29 22 6 Femoral, popliteal Pulse present or absent: ␬ = 0.53 (femoral); ␬=0.52 (popliteal) Pulse normal or reduced: ␬ = 0.15 (femoral); ␬ = 0.01 (popliteal) Brearly et al30 5 2 Femoral, popliteal, anterior tibial, posterior Pulse present or absent: ␬ = 0.92 tibial, dorsalis pedis

*Studies that expressed data as simple agreement 31-33 are not included in this table, unless raw data were available to calculate the ␬ statistic.30 †Conventional interpretation of ␬ statistic: 0-0.2, slight agreement; 0.2-0.4, fair agreement; 0.4-0.6, moderate agreement; 0.6-0.8, substantial agreement; and 0.8-1.0, almost perfect agreement.34

Table 3. Absent Pedal Pulses by Palpation in Healthy Subjects

% Extremities Missing*

Study No. of Extremities Age, y Dorsalis Pedis (DP) Posterior Tibial (PT) Both DP and PT Morrison35 2000 1-70+ 6.1 10.1 1.9 Silverman36 2028 20 (mean) 12.5 2.8 0.2 Stephens37* 400 35 (mean) 3.0 0.3 . . .† Barnhorst and Barner38 2000 1-10 8.7 0 0 Nuzzaci et al39 720 14-79 13.8 2.6 . . .† Robertson et al40 1094 9-30 4.5 1.3 . . .† Mean % NA† NA† 8.1 2.9 0.7

*The femoral pulse was present in every limb. †Ellipses indicate no data were available from these studies; NA, not applicable.

exists between an arterial blood pressure and whether its In some patients with false-negative examination pulse is palpable,30 this relationship is variable. In the dor- results (ie, disease is present but pedal pulses are pal- salis pedis artery, for example, the range of systolic pres- pable), the pedal pulses disappear during exercise.44-47 sures associated with palpable pulses (mean ± SD, 151±27 Just as exercise testing uncovers significant coronary mm Hg; range, 64-220 mm Hg) overlaps significantly with artery stenoses, exercise of the extremities presumably those that are nonpalpable (mean, 112±56 mm Hg; range, results in a disproportionate distribution of blood to 42-300 mm Hg).16 No absolute pressure threshold ex- better-perfused muscle and away from poorly perfused ists below which the pulse becomes nonpalpable, sug- distal tissue, causing the extremity to become symp- gesting that other variables such as a highly flexible ves- tomatic, the pressure to fall, and the pedal pulse to sel wall, high rate of arterial runoff distal to the pulse, disappear. Exactly what proportion of diseased high cardiac output, superficial location of artery, low extremities with false-negative examination findings viscosity of blood, or certain distributions of disease, may become positive with exercise has never been investi- preserve a palpable pulse despite low pressure.42,43 gated, but a related series of investigations has shown

ARCH INTERN MED/ VOL 158, JUNE 22, 1998 1359

Downloaded From: https://jamanetwork.com/ on 09/30/2021 Table 4. Physical Diagnosis for Presence of Vascular Disease: Sensitivity and Specificity*

