- Email: [email protected]
Does air plethysmography correlate with duplex scanning in patients with chronic venous insufficiency?
Does air plethysmography correlate with duplex scanning in patients with chronic venous insufficiency? Paul S. van B e m m e l e n , M D , P h D , M a r k A. Mattos, M D , K i m J. H o d g s o n , M D , L y n n e D. Barkmeier, M D , D o n E. Ramsey, M D , William E. F a u g h t , M D , and David S. S u m n e r , M D , Springfield, Ill.
Purpose: Duplex ultrasonography with distal cuff deflation was used to determine the presence and size of incompetent veins and compare the results with those of air plethysmography in patients with chronic venous insufficiency. Methods: Thirty-two legs underwent a detailed study with both modalities. Sixteen legs had venous ulceration, six had stasis dermatitis, and ten had symptomatic varicose veins without skin changes. Results:Although the venous filling index (VFI) in limbs with ulcers (5.4 + 3.8 ml/sec) and dermatitis (7.7 +- 4.6 ml/sec) was significantly higher (p < 0.05) than it was in limbs with varicose veins (2.6 -+ 1.7 ml/sec), there was a large amount of overlap. Only 13% of ulcerated legs had VFI greater than 10 ml/sec. Sixty-three percent of legs with ulcers, 33% of legs with dermatitis, and 90% of legs with varicose veins had VFIs less than 5 ml/sec. Mean ejection fractions (EFs) in the three groups were similar, ranging from 45% to 52%. Combining VFI and EF did not lessen the overlap between groups. Forty-one percent of limbs with ulcers or dermatitis had air plethysmography parameters in the normal or intermediate area (VFI < 5 ml/sec; EF > 40%), which in previous studies corresponded to an incidence of ulceration of only 2%. VFI had a significant but weak correlation (r = 0.39) with the diameter of incompetent veins at the knee and a somewhat stronger relationship (r = 0.55) with the diameter of lower leg veins. Total venous volume correlated moderately well with calf vein diameter (r = 0.75). The clinical status of the leg did not correlate with the diameters of incompetent veins at the knee or calf levels. All limbs with an obstructed outflow had EFs less than 60% and ulcers or dermatitis. Conclusions: We conclude that plethysmographic measurements of functional venous parameters (VFI, EF) do not discriminate well between limbs with uncomplicated varicose veins and limbs with ulcers or stasis dermatitis and that VFI correlates poorly with the presence of incompetent veins and their diameters. Both duplex scanning and plethysmography seem to be necessary for a complete evaluation of limbs with chronic venous insufficiency. (J VASC SURG 1993;18:796-807.)
Semiquantitative methods for evaluating venous function in patients with chronic venous insufficiency include venous pressure measurements, 1 photoplethysmography, 2 and air plethysmography, a Recently,
From the Department of Surgery, Section of Peripheral Vascular Surgery, Southern Illinois University School of Medicine, Springfield. Presented at the Fifth Annual Meeting of the American Venous Forum, Orlando, Fla., Feb. 24-26, 1993. Reprint requests: David S. Sumner, MD, Southern Illinois UniversitySchoolof MedicineDepartment of Surgery,PO Box 19230, Springfield,IL 62794-9230. Copyright © 1993 by The Society for Vascular Surgery and International Societyfor CardiovascularSurgery,North American Chapter. 0741-5214/93/$1.00 + .10 24]6/50279 796
the development o f an air plethysmograph (APG) 4 that surrounds the entire calf has renewed interest in assessing the function o f venous valves and the calf venous pump. Investigators at St. Mary's Hospital in L o n d o n have established normal values for the rate o f venous filling that accompanies the assumption o f an uptight position and for the reduction in calf volume that occurs with tiptoe exercise? -6 The venous filling index (VFI) is a measure o f reflux flow through incompetent venous valves, and the ejection fraction (EF) relates to the function o f the calf muscle pump. A decrease in the VFI has been described after varicose vein surgery 7 and with the use o f elastic support stockings. 8,9 Another group, reporting on their results with air plethysmography, noted a clear
JOURNAL OF VASCULAR SURGERY
Volume18, Number 5
distinction between the VFI in normal volunteers and in patients with severe chronic venous insufficiency.1° This group also observed a correlation between end-exercise venous volume (VV) and ambulatory venous pressure? ~ Other investigators, however, have not found air plethysmographic parameters to be helpful in differentiating between simple varicose veins and chronic venous insufficiency.~1 Ultrasonic duplex or color-flow imaging performed at multiple levels of the leg provides a precise method for locating incompetent veins and for quantitating reflux flow, ~3 but in patients with multiple incompetent veins, calculating the volume of reflux flow by this method is cumbersome? 4 The goals of our study were (1) to ascertain whether air plethysmographic parameters distinguish between uncomplicated varicose veins and limbs with severe skin changes; (2) to correlate the results of air plethysmography with anatomic measurements made with duplex scanning; and (3) to determine whether duplex scanning can be replaced by the simpler, less expensive APG method in certain patient groups. Furthermore, we attempted to determine whether the diameter of incompetent veins in patients with severe manifestations of chronic venous insufficiency (ulcers or stasis dermatitis) differs from that of patients with simple varicose veins. MATERIAL A N D M E T H O D S
Thirty-two legs of 28 patients with symptoms of chronic venous insuffÉciency were studied by both segmental color-flow duplex scanning and air plethysmography. Ulcers were present in 16 legs of 16 patients (mean age 56.8 + 15.0 years). In six legs there was severe swelling with wet dermatitis (four patients, mean age 50.3 + 6.0 years). Varicose veins were present in 10 legs of eight patients (mean age 47.1 _+ 15.6 years), who complained of fatigue with prolonged standing. According to the criteria estabfished by the Society for Vascular Surgery and the North American Chapter of the International Society for Cardiovascular Surgery, all limbs with dermatitis and ulcers were class 3 severity, and limbs with varicose veins were class 175 Air plethysmography. Methods for measuring changes in calf venous volume with air plethysmography have been described in detail by other authors, s The leg is enclosed from ankle to knee in a 14-inch polyurethane cuffthat acts as the sensing device (ACI Medical, Sun Valley, Calif.). Changes in air pressure are amplified and used to drive a strip chart recorder.
van Bemmelen et al.
797
In this study a standard protocol was used, which began with the subject resting in the supine position with the leg elevated on a foam block for 5 minutes. Pressure in the cuff was adjusted to 6 mm Hg. This is the "bias" pressure, which serves to keep the cuffin Contact with the skin. Calibration was performed by injecting 100 rnl of air into the system. The zero VV of the leg was determined by passively elevating the leg 45 °. After a stable baseline had been reached, the subject was instructed to assume a standing position without putting any weight on the leg being studied. In a normal situation, a slow increase in leg size is noted because of filling of veins from arterial inflow. The increase in calf volume in milliliters is called the VV, and the time in seconds required to reach a stable plateau is called the venous filling time. VFIs in ml/sec were calculated by dividing the VV at 90% refilling by the venous filling time required to achieve 90% refilling. The subject was then asked to perform a single tiptoe exercise, which was subsequently repeated. The resulting decrease in leg volume, divided by the maximal VV, is termed the EF. After this, 10 additional tiptoe maneuvers were performed to determine the residual venous volume; and afterwards the leg was elevated again to disclose any shift in the zero volume. Residual volume fraction (RVF) was calculated by dividing the residual venous volume by the VV. Normal legs have a VFI less than 2 ml/sec and an EF greater than 60%. 4,s VFIs greater than 5 ml/sec, EFs less than 40%, and RVFs greater than 40% are considered to be abnormal. 16 An increase in VFI or a decrease in EF are reported to be associated with an increased likelihood of venous ulceration.4,s,l° RVFs correlate well with ambulatory venous pressure, the gold standard for evaluating calf muscle pump function, u The incidence of ulceration rises steadily as the RVF increases. 6 Duplex scanning. The Quantum 2000 colorflow scanner (Siemens-Quantum, Issaquah, Wash.) was used for all studies. Evaluation of valve competence was performed with the distal cuff deflation maneuver? 3'17 This method allows evaluation of venous valvular function at distal levels of the leg, sites that cannot be tested with a Valsalva maneuver. 18 Studies were conducted with the patient standing holding onto an orthopedic frame, with his or her weight being supported by the contralateral leg. A Hokanson (Hokanson, Inc., D.E., Bellevue, Wash.) cuff inflator was used to ensure rapid (< 0.3 seconds) cuff inflation and deflation. Venous valvular competence was assessed in both the deep and superficial systems at the midthigh, knee, calf, and
798
v a n Bemraelen et al.