Definition of Test Performance Study No. of Physical Finding Study Level† Patients Abnormal Finding Disease Sensitivity Specificity LR+ LR− Abnormal pedal pulses Christensen et al13 1 132 Both PT/DP pulse absent AAIϽ0.9 0.63 0.99 44.6 0.4 AAIϽ0.5 0.95 0.73 3.5 0.1 Stoffers et al15 1 2455 PT/DP pulse: absent/absent AAIϽ0.9 0.73 0.92 9.0 0.3 or absent/weak Boyko et al12 1 631 PT/DP pulse: absent/absent, AAIϽ0.5 0.65-0.80‡ 0.78-0.79 3.0-3.8 0.3-0.4 absent/weak, or weak/weak Femoral arterial bruit Criqui et al10 1 613 Femoral bruit present AAIϽ0.8 0.20 0.96 4.7 0.8 Stoffers et al15 1 2455 Femoral bruit present AAIϽ0.9 0.29 0.95 5.7 0.7 Venous filling time Boyko et al12 2§ 631 Ͼ20 s AAIϽ0.5 0.22-0.25 0.94-0.95 3.6-4.6 0.8 Cool skin Stoffers et al15 1 2455 Unilateral cooler skin AAIϽ0.9 0.10 0.98 5.8 0.9 Boyko et al12 1 631 Dorsal foot cooler to touch AAIϽ0.5 0.65-0.80 0.46-0.47 1.2-1.5 0.4-0.7 than ipsilateral calf Color abnormality Stoffers et al15 1 2455 Pale, red, or blue color AAIϽ0.9 0.35 0.87 2.8 0.7 Boyko et al12 1 631 Blue, purple color AAIϽ0.5 0.24-0.32 0.84-0.85 1.6-2.0 0.8-0.9 Capillary refill time Boyko et al12 1 631 Ͼ5 s AAIϽ0.5 0.25-0.28 0.84-0.85 1.6-1.9 0.8-0.9 Trophic changes Boyko et al12 1 631 Absent distal lower-extremity AAIϽ0.5 0.47-0.48 0.70-0.71 1.6 0.7 hair growth Atrophic skin AAIϽ0.5 0.43-0.50 0.70 1.4-1.6 0.7-0.8

*LR+ indicates positive likelihood ratio; LR−, negative likelihood ratio; PT, posterior tibial; DP, dorsalis pedis; and AAI, ankle-arm index. †See “Materials and Methods” section for definitions. ‡Data from Boyko et al12 were derived from examination of both limbs. Because the findings in the right leg were not independent of those in the left leg of the same patient, the data for each were presented separately, yielding 2 values for each finding. §In contrast to all the other physical examination maneuvers in the study by Boyko et al,12 the venous filling time was performed after the measurement of the AAI, which potentially affects the blinding of the comparison and lowers the grade for this finding to level 2.

that 2% to 13% of abnormal AAIs occur only after The presence of color abnormalities (pale, red, or stress.8,11 (Most of these patients, however, never blue color) were weak predictors of disease (LR+=1.6- underwent angiography, raising the question whether 2.8), but in both of the studies that investigated this they all represent true-positive findings.) Still, many finding,12,15 foot discoloration added nothing to the vascular surgeons regard pulses that disappear with clinical prediction after multivariate analysis. Table 4 exercise as a significant phenomenon that occurs rela- also shows that an abnormal capillary refill time, absent tively commonly (in 16 patients over a 2-year period distal lower-extremity hair growth, and atrophic skin in 1 surgical practice).47 These surgeons propose a are all unhelpful to the clinician. Another level 3 study variety of simple bedside maneuvers to detect this also found no difference in hair loss among the lower phenomenon, all carried to the point of claudication limbs of 40 patients with vascular disease and 40 con- (running in place, walking, toe-stands, or ankle flex- trol subjects.23 ing repeatedly against resistance).44,45,47 DISTRIBUTION OF VASCULAR DISEASE Other Findings One study, graded level 3 because it omitted explicit in- The presence of a femoral arterial bruit is a strong indi- struction on the physical examination maneuvers per- cator of disease (LR+=4.7-5.7). Because the test is formed, showed that vascular surgeons using tradi- insensitive (sensitivity, 0.20-0.29), however, the tional bedside methods (similar to those in Table 1) absence of a femoral bruit does not meaningfully accurately predicted the distribution of occlusions and change the probability of disease (LR−=0.7-0.8) (Table stenoses in 96% of 102 symptomatic patients (the pre- 4). Two other positive findings that are useful, although dictions preceded measurement of the AAI or angiogra- both are insensitive (sensitivity, 0.10-0.25), are an phy).22 In another study, the finding of a severely re- abnormally prolonged venous filling time (LR+=3.6- duced or absent femoral pulse in patients with suspected 4.6) (see Table 5 for definition) and a limb that is vascular disease confirmed the diagnosis of aortoiliac dis- cooler when compared with the contralateral limb ease (LR+=ϱ; specificity, 1.0; 95% confidence interval, (LR+=5.8) (Table 4). In a different study,12 the finding 0.93-1.0; Table 6).19 Iliac bruits (bruits above the in- of a foot cooler than the ipsilateral calf was unhelpful guinal crease) are more modest predictors of aortoiliac (LR+=1.2-1.5), but this is not surprising because the disease (LR+=2.2; Table 6), suggesting that either trivial cutaneous blood flow, the main determinant of skin stenoses in the aortoiliac segment can produce bruits or surface temperature, often diminishes toward the feet to that bruits from significant stenoses distant from the aor- conserve body heat, causing the toes of normal extremi- toiliac segment can radiate to the groin. Even so, the find- ties to be 2°C to 6°C cooler than the ipsilateral groin.48 ing of femoral or iliac bruits, when combined with other A unilaterally cold foot, however, is abnormal and usu- clinical features of aortoiliac disease (thigh claudica- ally indicates arterial disease.15 tion, abnormal femoral pulse), is more meaningful. Among

ARCH INTERN MED/ VOL 158, JUNE 22, 1998 1360

Downloaded From: https://jamanetwork.com/ on 09/30/2021 studies, mostly case series that sometimes included in- Table 5. Glossary of Terms frared thermography,48-51 have confirmed this finding. Pre- sumably, femoropopliteal occlusions induce collateral ves- The Buerger Test sels to develop just beneath the skin surface around the According to Buerger, the clinician should elevate the supine patient’s foot perpendicular to the horizontal plane (ie, 180°), which will knee, thus raising its temperature. Table 6 reveals the cause foot pallor in those with vascular disease. The clinician then “warm-knee sign” to be a predictor of superficial fem- estimates the “angle of circulatory sufficiency” by slowly lowering oral artery occlusions or stenoses ending at the adduc- the leg and recording the angle at which a reddish hue returns (eg, tor hiatus (LR+=2.9, LR−=0.4). Important questions about at the examining table surface, 90°; below the table surface, Ͻ90°; the warm-knee sign, however, are whether popliteal oc- dangled perpendicular to the table, 0°). In a variation of this test cited in this review,18 clinicians elevated the clusions are important to the finding (some investiga- patient’s leg 60° above the table surface for 2 minutes (ie, 150° tors stating it is a prerequisite while others stating it is according to Buerger’s measures) and then dangled the leg over unnecessary) and whether this sign actually influences the edge of the table (ie, 0°) for another 2 minutes. Patients clinical decisions (ie, how many of these patients have with a positive test result had foot pallor with elevation and the appearance of a dusky red flush spreading proximally obvious femoropopliteal disease from the classic com- from the toes in the dependent position. bination of calf claudication, normal femoral pulses, and Capillary Refill Test absent popliteal pulses). The clinician applies firm digital pressure to the plantar skin of the distal great toes for 5 seconds. Transient local pallor is considered INCREMENTAL VALUE OF normal, while a delay greater than 5 seconds before return of the usual skin color is regarded as delayed refill.12 COMBINATIONS OF SYMPTOMS AND SIGNS: Venous Filling Time MULTIVARIATE ANALYSIS After positioning the patient supine and identifying a prominent pedal vein, the clinician elevates the patient’s leg 45° above the table The goal of multivariate analysis is to identify diagnos- surface for 1 minute. The patient then sits up and dangles the legs tic factors that are redundant, to identify the minimal num- over the side of the examining table, and the clinician records the time in seconds until the previously identified vein rises above the ber of useful findings. Using multivariate analysis, 2 stud- 12,15 level of the skin surface.12 ies have identified which clinical findings are Visual Flush Method independent predictors of peripheral arterial disease A maneuver similar to the Allen test, this method purportedly (Table 7). Although the studies enrolled different popu- estimates the ankle blood pressure when Doppler equipment is lations (100% vs 10.2% were patients with diabetes), in- unavailable. After placing the blood pressure cuff around the patient’s ankle and elevating the patient’s foot, the clinician inflates vestigated slightly different findings (Table 7), and de- the cuff and returns the limb to the horizontal position. As the fined disease differently (AAIϽ0.5 vs AAIϽ0.9), they both pressure in the cuff is slowly decreased, 5 mm Hg every 10 dealt with large numbers of elderly individuals of whom seconds, the first appearance of pink flush on the sole of the foot only 9% to 16% had claudication and 8% to 9% had pe- indicates the systolic pressure. In a study of 156 lower limbs with ripheral arterial disease. Importantly, both studies showed ankle-to-arm systolic pressure indexes (AAIs) between 0.3 and 1.4, 85% of the AAIs by the visual flush method were within 0.1 of the that certain historical items were important predictors conventional AAI.6 and that physical examination added significantly to any predictive model based on history alone. Independent predictors from the physical examination were abnormal pedal pulses, a prolonged venous patients with 2 of these 3 clinical features, 83% had aor- filling time, a unilaterally cool extremity, and a femoral toiliac disease in one study19 and in those with all 3 fea- artery bruit. The findings of foot discoloration, pro- tures, 100% had aortoiliac disease. In another study of longed capillary refill time, a hairless limb, and atrophic 50 patients with suspected vascular disease and intact pop- skin were all unhelpful diagnostically. liteal pulses (patients unlikely to have total vascular oc- Both models included mathematical formulas clusions), the finding of any limb bruit was a predictor (Table 8) that allow precise predictions of the prob- of significant stenoses on angiography (LR+=3.2) (Table ability of disease. Boyko et al12 validated the results of 6), which has important therapeutic implications be- their regression model using bootstrap methodology. cause these patients may be candidates for angioplasty Stoffers et al15 did not validate their regression model, (Table 1).14 but it is unlikely that their final model overfit the data A negative Buerger test result (Table 5) excluded dis- because they considered only 19 potential predictors of ease distal to the adductor canal in a study of 55 limbs low AAI in 2455 patients. According to the model by with symptomatic vascular disease (LR−=0; sensitivity, Stoffers et al, which yields a range of probabilities of 0.4% 1.0; 95% confidence interval, 0.72-1.0; Table 6).18 Fur- to 98%, the probability of an AAI less than 0.9 in a 75- thermore, a positive Buerger test result indicated more year-old man who smokes and has claudication, abnor- severe and extensive disease. Patients with a positive mal pedal pulses, a femoral artery bruit, and a unilater- test result had lower AAIs (mean ± SD, 0.37±0.29 vs ally cool leg is 93%. In such a patient, further diagnostic 0.62±0.23), more rest pain (60% vs 8%), and more gan- testing with a Doppler flowmeter is unnecessary. Using grene (23% vs 0%).18 the model of Boyko et al, which yields a range of prob- One level 3 study reported the intriguing finding that abilities of 0.02% to 93%, the clinician would determine about one half of patients with arterial disease confined that the probability of an AAI less than 0.5 in a 80-year- to the femoropopliteal segment had unilaterally warm old with diabetes, no history of peripheral vascular dis- knees in the same limb, whereas no patient with disease ease, normal pulses, and a venous filling time of 5 sec- in the aortoiliac segment had warm knees.25 Subsequent onds is only 3%. Further testing is again unnecessary. Only