ankle levels. At each level, cuffs were placed around the limb several centimeters below the site of duplex interrogation. To exceed hydrostatic pressures, cuffs were inflated to 80 mm Fig, 100 mm Hg, and 120 mm H g at the thigh, calf, and foot levels, respectively. Color-flow duplex scanning was used to identify incompetent veins and to record the direction and duration of flow during cuff deflation. Reflux flow persisting more than 0.5 seconds identifies venous incompetence. 13,19Although peak reflux flow velocities at the popliteal level were recorded, no attempt was made to calculate total reflux flow rates. The multiplicity of veins in this region and the errors inherent in calculating total flow make such measurements difficult and unreliable. Duplex scanning was used to measure the diameter of incompetent veins passing the knee. These veins were designated "reflux conduit" veins. The main pathways for venous reflux at the knee are the greater saphenous vein and the popliteal vein. An additional search was made to exclude other incompetent veins, such as anterolateral branches of the greater saphenous vein or lesser saphenous veins with a high termination. The diameters of incompetent posterior tibial and peroneal veins in the calf lower leg were also measured; these veins were termed "reflux reservoir" veins. Reservoir veins accommodate reflux flow once it reaches the lower leg. It is the increase in diameter of these veins that is responsible for the increase in calf volume measured with the APG. If incompetent veins were absent, a diameter of 0 mm was assigned. Outflow function. Duplex scanning was used to evaluate the superficial femoral vein as the main deep venous outflow conduit. In addition to visualizing residual thrombus in the superficial femoral vein, venous obstruction was identified by measuring outflow velocity in the proximal femoral vein during standardized augmentation of venous flow produced by rapid ( < 0 . 3 seconds) inflation of a 24 cm cuff placed around the calf. Legs were classified as normal when outflow velocities were in the 143 + 58 cm/sec range, z° Outflow velocities less than 85 cm/sec were considered to be consistent with incomplete recanalization of the deep system. Statistics. The Mann-Whitney U test (two-tailed) was used to compare grouped data, and Fisher's exact test was used to compare proportional data. A p value less than 0.05 was considered significant. Correlations between measurements were calculated by the Pearson's correlation coefficient method.
JOURNAL OF VASCULAR SURGERY November 1993
RESULTS Plethysmographic results are compared with clinical presentations in Tables I to III. In all clinical groups the VFI varied widely, but mean values obtained in limbs with ulcers and dermatitis were significantly higher than those found in limbs with simple varicose veins. Also, the proportion of legs with abnormal VFIs ( ___5 ml/sec) was higher in legs with dermatitis than it was in legs with varicose veins (Table I). There was little difference between the mean EFs of the three groups (Table II). RVFs were higher in limbs with ulcers than in limbs with dermatitis or varicose veins, and the proportion of limbs with abnormally high RVFs was significantly greater in the group with ulcers than it was in the group with varicose veins (Table III). In limbs with dermatitis, the increase in calf volume (172 + 92 ml) that accompanied assumption of a standing position (VV) was significantly greater (p = 0.04) than it was in limbs with varicose veins (101 _+ 35 ml). The VV in ulcerated legs was 120 -+ 44 ml. In Fig. 1, the results from each limb are plotted according to EF and VFI. The graph is divided into six compartments (I to VI) by a horizontal line corresponding to an EF of 40% and by vertical lines corresponding to VFIs of 5 and 10 ml/sec. The compartments of the matrix are numbered to correspond with increasing severity of hemodynamic compromise (with I as near normal and VI as the worst). 16 In our study 13 (59%) of 22 limbs with ulcers or dermatitis had values falling in compartments II to VI. Only three (30%) limbs with varicose veins had values in these compartments. This difference, however, failed to reach statistical significance (p = 0.25). Sixteen normal volunteers underwent APG testing before the start of this study. All had VFIs less than 2 ml/sec and EFs greater than 60%. Thus all normal limbs in our laboratory had APG values falling within the rectangle labeled "N" in Fig. 1. None of the venous segments in these subjects had reflux durations exceeding 0.5 seconds. Location and extent o f venous incompetence. Because four levels of the superficial and deep veins were studied, it was possible to record a maximum of eight incompetent segments in each limb. As shown in Table IV, there was no appreciable difference in the number of incompetent superficial venous segments in the three clinical groups. Limbs with ulcers, however, had more incompetent deepvenous segments than limbs with varicose veins (p = 0.008) or limbs with dermatitis (p = 0.06). Both the ulcer and dermatitis groups tended to have more total seg-
JOURNAL OF VASCULAR SURGERY Volume 18, Number 5
v a n B e m m e l e n et al.
799
Table I. VFI versus clinical presentation Abnormal ( >-5 ml/sec)
VFI (ml/sec)
Ulcer Dermatitis Varicose veins
n
Mean +- SD
Range
n
%
16 6 10
5.4 -+ 3.8 7.7 -+ 4.6 2.6 + 1.7"
2.0-16 1.4-13 1.2-7
6 4 1
38 67 10t
*Varicose veins versus dermatitis, p = 0.034; varicose veins versus ulcer, p = 0.003. tVaricose veins versus dermatitis, p = 0.036.
Table II. EF versus clinical presentation Abnormal (<40%)
EF (%)
Ulcer Dermatitis Varicose veins
n
Mean 4- SD
Range
n
%
16 6 10
45 -+ 22 49 -+ 20 52 _+ 16
14-92 22-72 25-76
8 2 3
50 33 30
N o statistically significant comparisons.
Table III. RVF versus clinical presentation Abnormal (>40%)
R V F (%)
Ulcer Dermatitis Varicose veins
n
Mean + SD
Range
n
%
16 6 10
60 -+ 23 44 _+ 17 43 -+ 21 ~
12-87 22-60 15-76
14 3 4
88 50 40j-
*Varicose veins versus ulcer, p = 0.067. j-Varicose veins versus ulcer, p = 0.026.
ments involved (superficial and deep) than the varicose vein group did (p = 0.008 and p = 0.08, respectively). Venous diameter. In all cases, duplex scanning identified incompetence of either the conduit or reservoir veins, as indicated by reflux persisting for greater than 0.5 sec after deflation of the calf or foot cuff. Incompetence was present at both levels in 23 (72%) of the 32 limbs. Diameters of both the reflux conduit veins and the reflux reservoir veins varied widely, and measurements for the three clinical groups overlapped over a wide range (Figs. 2 and 3). The mean diameters of the conduit veins in the ulcer group (6.3 - 3.1 mm) and in the dermatitis group (8.3--+ 4.8) did not differ significantly from the mean diameter of the varicose vein group (4.6 _+ 2.9 mm) (p = 0.18 and p = 0.06, respectively). Although the mean diameter of reservoir veins in the dermatitis group (7.8 +__5.2 mm) was twice that of reservoir veins in the varicose vein group (3.2 _+ 3.2
mm), this difference failed to reach statistical significance (p = 0.09). The difference was due largely to four veins with "zero" diameters in the latter group. The mean diameter of reservoir veins in the ulcer group was 4.9 _+ 2.1 mm (p = 0.16, versus varicose veins). Venous diameter versus air plethysmographic findings. VV was significantly correlated (r = 0.75, ,'2 = 0.56;p < 0.0001) with the diameter of incompetent calf veins (reservoir veins) as measured by duplex scanning (Fig. 4). A weak (r = 0.39, r e = 0.15) but statistically significant correlation (p = 0.03) was noted between the VFI and the diameter of incompetent conduit veins at knee level (Fig. 5). VFI also correlated (r = 0.55, re = 0.30; p = 0.001) with the diameter of incompetent reservoir veins in the lower leg (Fig. 6). VFIs greater than 5 ml/sec were not observed in limbs with incompetent calf veins less than 4 mm in diameter. On the other hand, when the diameters of incompetent
800
JOURNAL OF VASCULAR SURGERY November 1993
van Bemmelen et al.
I
IIl
20
• ulcer • dermatSIs o varicose veins
Vl
A•
&t)
~, o
0
4O
I
EF (%)
d • 0
IV
II
A
60--
80
1O0
,
O
•
I
0
5
0
15
20
VFI ( m l / s e e ) Fig. 1. EF plotted versus VFI. Triangles indicate limbs with ulcers; closedcircles indicate limbs with dermatitis; and open circles indicate limbs with varicose veins. Compartments are numbered (I to VI) according to increasing severity of physiologic abnormality. Rectangle (N) in lower left corner indicates reported range of normal values. Table IV. Number and location of incompetent venous segments Number of incompetent venous segments Location Ulcer (n = 16)
Dermatitis (n = 6)
Varicose veins (n = 10)
Superficial Deep Both Superficial Deep Both Superficial Deep Both
Mean +- SD 2.4 1.8 4.3 3.0 0.5 3.5 2.3 0.1 2.4
-+ 1.4 -+ 1 . 5 * t _+ 1.85 _+ 1.3 + 0.8 -+ 1.0§ -+ 1.3 -+ 0.3 _+ 1.2
Range
Median
0-4 0-4 1-8 1-4 0-2 2-5 1-4 0-1 1-4
3 1.5 4 3.5 0 3.5 2.5 0 2.5
*Ulcer versus dermatitis, p = 0.06 tUlcer versus varicose veins, p = 0.002 SUlcer versus varicose veins, p = 0.008 §Dermatitis versus varicose veins, p = 0.08
lower leg veins exceeded 8 mm, the VFI was always greater than the upper limit of normal (2.0 ml/sec). As shown in Fig. 7, the correlation between VV and VFI was r = 0 . 6 1 (r 2 = 0 . 3 8 ; p = 0 . 0 0 0 2 ) , but some correlation would be expected because the two measurements were not entirely independent. Although there was no apparent relationship between RVF and reservoir vein diameter (r = 0.12, r 2 -- 0.014,p -- 0.52), there was a poor (r -- 0.43, r 2 = 0.19) but statistically significant (p = 0.01) correlation between RVF and conduit vein diameter. Also, the VFI correlated (r = 0.46, r 2 = 0.21, p = 0.008) with the total number of venous segments that were found to be incompetent.