ARCH INTERN MED/ VOL 158, JUNE 22, 1998 1361

Downloaded From: https://jamanetwork.com/ on 09/30/2021 Table 6. Physical Diagnosis and Distribution of Vascular Disease: Sensitivity and Specificity*

Definition of Test Performance Study No. of Physical Findings Study Level† Patients Abnormal Finding Disease Sensitivity Specificity LR+ LR− Abnormal femoral Johnston et al19 2 78 Femoral pulse severely Ͼ10–mm Hg 0.38 1.0 ϱ‡ 0.6 pulse vs significant reduced or absent pressure gradient aortoiliac disease across aortoiliac segment Iliac bruit vs Johnston et al19 2 78 Iliac bruit present§ Ͼ10–mm Hg 0.28 0.87 2.2 0.8 aortoiliac disease pressure gradient across aortoiliac segment Limb bruit vs Nicholson et al14 1 50 Iliac, femoral, or Arterial stenosis 0.8 0.75 3.2 0.3 presence of popliteal bruit§ Ͼ25% on arterial stenosis (popliteal pulse angiogram present, see text) Buerger test Insall et al18 2 55 limbs See Table 5 Disease distal to 1.0 0.54 2.2 0 adductor canal on angiography Warm knees O’Brien et al20 2 30 Knee of symptomatic Distal end of 0.73 0.75 2.9 0.4 limb warmer than angiographic contralateral knee occlusion ends at adductor hiatus

*LR+ indicates positive likelihood ratio; LR−, negative likelihood ratio. †See the “Materials and Methods” section for definitions. ‡Although the finding of an abnormal femoral pulse is undoubtedly helpful, an LR+ value of “infinity” may considerably overestimate the true value due to a small sample size. §The inguinal crease distinguishes the iliac bruit (heard above the crease) from the femoral bruit (heard below).

Table 7. Incremental Value of Symptoms and Signs: Multivariate Analysis*

Boyko et al12 Stoffers et al15 No. of patients 631 2455 Type of study Cross-sectional study of diabetic veteran Cross-sectional random sampling of enrollees in a general medicine clinic 18 general practice centers in the Netherlands Mean age, y 63.4 59.3 % with diabetes 100 10.2 % with claudication 15.7 9.4 % with disease, definition of disease 7.6, AAIϽ0.5 9.2, AAIϽ0.9 Clinical findings that independently predict disease† Symptoms, risk factors Older age (Ͼ65 y), history of PVD Intermittent claudication, male gender, older age (Ͼ60 y), history of ischemic heart disease, history of diabetes, history of smoking Physical signs Abnormal pedal pulses, venous filling Abnormal pedal pulses, femoral artery bruit, time Ͼ20 s unilaterally cooler limb, hypertension

*AAI indicates ankle-arm index; PVD, peripheral vascular disease. †Findings found not to be independent predictors: Boyko et al12—duration of diabetes, current smoking status, other symptoms (lower extremity ulcers or amputation, cold or blue feet, claudication, prior leg bypass surgery) and signs (absent leg hair; atrophic, blue, or cool skin; or capillary refill time Ͼ5 seconds); Stoffers et al15—history of hypercholesterolemia, history of cerebral vascular disease, amount of physical exercise, other symptoms (cold feet, erectile dysfunction), and signs (abnormal femoral pulse, absent leg hair, foot discoloration, or foot ulcers).

individuals with intermediate probabilities require fur- COMMENT ther testing—for example, the 29% probability of an AAI less than 0.9 in a 64-year-old woman who smokes and Clinicians should suspect chronic arterial occlusion in has claudication, weak pedal pulses, no unilateral limb patients who present with claudication, gangrene, rest coolness, and no femoral artery bruit.15 pain that is relieved by dangling the limb, or ulcers with Although the mathematical formulas in Table 8 are a necrotic base involving the distal foot. Lumbar spinal complicated and too burdensome for general clinical use stenosis also may cause claudication, but the limb dis- in individual patients, clinicians can generate simple-to- comfort in these patients characteristically worsens with use figures, examples of which appear in one of the origi- extension of the spine (eg, standing), improves with flex- nal studies.12 ion of the spine (eg, sitting), and varies in severity from

ARCH INTERN MED/ VOL 158, JUNE 22, 1998 1362

Downloaded From: https://jamanetwork.com/ on 09/30/2021 some of Buerger’s other prodromal findings, including Table 8. Probability of Peripheral Arterial hairless legs, atrophic skin, and foot discoloration. Disease, According to 2 Models Unresolved in this review are the clinical utility of the warm-knee sign, the diagnostic value of the abnor- 12 Model by Boyko et al mal popliteal pulse (as a sign of femoropopliteal disease The probability of an ankle-to-arm systolic pressure index below 0.5 is estimated to be (odds)/(1 + odds), where odds = ex and when the femoral pulse is present), the importance of ex- x = −14.6077 + (0.0935) (age) + (0.0626) (venous filling ercising patients with suspected disease and palpable pedal time) + (1.4922) (pulse code) + (1.9007) (PVD history code). Age pulses, and the merit of the “visual flush method,” a po- and venous filling time are coded as continuous variables, with age tentially important maneuver (Table 5) that allows cal- entered as years and venous filling time as seconds. The pulse code is: pedal pulses present = 1, pedal pulses diminished or absent = 2. culation of the AAI without a Doppler flowmeter. Even The peripheral vascular disease (PVD) history code is: no prior without answers to these questions, however, the physi- history = 1, prior history = 2. cal examination remains a powerful diagnostic tool for Model by Stoffers et al15 clinicians who have patients with suspected chronic lower- The probability of an AAI below 0.9 is estimated to be 1/ (1 + e−g), extremity arterial disease. where g = (−5.623 + ␤1+␤2+...+␤i). The relation between coefficients (␤i) and OR is given by the formula ␤i = ln (ORi), where ln is the natural logarithm. The OR for various findings is: male sex, Accepted for publication August 29, 1997. 1.5; age greater than 60 y, 2.2; intermittent claudication, 3.5; The authors thank Thomas Hatsukami, MD, who unilateral lower skin temperature, 2.5; pedal pulses reviewed the manuscript and provided many helpful abnormal (absent/absent or absent/weak), 16.4; pedal pulses doubtful (weak/weak or strong/absent), 7.0; femoral artery bruit, comments. 3.5; ischemic heart disease, 1.7; diabetes mellitus, 1.6; smoking, Reprints: Steven R. McGee, MD, Veterans Affairs Puget 2.1; hypertension, 1.5. Sound Health Care System (Mailstop 111), 1660 S Colum- bian Way, Seattle, WA 98108.