O u t f l o w obstruction versus EF. The relationship between EF and the presence or absence of superficial femoral vein obstruction (determined by duplex scanning) is shown in Fig. 8. In all limbs with an occluded deep venous outflow, the EF was less than 60%. EFs in the intermediate range (40% to 60%) were found in limbs with occluded, or partially recanalized, superficial femoral veins. A wide range of EFs was observed in legs with nonobstructed outflow; therefore, a low EF was not predictive of deep venous obstruction. Nine (56%) of 16 legs with ulcers, one (17%) of six legs with dermatitis, and none (0%) of 10 legs with varicose vein had venous outflow obstruction.
JOURNAL OF VASCULAR SURGERY Volume 18, Number 5
van Bemmelen et al.
801
Diameter (mm) 16
• ulcer • dermatitis 0 varicose veins
14 12
10
•
•
0 8
•
00 0 0 0 0 0
•
oo
•
Ulcer
Dermatitis
Varicose Veins
Fig.. 2. Relationship b e t w e e n diameters o f c o n d u i t veins a n d presence o f ulcers, dermatitis, or varicose veins.
D i a m e t e r (mm) 18
• ulcer • dermatitis
16
0 varicose veins
14 12 10 0
8 6
&&&
4
A A A A A A A A A •
• • @
0 O0 0 0
2 0
•
Ulcer
0 0 0 0
Dermatitis
V a r i c o s e Veins
Fig. 3. Relationship between diameters of reservoir veins in calf and presence of ulcers, dermatitis, or varicose veins.
Thus, with regard to the presence of venous obstruction, legs with ulcers differed significantly from those with varicose veins (p = 0.013). DISCUSSION Theoretically, the rate of venous reflux, as measured with the VFI, and the efficacy of the calf muscle pump, as measured with the EF, should correlate with the cutaneous manifestations of chronic venous insufficiency. In this study, however, air plethysmo-
graphic parameters did not discriminate well between limbs with severe manifestation of chronic venous insufficiency and limbs with uncomplicated varicose veins. Although the mean VFIs of legs with ulcers and legs with dermatitis were significantly greater than the mean VFI in legs with varicose veins, there was a great deal of overlap (Table I and Fig. 1). Moreover, 62% of limbs with ulcers and 33% with dermatitis had VFIs in the near normal range (< 5 ml/sec). Only one (4.5%) of the 22 legs in the total
802
JOURNAL OF VASCULARSURGERY November 1993
van B e m m d e n et al.
300-
- ulcer • dermatitis o varicose veins
250.A
2000
A •
VV (ml)
A
150100-
5o! 0
,m.
~o
. . . .
f
. . . .
I
6
0
. . . .
I
10
. . . .
16
I
30
d i a m e t e r (ml)
Fig. 4. Scattergram demonstrates correlation between VV measured with APG and diameter of incompetent lower leg veins (reservoir veins) measured by duplex scanning (r = 0.75).
10-
- ulcer • dermatitis o varicose veins
161412A
VFI (ml/sec)
108 0
6 A
A
A A A
A
4 2 0
A
0
A
8
0"
' ' 1 ' ' ' 1 ' ' ' 1 ' ' ' 1
2
4
6
" .
8
.
.
.
.
.
.
.
.
.
110 112 114
conduit diameter (mm) Fig. 5. Scattergram demonstrates correlation between VFI measured with APG and diameter of incompetent veins (conduit veins) at knee level measured by duplex scanning (r = 0.39).
group with severe skin changes had a VFI in the strictly normal range ( < 2 ml/sec), however. The mean EFs of all three groups were quite similar and were within the near normal range (> 40%). EFs were greater than 40% in 50% of the limbs with ulcers and in 67% of limbs with dermatitis. Combining the two parameters (VFI and EF) did not seem to improve the discrimination greatly (Fig. 1). Although the proportion of limbs with an elevated RVF was significantly greater in the ulcer group than it was in the group with varicose veins, there were many overlapping values (Table III). These observations are similar to those reported by Cordts et al. 12 in a much larger series of patients. Although they found that the mean VFI in extremities with chronic venous insufficiency (6.8 + 0.6
SEM) was greater than that in legs with varicose vein ( 5 . 4 _ 0.6 SEM), the difference failed to reach statistical significance (p = 0.06) and there was a remarkably similar pattern of distribution of the individual values. Forty-six percent of the legs with chronic venous insufficiency had VFIs less than 5 ml/sec. Also, there was little difference between the mean EFs of the two groups (chronic venous insufficiency, 5 1 . 9 _ 1.6 SEM; varicose veins, 48.6 _+ 1.8 SEM); again with markedly overlapping distributions. The authors concluded that "hemodynamic variables provided by air plethysmography are not predictive of the presence or absence of venous ulceration."12 In contrast, Christopoulos, Nicolaides, and Szendr• 4 reported a strikingly good correlation between
JOURNAL OF VASCULAR SURGERY
Volume18, Number 5
van Bemmelen et al.
., ulcer • dermatitis o varicose veins
18-
lei 14i 12~ VFI (ml/sec)
803
,o! 8-
o
6-
4. 2.
oot
0
''1'''1'''1'''1
0
2
4
6
.............
8 110 1'2 1'4 1'6 118
d i a m e t e r (ml)
Fig. 6. Scattergram demonstrates correlation between VFI and diameter of incompetent lower leg veins (reservoir veins) (r -- 0.55).
18le-
- ulcer • dermatitis o varicose veins
14! 12 i A A
VFI (ml/sec)
8-
O
6. 4. 2-
0
. . . _ , _ . . . ,
0
5O
. . . .
100
, . . . .
150
,
. . . .
200
,
. . . .
250
,
300
VV (ml) Fig. 7. Scattergram demonstrates relationship between VV and VFI (r = 0.61).
VFI and the incidence of ulcers or skin changes. Whereas no limbs with VFIs less than 3 ml/sec had chronic swelling or cutaneous manifestations of venous disease; 19% with VFIs between 3 and 5 ml/sec, 61% with VFIs of 5 to 10 ml/sec, and 76% with VFIs greater than 10 ml/sec had ulcers or severe skin changes. In our study, only 16% of the VFIs exceeded the 10 ml/sec limit, in spite of the large fraction (50%) of legs with severe ulceration. An explanation for the discrepancy between our results and those of Christopoulos et al. ~ is the exclusion by that group of patients with phlebographic evidence of chronic venous obstruction. Nicolaides et al.16 have also shown that the combination of VFI and EF has prognostic significance. When the results of these two examinations were plotted on a graph to provide six compa~:u,~ents
(I to VI in Fig. 1), the incidence of leg ulceration was found to be as follows: I (2%), II (30%), III (32%), IV (41%), V (63%), and VI (70%). Our study was too small to permit an estimation of the incidence of ulceration relative to the VFI-EF matrix and was not designed to do so; however, the relatively high proportion of limbs with ulcers or dermatitis in compaam~ent I seems inconsistent with the 2% incidence of ulcer reported for limbs with values in this compartment. Tourniquet tests have been described in conjunction with the APG to determine whether reflux can be reduced by occluding the superficial veins. A recent study by McMullin et al. 21 found that the amount of pressure necessary to occlude incompetent superficial veins depends on the circumference of the limb and is highly variable. They concluded that tourniquets
804
JOURNAL OF VASCULAR SURGERY November 1993
van Bemmelen et al.
100-
- ulcer • dermatitis o varicose veins
80" 8 8 8
/,m..
o~
60
.
tL
UJ
40.
:; A
8
-
!
20. 0
I
obstructed
I
non obstructed
Fig. 8. Scattergram demonstrates relationship between EF (%) and deep venous obstruction detected by duplex scanning. are a source of error because they often fail to exclude superficial reflux. Because of these observations and because the results of use of tourniquets in conjunction with photoplethysmography have proved to be disappointing, 22 we did not attempt to use tourniquets in this study. Duplex measurements of venous diameter were not significantly correlated with clinical presentation. In this respect, they were somewhat less discriminatory than the APG. Nonetheless, the mean diameter of reservoir veins was greatest (7.8 --- 5.2 mm) and the mean VFI was highest (7.7 +_ 4.6 ml/sec) in legs with the Society for Vascular Surgery and the North American Chapter of the International Society for Cardiovascular Surgery class 3 dermatitis. Legs with ulceration had reservoir vein diameters of 4.9 +__2.1 mm, which corresponded to a mean VFI of 5.4 __ 3.8 ml/sec. The smallest reservoir vein diameters (3.2 ___ 3.2 mm) and the lowest mean VFI (2.6 ___ 1.7 ml/sec) were found in class 1 limbs with uncomplicated varicose veins. It seems reasonable to assume that calf W and the rate of venous reflux (VFI) should be related to the presence of large incompetent conduit and lower leg reservoir veins. This assumption is supported by the statistically significant (but mostly rather poor) correlations that were observed between W and calf vein (reservoir vein) diameters, VV and VFI, VFI and conduit vein diameters, VFI and reservoir vein diameters, and VFI and the number of incompetent venous segments. O f the five limbs with a VFI equal to or greater than 10 ml/sec, all had venous diameters greater than 6.0 mm in either the calf or conduit veins, and four had venous diameters greater than 6.0 mm in both areas, suggesting that the presence of both incompetent conduit and lower-leg reservoir
veins of these dimensions are required to attain a VFI in excess of 10 ml/sec. Many of the ulcerated legs had VFIs in the intermediate (near normal) 2 to 5 ml/sec range (Fig. 1). Values in this range can occur when a large incompetent conduit-vein is present at the knee level but the calf reservoir is small. Likewise, if the reflux conduit veins are small and the calf reservoir is large, the VFI can be in the 2 to 5 ml/sec range. VFIs in the intermediate range also occurred in the absence of any detectable incompetent veins crossing the knee level-in other words, in the absence of any reflux conduit. This raises the question whether an increased calf volume capacity alone is sufficient to increase VFI over normal levels. The VFI may also be affected by variations in the rate of arterial inflow. 23 Although arteriolar constriction reduces blood flow in dependent legs (the venoarterial reflex),24 we have observed that "resting" blood flow in the popliteal artery may be transiently increased several fold when subjects assume an erect position, largely as a result of an elevated diastolic velocity (Fig. 9). Because the APG test involves emptying the calf veins by leg elevation before standing, the venous pressure approaches zero. The arterial pressure, on the other hand, increases commensurate with the length of the hydrostatic column to the heart. This combination of increased arterial pressure and (temporarily) reduced venous pressure augments the pressure gradient across the resistance vessels and, in so doing, increases the rate of arterial blood flow. The increased flow occurs during the portion of the refilling curve used to calculate the VFI. Therefore, it is possible that the increase in arterial filling might be prolonged when a larger than normal, relatively empty, low pressure venous reser-
JOURNALOF VASCULARSURGERY Volume 18, Number 5
v a n B e m m e l e n et al.