REFERENCES day to day much more than the discomfort of vascular claudication.52,53 1. Buerger L. The Circulatory Disturbances of the Extremities: Including Gangrene, Since the AAI is the diagnostic criterion standard for Vasomotor and Trophic Disorders. Philadelphia, Pa: WB Saunders Co; 1924. peripheral arterial disease, clinicians who have Doppler 2. Pickering GW. On the clinical recognition of structural disease of the peripheral equipment should use flowmeters, not physical exami- vessels. BMJ. 1933;2:1106-1110. nation, to make the diagnosis in most of their patients. 3. Lewis T. Vascular Disorders of the Limbs, Described for Practitioners and Stu- Primary care physicians, however, do not always have dents. New York, NY: Macmillan Co; 1936. 4. Mannick JA. Evaluation of chronic lower-extremity ischemia. N Engl J Med. 1983; easy access to Doppler equipment, and this review has 309:841-843. confirmed the value of the traditional approach to diag- 5. Pentecost MJ, Criqui MH, Dorros G,et al. Guidelines for peripheral percutane- nosis, combining information from both the patient in- ous transluminal angioplasty of the abdominal aorta and lower extremity ves- terview and physical examination. The important physi- sels. Circulation. 1994;89:511-531. cal signs—presence or absence of pedal pulses, a 6. Carter SA. Indirect systolic pressures and pulse waves in arterial occlusive disease of the lower extremities. Circulation. 1968;37:624-637. unilaterally cool extremity, prolonged venous filling time, 7. Carter SA. Clinical measurement of systolic pressures in limbs with arterial oc- and femoral artery bruit—help clinicians determine the clusive disease. JAMA. 1969;207:1869-1874. estimated probability of disease in their patients, which, 8. Ouriel K, McDonnell AE, Metz CE, Zarins CK. A critical evaluation of stress test- in turn, appropriately influences decisions about the need ing in the diagnosis of peripheral vascular disease. Surgery. 1982;91:686-693. 9. Hiatt WR, Hoag S, Hamman RF. Effect of diagnostic criteria on the prevalence of for further testing. peripheral arterial disease: the San Luis Valley Diabetes Study. Circulation. 1995; Even clinicians who have a Doppler flowmeter, how- 91:1472-1479. ever, still benefit from examining the patient carefully. 10. Criqui MH, Fronek A, Klauber MR, Barrett-Connor E, Gabriel S. The sensitivity, Some patients with diabetes have calcified distal arteries specificity, and predictive value of traditional clinical evaluation of peripheral ar- that are noncompressible, causing the AAI to be falsely terial disease: results from noninvasive testing in a defined population. Circula- tion. 1985;71:516-522. high (ie, Ͼ0.9) even though the clinical evaluation still 11. Hiatt WR, Marshall JA, Baxter J,et al. Diagnostic methods for peripheral arte- 4 suggests disease. Furthermore, in patients whose AAIs rial disease in the San Luis Valley Diabetes Study. J Clin Epidemiol. 1990;43: are less than 0.9, the presence or absence of thigh clau- 597-606. dication, femoral pulse abnormalities, lower-extremity 12. Boyko EJ, Ahroni JH, Davignon D, Stensel V, Prigeon RL, Smith DG. Diagnostic bruits, warm knees, and a positive Buerger test all help utility of the history and physical examination for peripheral vascular disease among patients with diabetes mellitus. J Clin Epidemiol. 1997;50:659-668. determine the extent of the patient’s disease, its distri- 13. Christensen JH, Freundlich M, Jacobsen BA, Falstie-Jensen N. Clinical rel- bution, and, if the patient is a candidate for surgery, the evance of pedal pulse palpation in patients suspected of peripheral arterial in- necessary procedure. sufficiency. J Intern Med. 1989;226:95-99. Other findings from the examination, although tra- 14. Nicholson ML, Byrne RL, Steele GA, Callum KG. Predictive value of bruits and Doppler pressure measurements in detecting lower limb arterial stenosis. Eur J ditionally thought to be useful, do not help the diagnos- Vasc Surg. 1993;7:59-61. 3 2 tic process. Lewis and Pickering, writing soon after pub- 15. Stoffers HEJH, Kester ADM, Saiser V, Rindens PELM, Knottnerus JA. Diagnos- lication of Buerger’s book,1 showed that the capillary refill tic value of signs and symptoms associated with peripheral arterial occlusive dis- test was unreliable because the capillary blood ex- ease seen in general practice: a multivariable approach. Med Decis Making. 1997; pressed by firm pressure from the examiner’s hand could 17:61-70. 16. Kazmers A, Koski MF, Groehn H,et al. Assessment of noninvasive lower extrem- return from adjacent veins, allowing prompt refill even ity arterial testing versus pulse exam. Am Surg. 1996;62:315-319. in a limb rendered completely ischemic experimentally. 17. Carter SA. Arterial ausculatation in peripheral vascular disease. JAMA. 1981; Our review also finds capillary refill unhelpful, along with 246:1682-1686.

ARCH INTERN MED/ VOL 158, JUNE 22, 1998 1363

Downloaded From: https://jamanetwork.com/ on 09/30/2021 18. Insall RL, Davies RJ, Prout WG. Significance of Buerger’s test in the assess- thousand individuals without symptoms of circulatory affections of the extremi- ment of lower limb ischaemia. J R Soc Med. 1989;82:729-731. ties. N Engl J Med. 1933;208:438-440. 19. Johnston KW, Demorais D, Colapinto RF. Difficulty in assessing the severity of 36. Silverman JJ. The incidence of palpable dorsalis pedis and posterior tibial pul- aorto-iliac disease by clinical and arteriographic methods. Angiology. 1981;32: sations in soldiers. Am Heart J. 1946;32:82-87. 609-614. 37. Stephens GL. Palpable dorsalis pedis and posterior tibial pulses. Arch Surg. 1962; 20. O’Brien DP, Walsh TN, Given HF. The warm knee sign: an evaluation. Eur J Vasc 84:662-664. Surg. 1990;4:531-534. 38. Barnhorst DA, Barner HB. Prevalence of congenitally absent pedal pulses. N Engl 21. Marinelli MR, Beach KW, Glass MJ, Primozich JF, Strandness DE. Noninvasive J Med. 1968;278:264-265. testing vs clinical evaluation of arterial disease. JAMA. 1979;241:2031-2034. 39. Nuzzaci G, Giuliano G, Righi D, Baroncelli T, Lotti A, Marinoni M. A study of 22. Baker WH, String T, Hayes AC. Diagnosis of peripheral occlusive disease: com- the semeiological reliability of dorsalis pedis artery and posterior tibial artery parison of clinical evaluation and noninvasive laboratory. Arch Surg. 1978;113: in the diagnosis of lower limb arterial occlusive disease. Angiology. 1984;35: 1308-1310. 767-772. 23. Parfrey N, Ryan JF, Shanahan L, Brady MP. Hairless lower limbs and occlusive 40. Robertson GSM, Ristic CD, Bullen BR. The incidence of congenitally absent foot arterial disease. Lancet. 1976;1:276. pulses. Ann R Coll Surg Engl. 1980;72:99-100. 24. Sobinsky KR, Borozan PG, Gray B, Schuler JJ, Flanigan DP. Is femoral pulse pal- 41. Reich RS. The pulses of the foot: their value in the diagnosis of peripheral cir- pation accurate in assessing the hemodynamic significance of aortoiliac occlu- culatory disease. Ann Surg. 1934;99:613-622. sive disease? Am J Surg. 1984;148:214-217. 42. Marx HJ, Yu PN. Clinical examination of the arterial pulse. Prog Cardiovasc Dis. 25. Zweifler AJ. Significance of knee skin temperature in ischemic legs. Arch Intern 1967;10:207-235. Med. 1965;115:151-154. 43. Milnor WR. Pulsatile blood flow. N Engl J Med. 1972;287:27-34. 26. Sackett DL, Haynes RB, Tugwell P. Clinical Epidemiology: A Basic Science for 44. DeWeese JA. Pedal pulses disappearing with exercise: a test for intermittent clau- Clinical Medicine. Boston, Mass: Little Brown & Co; 1985. dication. N Engl J Med. 1960;262:1214-1217. 