805
Fig. 9. A, Resting popliteal artery flow-velocity signal in normal subject in supine position with elevation of leg. Note absence of antegrade diastofic flow and large retrograde flow component (Peak systolic velocity = 72 cm/sec and peak reverse flow velocity = - 37 cm/sec). B, Same artery immediately after assuming erect position, without weight-bearing. Note increase in antegrade diastolic flow. In absence of venous valvular reflux, increased arterial volume-flow is related to rate of filling of lower leg veins (Peak systolic velocity -- 67 cm/sec and peak diastolic velocity = 32 cm/sec).
voir is present, regardless of valve function at the knee level. Although measurement of blood flow by venous occlusion plethysmography has been recommended as a method for assessing the contribution of arterial inflow to the rate of venous filling,23 it is by no means certain that flow rates measured in the supine position are predictive of those that occur with standing. Total limb blood flow may decrease by 50% as a result of the venoarterial reflex24 or increase
(as we have observed) owing to changes in the arteriovenous pressure gradient. In severely ischemic limbs with an impaired venoarterial reflex, blood flow may increase by as much as 40% on standing. 24 Because of the variability of the blood flow response to position change, we made no attempt to estimate flow rates plethysmographically. The incidence of significant arterial disease is often underestimated in patients with chronic ulcers. 2s27 H o w arterial obstruction might affect
806
van Bemmelen et al.
orthostatic changes in blood flow is difficult to predict in any given subject, but it may be responsible for some of the variation in VFI reported in the literature. Because none of the patients in our study had overt arterial disease, this factor could not be responsible for the disparity in APG and duplex results that we observed. An unresolved issue is whether the volume of venous reflux has major clinical significance. According to Bergan's study, 2s which focused on the greater saphenous vein, there is a disappointing lack of correlation between the quantity of reflux as measured by duplex scanning and clinical classification of disease severity. This is not in accordance with the results ofVasdekis et al., 14which indicated that reflux greater than 10 ml/second in incompetent veins was associated with a high incidence of skin changes (66%), regardless of whether the reflux was in the deep or superficial system. Most of the patients in the study by Vasdekis et al. 14 had isolated superficial reflux, similar to Bergan's2s series. Air plethysmographic studies, which measure reflux to the entire calf, emphasize reflux quantity as an important parameter of venous disease. ~ Results of the present study suggest that duplex scanning and air plethysmography may not measure the same reflux (only 4 of 22 limbs with class 3 disease had reflux greater than 10 ml/sec). This might be explained by the effect of variable arterial inflow on plethysmographic measurements and by the disparate nature of the venous structures being interrogated. Calf VV and its reduction on muscle contraction (EF) are of interest to distinguish between postthrombotic calf pumps, with obliterated veins that have a small volume, and varicose disease with an increase in the size of the venous reservoir below the knee. In our study, EF was often reduced below 60% by an increase in the denominator (e.g. venous volume increased in limbs with superficial varicosities), rather than by a reduction in the numerator (stroke volume). In legs with numerous large functional collateral vessels, the EF was in the 40% to 60% range, even in the presence of an obstructed deep venous outflow (Fig. 8). Because EFs vary widely in limbs with nonobstructed outflow, a low EF does not predict deep obstruction; but EFs greater than 60% were not observed in legs with deep venous obstruction, All legs with an obstructed venous outflow had severe manifestations of chronic venous insufficiency, however. In summary, the APG parameters measured in this study did not effectively distinguish limbs with chronic venous insufficiencyfrom limbs with uncom-
JOURNAL OF VASCULARSURGERY November 1993
plicated varicose veins; nor did they reliably predict the presence or size of incompetent reflux conduit veins or reflux-reservoir veins. In part, this may be related to orthostatic variations in arterial inflow. For the treatment of patients with chronic venous insufficiency, a noninvasive method that is independent of arterial blood flow, that is not affected by arterial disease, and that does not rely on the use of tourniquets for the localization of incompetent veins is desirable. Duplex scanning findings are pertinent to the selection of patients for surgical therapy and for ascertaining what operation should be performed. For these reasons, duplex ultrasonography is unlikely to be replaced by functional measurements. Measurements of VV reflux by APG and the effect of muscle contraction continue to be of interest, however, to help solve therapeutic dilemmas, which occur when superficial incompetence is complicated by moderate deep venous obstruction. APG results also provide functional information that may be important in assessing the results of therapy. For the comprehensive workup of patients with dermatitis and ulceration, duplex scanning, air plethysmography, and arterial evaluation are all important. We thank Jeffrey L. Collins, Dianne L. Raymond, RVT, and Paula PetriUi for their technical assistance. REFERENCES 1. Hojensgard IC, Smrup H. Venous pressure in pr~ary and post-thrombotic varicose veins. Study of statics and dynamics of venous system of lower extremities under pathologic conditions. Acta Chir Scand 1949;99:133-53. 2. Abramowitz HB, Queral LA, Flinn WR, et al. The use of photoplethysmography in the assessment of venous insufficiency: a comparison to venous pressure measurements. Surgery 1979;86:434-41. 3. Allan 1C. Volume changes in the lower limb in response to postural alterations and muscular exercise. S. Afr J Surg 1964;2:75-90. 4. Christopoulos D, Nicolaides, Szendro G. Venous reflux: quantification and correlation with the clinical severity of chronic venous disease. Br J Surg 1988;75:352-6. 5. Christopoulos D, Nicolaides AN. Noninvasive diagnosis and quantitation of popliteal reflux in the swollen and ulcerated leg. J Cardiovasc Surg 1988;29:535-9. 6. Christopoulos D, Nicolaides AN, Cook A, Irvine A, Galloway JMD, Wilkinson A. Pathogenesis of venous ulceration in relation to the calf muscle pump function. Surgery 1989;106: 829-35. 7. Christopoulos D, Nicolaides AN, Galloway JMD, Wilkinson A. Objective noninvasiveevaluation of venous surgical results. J VASC SUP,G 1988;8:683-7. 8. Christopoulos DG, Nicolaides AN, Szendro G, Irvine AT, Bull M, Eastcott HHG. Air-plethysmography and the effect of elastic compression on venous hemodynamics of the leg. J V^sc SURG 1987;5:148-59.
JOURNAL OF VASCULARSURGERY Volume 18, Number 5
9. Christopoulos D, Nicolaides AN, Belcaro G, Duffy P. The effect of elastic compression on calf mnsde pump function. Phlebology 1990;5:13-9. 10. Welkie JF, Comerota AJ, Kerr RP, Katz ML, Jayheimer EC, Brigham RA. The hemodynamics of venous ulceration. Ann Vasc Surg 1992;6:1-4. 1I. Welkie1F, Kerr RP, Katz ML, Comerota AJ. Can noninvasive venous volume determinations accuratelypredict ambulatory venous pressure? J Vasc Technology 1991;15:186-90. 12. Cordts PR, Hartono C, LaMorte WW, Menzoian JO. Physiologic similarities between extremities with varicose veins and with chronic insufficiencyutilizing air plethysmography. Am J Surg 1992;164:260-4. 13. van Bemmelen PS, Bedford G, Beach K, Strandness DE. Quantitative segmental evaluation of venous valvular reflux with duplex ultrasound scanning. J Vasc Surg 1989;10:42531. 14. Vasdekis SN, Clarke GH, Nicolaides AN. Quantification of venous: reflux by means of duplex scanning. J VAsc SURG 1989;10:670-7. 15. Porter JM, Rutherford RB, Clagett GP, et al. Reporting standards in venous disease. J VASC St3RG 1988;8:172-81. 16. Nicolaides AN, Sumner DS. Investigation of patients with deep venous thrombosis and chronic venous insufficiency. London: Med-Orion Publishing Co., 1991:1-73. 17. CzeredarzukM, Branas CC, Weingarten MS. Duplex imaging and distal cuff deflation to measure venous reflux time. J Vasc Technol 1992;15:284-7. 18. van Bemmelen PS, Beach K, Bedford G, Strandness DE. The mechanism of venous valve closure. Arch Surg 1990;125: 617-9. 19. van Bemmelen PS, Bedford G, Beach K, Strandness DE Jr. Status of the valves in the superficialand deep venous system in chronic venous disease. Surgery 1991;109:730-4.
van Bemmelen et al.