27. Boyko EJ. Ruling out or ruling in disease with the most sensitive or specific di- 45. Carter SA. Response of ankle systolic pressure to leg exercise in mild or ques- agnostic test: short cut or wrong turn? Med Decis Making. 1994;14:175-179. tionable arterial disease. N Engl J Med. 1972;287:578-582. 28. Lawson IR, Ingman SR, Masih Y, Freeman B. Reliability of palpation of pedal 46. Laing SP, Greenhalgh RM. Standard exercise test to assess peripheral arterial pulses as ascertained by the kappa statistic. J Am Geriatr Soc. 1980;28: disease. BMJ. 1980;280:13-16. 300-303. 47. Barner HB, Kaiser GC, Willman VL, Hanlon CR. Intermittent claudication with pedal 29. Myers KA, Scott DF, Devine TJ, Johnston AH, Denton MJ, Gilfillan IS. Palpation pulses. JAMA. 1968;204:958-962. of the femoral and popliteal pulses: a study of the accuracy as assessed by agree- 48. Soulen RL, Lapayowker MS, Tyson RR, Korangy AA. Angiography, ultrasound, ment between multiple observers. Eur J Vasc Surg. 1987;1:245-249. and thermography in the study of peripheral vascular disease. Radiology. 1972; 30. Brearly S, Shearman CP, Simms MH. Peripheral pulse palpation: an unreliable 105:115-119. physical sign. Ann R Coll Surg Engl. 1992;74:169-171. 49. Evans AL, James WB, Forrest H. Thermography in lower limb arterial disease. 31. Ludbrook J, Clarke AM, McKenzie JK. Significance of absent ankle pulse. BMJ. Clin Radiol. 1976;27:383-388. 1962;5294:1724-1726. 50. Gaylis H. Warm knees and cold feet: a new sign of arterial occlusion. Lancet. 32. Magee TR, Stanley PRW, Al Mufti R, Simpson L, Campbell WB. Should we pal- 1966;1:792-793. pate foot pulses? Ann R Coll Surg Engl. 1992;74:166-168. 51. Spence VA, Kester RC, Howie G, Walker WF. Current status of thermography in 33. Meade TW, Gardner MJ, Cannon P, Richardson PC. Observer variability in re- peripheral vascular disease. J Cardiovasc Surg. 1975;16:572-579. cording the peripheral pulses. Br Heart J. 1968;30:661-665. 52. Dong GX, Porter RW. Walking and cycling tests in neurogenic and intermittent 34. Sackett DL. A primer on the precision and accuracy of the clinical examination. claudication. Spine. 1989;14:965-969. JAMA. 1992;267:2638-2644. 53. Goodreau JJ, Creasy JK, Flanigan P, et al. Rational approach to the differentia- 35. Morrison H. A study of the dorsalis pedis and posterior tibial pulses in one tion of vascular and neurogenic claudication. Surgery. 1978;84:749-757.

ARCH INTERN MED/ VOL 158, JUNE 22, 1998 1364

Downloaded From: https://jamanetwork.com/ on 09/30/2021