807
20. van Bemmelen PS, Bedford G, Beach K, Strandness DE. Functional status of the deep venous system after an episode of deep venous thrombosis. Ann Vase Surg 1990;5: 455-9. 21. McMullin GM, Coleridge Smith PD, Scurr JH. A study of tourniquets in the investigation of venous insufficiency. Phlebology 1991;6:133-9. 22. van Bemmelen PS, van Ramshorst B, Eikelboom B. Photoplethysmography reexamined: lack of correlation with duplex scanning. Surgery 1992;112:544-8. 23. Pierce EC II, Chiang K, SchanzerH. Volume tests for chronic venous insufficiency: an appraisal. Surgery 1991;109:56774. 24. Morgan RH, Psaila JV, Stone J, Carolan G, Woodcock JP. Postural changes in femoral artery blood flow in normal subjects, patients with peripheral vascular occlusive disease and patients undergoing lumbar sympathectomy, measured by duplex ultrasound flowmetry. Eur J Vase Surg 1992;6: 408-15. 25. Cornwall JV, Dore CJ, Lewis JD. Leg ulcers: epidemiology and aetiology. Br J Surg 1986;73:693-6. 26. CallamMJ, Harper DR, Dale JJ, Ruckley CV. Arterial disease in chronic leg ulceration; an underestimated hazard? Lothian and Forth Valley leg ulcer study. BMJ 1987;294:929-31. 27. Nelzen O, Bergqvist D, Lindhagen A. Leg ulcer etiology-a cross sectional population study. }"Vase SuRG 1991;14:55764. 28. Bergan JJ. Clinical application of duplex testing in treatment of primary venous stasis,varicoseveins. In: van BemmelenPS, Bergan JJ, eds. Quantitative measurement of venous incompetence. Austin: RG Landis, 1992:78-104.
Submitted April 7, 1993; accepted July 21, 1993.
Purpose: Duplex ultrasonography with distal cuff deflation was used to determine the presence and size of incompetent veins and compare the results with those of air plethysmography in patients with chronic venous insufficiency. Methods: Thirty-two legs underwent a detailed study with both modalities. Sixteen legs had venous ulceration, six had stasis dermatitis, and ten had symptomatic varicose veins without skin changes. Results:Although the venous filling index (VFI) in limbs with ulcers (5.4 + 3.8 ml/sec) and dermatitis (7.7 +- 4.6 ml/sec) was significantly higher (p < 0.05) than it was in limbs with varicose veins (2.6 -+ 1.7 ml/sec), there was a large amount of overlap. Only 13% of ulcerated legs had VFI greater than 10 ml/sec. Sixty-three percent of legs with ulcers, 33% of legs with dermatitis, and 90% of legs with varicose veins had VFIs less than 5 ml/sec. Mean ejection fractions (EFs) in the three groups were similar, ranging from 45% to 52%. Combining VFI and EF did not lessen the overlap between groups. Forty-one percent of limbs with ulcers or dermatitis had air plethysmography parameters in the normal or intermediate area (VFI < 5 ml/sec; EF > 40%), which in previous studies corresponded to an incidence of ulceration of only 2%. VFI had a significant but weak correlation (r = 0.39) with the diameter of incompetent veins at the knee and a somewhat stronger relationship (r = 0.55) with the diameter of lower leg veins. Total venous volume correlated moderately well with calf vein diameter (r = 0.75). The clinical status of the leg did not correlate with the diameters of incompetent veins at the knee or calf levels. All limbs with an obstructed outflow had EFs less than 60% and ulcers or dermatitis. Conclusions: We conclude that plethysmographic measurements of functional venous parameters (VFI, EF) do not discriminate well between limbs with uncomplicated varicose veins and limbs with ulcers or stasis dermatitis and that VFI correlates poorly with the presence of incompetent veins and their diameters. Both duplex scanning and plethysmography seem to be necessary for a complete evaluation of limbs with chronic venous insufficiency. (J VASC SURG 1993;18:796-807.)
Semiquantitative methods for evaluating venous function in patients with chronic venous insufficiency include venous pressure measurements, 1 photoplethysmography, 2 and air plethysmography, a Recently,
From the Department of Surgery, Section of Peripheral Vascular Surgery, Southern Illinois University School of Medicine, Springfield. Presented at the Fifth Annual Meeting of the American Venous Forum, Orlando, Fla., Feb. 24-26, 1993. Reprint requests: David S. Sumner, MD, Southern Illinois UniversitySchoolof MedicineDepartment of Surgery,PO Box 19230, Springfield,IL 62794-9230. Copyright © 1993 by The Society for Vascular Surgery and International Societyfor CardiovascularSurgery,North American Chapter. 0741-5214/93/$1.00 + .10 24]6/50279 796
the development o f an air plethysmograph (APG) 4 that surrounds the entire calf has renewed interest in assessing the function o f venous valves and the calf venous pump. Investigators at St. Mary's Hospital in L o n d o n have established normal values for the rate o f venous filling that accompanies the assumption o f an uptight position and for the reduction in calf volume that occurs with tiptoe exercise? -6 The venous filling index (VFI) is a measure o f reflux flow through incompetent venous valves, and the ejection fraction (EF) relates to the function o f the calf muscle pump. A decrease in the VFI has been described after varicose vein surgery 7 and with the use o f elastic support stockings. 8,9 Another group, reporting on their results with air plethysmography, noted a clear
JOURNAL OF VASCULAR SURGERY
Volume18, Number 5
distinction between the VFI in normal volunteers and in patients with severe chronic venous insufficiency.1° This group also observed a correlation between end-exercise venous volume (VV) and ambulatory venous pressure? ~ Other investigators, however, have not found air plethysmographic parameters to be helpful in differentiating between simple varicose veins and chronic venous insufficiency.~1 Ultrasonic duplex or color-flow imaging performed at multiple levels of the leg provides a precise method for locating incompetent veins and for quantitating reflux flow, ~3 but in patients with multiple incompetent veins, calculating the volume of reflux flow by this method is cumbersome? 4 The goals of our study were (1) to ascertain whether air plethysmographic parameters distinguish between uncomplicated varicose veins and limbs with severe skin changes; (2) to correlate the results of air plethysmography with anatomic measurements made with duplex scanning; and (3) to determine whether duplex scanning can be replaced by the simpler, less expensive APG method in certain patient groups. Furthermore, we attempted to determine whether the diameter of incompetent veins in patients with severe manifestations of chronic venous insufficiency (ulcers or stasis dermatitis) differs from that of patients with simple varicose veins. MATERIAL A N D M E T H O D S
Thirty-two legs of 28 patients with symptoms of chronic venous insuffÉciency were studied by both segmental color-flow duplex scanning and air plethysmography. Ulcers were present in 16 legs of 16 patients (mean age 56.8 + 15.0 years). In six legs there was severe swelling with wet dermatitis (four patients, mean age 50.3 + 6.0 years). Varicose veins were present in 10 legs of eight patients (mean age 47.1 _+ 15.6 years), who complained of fatigue with prolonged standing. According to the criteria estabfished by the Society for Vascular Surgery and the North American Chapter of the International Society for Cardiovascular Surgery, all limbs with dermatitis and ulcers were class 3 severity, and limbs with varicose veins were class 175 Air plethysmography. Methods for measuring changes in calf venous volume with air plethysmography have been described in detail by other authors, s The leg is enclosed from ankle to knee in a 14-inch polyurethane cuffthat acts as the sensing device (ACI Medical, Sun Valley, Calif.). Changes in air pressure are amplified and used to drive a strip chart recorder.
van Bemmelen et al.
797
In this study a standard protocol was used, which began with the subject resting in the supine position with the leg elevated on a foam block for 5 minutes. Pressure in the cuff was adjusted to 6 mm Hg. This is the "bias" pressure, which serves to keep the cuffin Contact with the skin. Calibration was performed by injecting 100 rnl of air into the system. The zero VV of the leg was determined by passively elevating the leg 45 °. After a stable baseline had been reached, the subject was instructed to assume a standing position without putting any weight on the leg being studied. In a normal situation, a slow increase in leg size is noted because of filling of veins from arterial inflow. The increase in calf volume in milliliters is called the VV, and the time in seconds required to reach a stable plateau is called the venous filling time. VFIs in ml/sec were calculated by dividing the VV at 90% refilling by the venous filling time required to achieve 90% refilling. The subject was then asked to perform a single tiptoe exercise, which was subsequently repeated. The resulting decrease in leg volume, divided by the maximal VV, is termed the EF. After this, 10 additional tiptoe maneuvers were performed to determine the residual venous volume; and afterwards the leg was elevated again to disclose any shift in the zero volume. Residual volume fraction (RVF) was calculated by dividing the residual venous volume by the VV. Normal legs have a VFI less than 2 ml/sec and an EF greater than 60%. 4,s VFIs greater than 5 ml/sec, EFs less than 40%, and RVFs greater than 40% are considered to be abnormal. 16 An increase in VFI or a decrease in EF are reported to be associated with an increased likelihood of venous ulceration.4,s,l° RVFs correlate well with ambulatory venous pressure, the gold standard for evaluating calf muscle pump function, u The incidence of ulceration rises steadily as the RVF increases. 6 Duplex scanning. The Quantum 2000 colorflow scanner (Siemens-Quantum, Issaquah, Wash.) was used for all studies. Evaluation of valve competence was performed with the distal cuff deflation maneuver? 3'17 This method allows evaluation of venous valvular function at distal levels of the leg, sites that cannot be tested with a Valsalva maneuver. 18 Studies were conducted with the patient standing holding onto an orthopedic frame, with his or her weight being supported by the contralateral leg. A Hokanson (Hokanson, Inc., D.E., Bellevue, Wash.) cuff inflator was used to ensure rapid (< 0.3 seconds) cuff inflation and deflation. Venous valvular competence was assessed in both the deep and superficial systems at the midthigh, knee, calf, and
798
v a n Bemraelen et al.
ankle levels. At each level, cuffs were placed around the limb several centimeters below the site of duplex interrogation. To exceed hydrostatic pressures, cuffs were inflated to 80 mm Fig, 100 mm Hg, and 120 mm H g at the thigh, calf, and foot levels, respectively. Color-flow duplex scanning was used to identify incompetent veins and to record the direction and duration of flow during cuff deflation. Reflux flow persisting more than 0.5 seconds identifies venous incompetence. 13,19Although peak reflux flow velocities at the popliteal level were recorded, no attempt was made to calculate total reflux flow rates. The multiplicity of veins in this region and the errors inherent in calculating total flow make such measurements difficult and unreliable. Duplex scanning was used to measure the diameter of incompetent veins passing the knee. These veins were designated "reflux conduit" veins. The main pathways for venous reflux at the knee are the greater saphenous vein and the popliteal vein. An additional search was made to exclude other incompetent veins, such as anterolateral branches of the greater saphenous vein or lesser saphenous veins with a high termination. The diameters of incompetent posterior tibial and peroneal veins in the calf lower leg were also measured; these veins were termed "reflux reservoir" veins. Reservoir veins accommodate reflux flow once it reaches the lower leg. It is the increase in diameter of these veins that is responsible for the increase in calf volume measured with the APG. If incompetent veins were absent, a diameter of 0 mm was assigned. Outflow function. Duplex scanning was used to evaluate the superficial femoral vein as the main deep venous outflow conduit. In addition to visualizing residual thrombus in the superficial femoral vein, venous obstruction was identified by measuring outflow velocity in the proximal femoral vein during standardized augmentation of venous flow produced by rapid ( < 0 . 3 seconds) inflation of a 24 cm cuff placed around the calf. Legs were classified as normal when outflow velocities were in the 143 + 58 cm/sec range, z° Outflow velocities less than 85 cm/sec were considered to be consistent with incomplete recanalization of the deep system. Statistics. The Mann-Whitney U test (two-tailed) was used to compare grouped data, and Fisher's exact test was used to compare proportional data. A p value less than 0.05 was considered significant. Correlations between measurements were calculated by the Pearson's correlation coefficient method.
JOURNAL OF VASCULAR SURGERY November 1993
RESULTS Plethysmographic results are compared with clinical presentations in Tables I to III. In all clinical groups the VFI varied widely, but mean values obtained in limbs with ulcers and dermatitis were significantly higher than those found in limbs with simple varicose veins. Also, the proportion of legs with abnormal VFIs ( ___5 ml/sec) was higher in legs with dermatitis than it was in legs with varicose veins (Table I). There was little difference between the mean EFs of the three groups (Table II). RVFs were higher in limbs with ulcers than in limbs with dermatitis or varicose veins, and the proportion of limbs with abnormally high RVFs was significantly greater in the group with ulcers than it was in the group with varicose veins (Table III). In limbs with dermatitis, the increase in calf volume (172 + 92 ml) that accompanied assumption of a standing position (VV) was significantly greater (p = 0.04) than it was in limbs with varicose veins (101 _+ 35 ml). The VV in ulcerated legs was 120 -+ 44 ml. In Fig. 1, the results from each limb are plotted according to EF and VFI. The graph is divided into six compartments (I to VI) by a horizontal line corresponding to an EF of 40% and by vertical lines corresponding to VFIs of 5 and 10 ml/sec. The compartments of the matrix are numbered to correspond with increasing severity of hemodynamic compromise (with I as near normal and VI as the worst). 16 In our study 13 (59%) of 22 limbs with ulcers or dermatitis had values falling in compartments II to VI. Only three (30%) limbs with varicose veins had values in these compartments. This difference, however, failed to reach statistical significance (p = 0.25). Sixteen normal volunteers underwent APG testing before the start of this study. All had VFIs less than 2 ml/sec and EFs greater than 60%. Thus all normal limbs in our laboratory had APG values falling within the rectangle labeled "N" in Fig. 1. None of the venous segments in these subjects had reflux durations exceeding 0.5 seconds. Location and extent o f venous incompetence. Because four levels of the superficial and deep veins were studied, it was possible to record a maximum of eight incompetent segments in each limb. As shown in Table IV, there was no appreciable difference in the number of incompetent superficial venous segments in the three clinical groups. Limbs with ulcers, however, had more incompetent deepvenous segments than limbs with varicose veins (p = 0.008) or limbs with dermatitis (p = 0.06). Both the ulcer and dermatitis groups tended to have more total seg-
JOURNAL OF VASCULAR SURGERY Volume 18, Number 5
v a n B e m m e l e n et al.
799
Table I. VFI versus clinical presentation Abnormal ( >-5 ml/sec)
VFI (ml/sec)
Ulcer Dermatitis Varicose veins
n
Mean +- SD
Range
n
%
16 6 10
5.4 -+ 3.8 7.7 -+ 4.6 2.6 + 1.7"
2.0-16 1.4-13 1.2-7
6 4 1
38 67 10t
*Varicose veins versus dermatitis, p = 0.034; varicose veins versus ulcer, p = 0.003. tVaricose veins versus dermatitis, p = 0.036.
Table II. EF versus clinical presentation Abnormal (<40%)
EF (%)
Ulcer Dermatitis Varicose veins
n
Mean 4- SD
Range
n
%
16 6 10
45 -+ 22 49 -+ 20 52 _+ 16
14-92 22-72 25-76
8 2 3
50 33 30
N o statistically significant comparisons.
Table III. RVF versus clinical presentation Abnormal (>40%)
R V F (%)
Ulcer Dermatitis Varicose veins
n
Mean + SD
Range
n
%
16 6 10
60 -+ 23 44 _+ 17 43 -+ 21 ~
12-87 22-60 15-76
14 3 4
88 50 40j-
*Varicose veins versus ulcer, p = 0.067. j-Varicose veins versus ulcer, p = 0.026.
ments involved (superficial and deep) than the varicose vein group did (p = 0.008 and p = 0.08, respectively). Venous diameter. In all cases, duplex scanning identified incompetence of either the conduit or reservoir veins, as indicated by reflux persisting for greater than 0.5 sec after deflation of the calf or foot cuff. Incompetence was present at both levels in 23 (72%) of the 32 limbs. Diameters of both the reflux conduit veins and the reflux reservoir veins varied widely, and measurements for the three clinical groups overlapped over a wide range (Figs. 2 and 3). The mean diameters of the conduit veins in the ulcer group (6.3 - 3.1 mm) and in the dermatitis group (8.3--+ 4.8) did not differ significantly from the mean diameter of the varicose vein group (4.6 _+ 2.9 mm) (p = 0.18 and p = 0.06, respectively). Although the mean diameter of reservoir veins in the dermatitis group (7.8 +__5.2 mm) was twice that of reservoir veins in the varicose vein group (3.2 _+ 3.2
mm), this difference failed to reach statistical significance (p = 0.09). The difference was due largely to four veins with "zero" diameters in the latter group. The mean diameter of reservoir veins in the ulcer group was 4.9 _+ 2.1 mm (p = 0.16, versus varicose veins). Venous diameter versus air plethysmographic findings. VV was significantly correlated (r = 0.75, ,'2 = 0.56;p < 0.0001) with the diameter of incompetent calf veins (reservoir veins) as measured by duplex scanning (Fig. 4). A weak (r = 0.39, r e = 0.15) but statistically significant correlation (p = 0.03) was noted between the VFI and the diameter of incompetent conduit veins at knee level (Fig. 5). VFI also correlated (r = 0.55, re = 0.30; p = 0.001) with the diameter of incompetent reservoir veins in the lower leg (Fig. 6). VFIs greater than 5 ml/sec were not observed in limbs with incompetent calf veins less than 4 mm in diameter. On the other hand, when the diameters of incompetent
800
JOURNAL OF VASCULAR SURGERY November 1993
van Bemmelen et al.
I
IIl
20
• ulcer • dermatSIs o varicose veins
Vl
A•
&t)
~, o
0
4O
I
EF (%)
d • 0
IV
II
A
60--
80
1O0
,
O
•
I
0
5
0
15
20
VFI ( m l / s e e ) Fig. 1. EF plotted versus VFI. Triangles indicate limbs with ulcers; closedcircles indicate limbs with dermatitis; and open circles indicate limbs with varicose veins. Compartments are numbered (I to VI) according to increasing severity of physiologic abnormality. Rectangle (N) in lower left corner indicates reported range of normal values. Table IV. Number and location of incompetent venous segments Number of incompetent venous segments Location Ulcer (n = 16)
Dermatitis (n = 6)
Varicose veins (n = 10)
Superficial Deep Both Superficial Deep Both Superficial Deep Both
Mean +- SD 2.4 1.8 4.3 3.0 0.5 3.5 2.3 0.1 2.4
-+ 1.4 -+ 1 . 5 * t _+ 1.85 _+ 1.3 + 0.8 -+ 1.0§ -+ 1.3 -+ 0.3 _+ 1.2
Range
Median
0-4 0-4 1-8 1-4 0-2 2-5 1-4 0-1 1-4
3 1.5 4 3.5 0 3.5 2.5 0 2.5
*Ulcer versus dermatitis, p = 0.06 tUlcer versus varicose veins, p = 0.002 SUlcer versus varicose veins, p = 0.008 §Dermatitis versus varicose veins, p = 0.08
lower leg veins exceeded 8 mm, the VFI was always greater than the upper limit of normal (2.0 ml/sec). As shown in Fig. 7, the correlation between VV and VFI was r = 0 . 6 1 (r 2 = 0 . 3 8 ; p = 0 . 0 0 0 2 ) , but some correlation would be expected because the two measurements were not entirely independent. Although there was no apparent relationship between RVF and reservoir vein diameter (r = 0.12, r 2 -- 0.014,p -- 0.52), there was a poor (r -- 0.43, r 2 = 0.19) but statistically significant (p = 0.01) correlation between RVF and conduit vein diameter. Also, the VFI correlated (r = 0.46, r 2 = 0.21, p = 0.008) with the total number of venous segments that were found to be incompetent.
O u t f l o w obstruction versus EF. The relationship between EF and the presence or absence of superficial femoral vein obstruction (determined by duplex scanning) is shown in Fig. 8. In all limbs with an occluded deep venous outflow, the EF was less than 60%. EFs in the intermediate range (40% to 60%) were found in limbs with occluded, or partially recanalized, superficial femoral veins. A wide range of EFs was observed in legs with nonobstructed outflow; therefore, a low EF was not predictive of deep venous obstruction. Nine (56%) of 16 legs with ulcers, one (17%) of six legs with dermatitis, and none (0%) of 10 legs with varicose vein had venous outflow obstruction.
JOURNAL OF VASCULAR SURGERY Volume 18, Number 5
van Bemmelen et al.
801
Diameter (mm) 16
• ulcer • dermatitis 0 varicose veins
14 12
10
•
•
0 8
•
00 0 0 0 0 0
•
oo
•
Ulcer
Dermatitis
Varicose Veins
Fig.. 2. Relationship b e t w e e n diameters o f c o n d u i t veins a n d presence o f ulcers, dermatitis, or varicose veins.
D i a m e t e r (mm) 18
• ulcer • dermatitis
16
0 varicose veins
14 12 10 0
8 6
&&&
4
A A A A A A A A A •
• • @
0 O0 0 0
2 0
•
Ulcer
0 0 0 0
Dermatitis
V a r i c o s e Veins
Fig. 3. Relationship between diameters of reservoir veins in calf and presence of ulcers, dermatitis, or varicose veins.
Thus, with regard to the presence of venous obstruction, legs with ulcers differed significantly from those with varicose veins (p = 0.013). DISCUSSION Theoretically, the rate of venous reflux, as measured with the VFI, and the efficacy of the calf muscle pump, as measured with the EF, should correlate with the cutaneous manifestations of chronic venous insufficiency. In this study, however, air plethysmo-
graphic parameters did not discriminate well between limbs with severe manifestation of chronic venous insufficiency and limbs with uncomplicated varicose veins. Although the mean VFIs of legs with ulcers and legs with dermatitis were significantly greater than the mean VFI in legs with varicose veins, there was a great deal of overlap (Table I and Fig. 1). Moreover, 62% of limbs with ulcers and 33% with dermatitis had VFIs in the near normal range (< 5 ml/sec). Only one (4.5%) of the 22 legs in the total
802
JOURNAL OF VASCULARSURGERY November 1993
van B e m m d e n et al.
300-
- ulcer • dermatitis o varicose veins
250.A
2000
A •
VV (ml)
A
150100-
5o! 0
,m.
~o
. . . .
f
. . . .
I
6
0
. . . .
I
10
. . . .
16
I
30
d i a m e t e r (ml)
Fig. 4. Scattergram demonstrates correlation between VV measured with APG and diameter of incompetent lower leg veins (reservoir veins) measured by duplex scanning (r = 0.75).
10-
- ulcer • dermatitis o varicose veins
161412A
VFI (ml/sec)
108 0
6 A
A
A A A
A
4 2 0
A
0
A
8
0"
' ' 1 ' ' ' 1 ' ' ' 1 ' ' ' 1
2
4
6
" .
8
.
.
.
.
.
.
.
.
.
110 112 114
conduit diameter (mm) Fig. 5. Scattergram demonstrates correlation between VFI measured with APG and diameter of incompetent veins (conduit veins) at knee level measured by duplex scanning (r = 0.39).
group with severe skin changes had a VFI in the strictly normal range ( < 2 ml/sec), however. The mean EFs of all three groups were quite similar and were within the near normal range (> 40%). EFs were greater than 40% in 50% of the limbs with ulcers and in 67% of limbs with dermatitis. Combining the two parameters (VFI and EF) did not seem to improve the discrimination greatly (Fig. 1). Although the proportion of limbs with an elevated RVF was significantly greater in the ulcer group than it was in the group with varicose veins, there were many overlapping values (Table III). These observations are similar to those reported by Cordts et al. 12 in a much larger series of patients. Although they found that the mean VFI in extremities with chronic venous insufficiency (6.8 + 0.6
SEM) was greater than that in legs with varicose vein ( 5 . 4 _ 0.6 SEM), the difference failed to reach statistical significance (p = 0.06) and there was a remarkably similar pattern of distribution of the individual values. Forty-six percent of the legs with chronic venous insufficiency had VFIs less than 5 ml/sec. Also, there was little difference between the mean EFs of the two groups (chronic venous insufficiency, 5 1 . 9 _ 1.6 SEM; varicose veins, 48.6 _+ 1.8 SEM); again with markedly overlapping distributions. The authors concluded that "hemodynamic variables provided by air plethysmography are not predictive of the presence or absence of venous ulceration."12 In contrast, Christopoulos, Nicolaides, and Szendr• 4 reported a strikingly good correlation between
JOURNAL OF VASCULAR SURGERY
Volume18, Number 5
van Bemmelen et al.
., ulcer • dermatitis o varicose veins
18-
lei 14i 12~ VFI (ml/sec)
803
,o! 8-
o
6-
4. 2.
oot
0
''1'''1'''1'''1
0
2
4
6
.............
8 110 1'2 1'4 1'6 118
d i a m e t e r (ml)
Fig. 6. Scattergram demonstrates correlation between VFI and diameter of incompetent lower leg veins (reservoir veins) (r -- 0.55).
18le-
- ulcer • dermatitis o varicose veins
14! 12 i A A
VFI (ml/sec)
8-
O
6. 4. 2-
0
. . . _ , _ . . . ,
0
5O
. . . .
100
, . . . .
150
,
. . . .
200
,
. . . .
250
,
300
VV (ml) Fig. 7. Scattergram demonstrates relationship between VV and VFI (r = 0.61).
VFI and the incidence of ulcers or skin changes. Whereas no limbs with VFIs less than 3 ml/sec had chronic swelling or cutaneous manifestations of venous disease; 19% with VFIs between 3 and 5 ml/sec, 61% with VFIs of 5 to 10 ml/sec, and 76% with VFIs greater than 10 ml/sec had ulcers or severe skin changes. In our study, only 16% of the VFIs exceeded the 10 ml/sec limit, in spite of the large fraction (50%) of legs with severe ulceration. An explanation for the discrepancy between our results and those of Christopoulos et al. ~ is the exclusion by that group of patients with phlebographic evidence of chronic venous obstruction. Nicolaides et al.16 have also shown that the combination of VFI and EF has prognostic significance. When the results of these two examinations were plotted on a graph to provide six compa~:u,~ents
(I to VI in Fig. 1), the incidence of leg ulceration was found to be as follows: I (2%), II (30%), III (32%), IV (41%), V (63%), and VI (70%). Our study was too small to permit an estimation of the incidence of ulceration relative to the VFI-EF matrix and was not designed to do so; however, the relatively high proportion of limbs with ulcers or dermatitis in compaam~ent I seems inconsistent with the 2% incidence of ulcer reported for limbs with values in this compartment. Tourniquet tests have been described in conjunction with the APG to determine whether reflux can be reduced by occluding the superficial veins. A recent study by McMullin et al. 21 found that the amount of pressure necessary to occlude incompetent superficial veins depends on the circumference of the limb and is highly variable. They concluded that tourniquets
804
JOURNAL OF VASCULAR SURGERY November 1993
van Bemmelen et al.
100-
- ulcer • dermatitis o varicose veins
80" 8 8 8
/,m..
o~
60
.
tL
UJ
40.
:; A
8
-
!
20. 0
I
obstructed
I
non obstructed
Fig. 8. Scattergram demonstrates relationship between EF (%) and deep venous obstruction detected by duplex scanning. are a source of error because they often fail to exclude superficial reflux. Because of these observations and because the results of use of tourniquets in conjunction with photoplethysmography have proved to be disappointing, 22 we did not attempt to use tourniquets in this study. Duplex measurements of venous diameter were not significantly correlated with clinical presentation. In this respect, they were somewhat less discriminatory than the APG. Nonetheless, the mean diameter of reservoir veins was greatest (7.8 --- 5.2 mm) and the mean VFI was highest (7.7 +_ 4.6 ml/sec) in legs with the Society for Vascular Surgery and the North American Chapter of the International Society for Cardiovascular Surgery class 3 dermatitis. Legs with ulceration had reservoir vein diameters of 4.9 +__2.1 mm, which corresponded to a mean VFI of 5.4 __ 3.8 ml/sec. The smallest reservoir vein diameters (3.2 ___ 3.2 mm) and the lowest mean VFI (2.6 ___ 1.7 ml/sec) were found in class 1 limbs with uncomplicated varicose veins. It seems reasonable to assume that calf W and the rate of venous reflux (VFI) should be related to the presence of large incompetent conduit and lower leg reservoir veins. This assumption is supported by the statistically significant (but mostly rather poor) correlations that were observed between W and calf vein (reservoir vein) diameters, VV and VFI, VFI and conduit vein diameters, VFI and reservoir vein diameters, and VFI and the number of incompetent venous segments. O f the five limbs with a VFI equal to or greater than 10 ml/sec, all had venous diameters greater than 6.0 mm in either the calf or conduit veins, and four had venous diameters greater than 6.0 mm in both areas, suggesting that the presence of both incompetent conduit and lower-leg reservoir
veins of these dimensions are required to attain a VFI in excess of 10 ml/sec. Many of the ulcerated legs had VFIs in the intermediate (near normal) 2 to 5 ml/sec range (Fig. 1). Values in this range can occur when a large incompetent conduit-vein is present at the knee level but the calf reservoir is small. Likewise, if the reflux conduit veins are small and the calf reservoir is large, the VFI can be in the 2 to 5 ml/sec range. VFIs in the intermediate range also occurred in the absence of any detectable incompetent veins crossing the knee level-in other words, in the absence of any reflux conduit. This raises the question whether an increased calf volume capacity alone is sufficient to increase VFI over normal levels. The VFI may also be affected by variations in the rate of arterial inflow. 23 Although arteriolar constriction reduces blood flow in dependent legs (the venoarterial reflex),24 we have observed that "resting" blood flow in the popliteal artery may be transiently increased several fold when subjects assume an erect position, largely as a result of an elevated diastolic velocity (Fig. 9). Because the APG test involves emptying the calf veins by leg elevation before standing, the venous pressure approaches zero. The arterial pressure, on the other hand, increases commensurate with the length of the hydrostatic column to the heart. This combination of increased arterial pressure and (temporarily) reduced venous pressure augments the pressure gradient across the resistance vessels and, in so doing, increases the rate of arterial blood flow. The increased flow occurs during the portion of the refilling curve used to calculate the VFI. Therefore, it is possible that the increase in arterial filling might be prolonged when a larger than normal, relatively empty, low pressure venous reser-
JOURNALOF VASCULARSURGERY Volume 18, Number 5
v a n B e m m e l e n et al.
805
Fig. 9. A, Resting popliteal artery flow-velocity signal in normal subject in supine position with elevation of leg. Note absence of antegrade diastofic flow and large retrograde flow component (Peak systolic velocity = 72 cm/sec and peak reverse flow velocity = - 37 cm/sec). B, Same artery immediately after assuming erect position, without weight-bearing. Note increase in antegrade diastolic flow. In absence of venous valvular reflux, increased arterial volume-flow is related to rate of filling of lower leg veins (Peak systolic velocity -- 67 cm/sec and peak diastolic velocity = 32 cm/sec).
voir is present, regardless of valve function at the knee level. Although measurement of blood flow by venous occlusion plethysmography has been recommended as a method for assessing the contribution of arterial inflow to the rate of venous filling,23 it is by no means certain that flow rates measured in the supine position are predictive of those that occur with standing. Total limb blood flow may decrease by 50% as a result of the venoarterial reflex24 or increase
(as we have observed) owing to changes in the arteriovenous pressure gradient. In severely ischemic limbs with an impaired venoarterial reflex, blood flow may increase by as much as 40% on standing. 24 Because of the variability of the blood flow response to position change, we made no attempt to estimate flow rates plethysmographically. The incidence of significant arterial disease is often underestimated in patients with chronic ulcers. 2s27 H o w arterial obstruction might affect
806
van Bemmelen et al.
orthostatic changes in blood flow is difficult to predict in any given subject, but it may be responsible for some of the variation in VFI reported in the literature. Because none of the patients in our study had overt arterial disease, this factor could not be responsible for the disparity in APG and duplex results that we observed. An unresolved issue is whether the volume of venous reflux has major clinical significance. According to Bergan's study, 2s which focused on the greater saphenous vein, there is a disappointing lack of correlation between the quantity of reflux as measured by duplex scanning and clinical classification of disease severity. This is not in accordance with the results ofVasdekis et al., 14which indicated that reflux greater than 10 ml/second in incompetent veins was associated with a high incidence of skin changes (66%), regardless of whether the reflux was in the deep or superficial system. Most of the patients in the study by Vasdekis et al. 14 had isolated superficial reflux, similar to Bergan's2s series. Air plethysmographic studies, which measure reflux to the entire calf, emphasize reflux quantity as an important parameter of venous disease. ~ Results of the present study suggest that duplex scanning and air plethysmography may not measure the same reflux (only 4 of 22 limbs with class 3 disease had reflux greater than 10 ml/sec). This might be explained by the effect of variable arterial inflow on plethysmographic measurements and by the disparate nature of the venous structures being interrogated. Calf VV and its reduction on muscle contraction (EF) are of interest to distinguish between postthrombotic calf pumps, with obliterated veins that have a small volume, and varicose disease with an increase in the size of the venous reservoir below the knee. In our study, EF was often reduced below 60% by an increase in the denominator (e.g. venous volume increased in limbs with superficial varicosities), rather than by a reduction in the numerator (stroke volume). In legs with numerous large functional collateral vessels, the EF was in the 40% to 60% range, even in the presence of an obstructed deep venous outflow (Fig. 8). Because EFs vary widely in limbs with nonobstructed outflow, a low EF does not predict deep obstruction; but EFs greater than 60% were not observed in legs with deep venous obstruction, All legs with an obstructed venous outflow had severe manifestations of chronic venous insufficiency, however. In summary, the APG parameters measured in this study did not effectively distinguish limbs with chronic venous insufficiencyfrom limbs with uncom-
JOURNAL OF VASCULARSURGERY November 1993
plicated varicose veins; nor did they reliably predict the presence or size of incompetent reflux conduit veins or reflux-reservoir veins. In part, this may be related to orthostatic variations in arterial inflow. For the treatment of patients with chronic venous insufficiency, a noninvasive method that is independent of arterial blood flow, that is not affected by arterial disease, and that does not rely on the use of tourniquets for the localization of incompetent veins is desirable. Duplex scanning findings are pertinent to the selection of patients for surgical therapy and for ascertaining what operation should be performed. For these reasons, duplex ultrasonography is unlikely to be replaced by functional measurements. Measurements of VV reflux by APG and the effect of muscle contraction continue to be of interest, however, to help solve therapeutic dilemmas, which occur when superficial incompetence is complicated by moderate deep venous obstruction. APG results also provide functional information that may be important in assessing the results of therapy. For the comprehensive workup of patients with dermatitis and ulceration, duplex scanning, air plethysmography, and arterial evaluation are all important. We thank Jeffrey L. Collins, Dianne L. Raymond, RVT, and Paula PetriUi for their technical assistance. REFERENCES 1. Hojensgard IC, Smrup H. Venous pressure in pr~ary and post-thrombotic varicose veins. Study of statics and dynamics of venous system of lower extremities under pathologic conditions. Acta Chir Scand 1949;99:133-53. 2. Abramowitz HB, Queral LA, Flinn WR, et al. The use of photoplethysmography in the assessment of venous insufficiency: a comparison to venous pressure measurements. Surgery 1979;86:434-41. 3. Allan 1C. Volume changes in the lower limb in response to postural alterations and muscular exercise. S. Afr J Surg 1964;2:75-90. 4. Christopoulos D, Nicolaides, Szendro G. Venous reflux: quantification and correlation with the clinical severity of chronic venous disease. Br J Surg 1988;75:352-6. 5. Christopoulos D, Nicolaides AN. Noninvasive diagnosis and quantitation of popliteal reflux in the swollen and ulcerated leg. J Cardiovasc Surg 1988;29:535-9. 6. Christopoulos D, Nicolaides AN, Cook A, Irvine A, Galloway JMD, Wilkinson A. Pathogenesis of venous ulceration in relation to the calf muscle pump function. Surgery 1989;106: 829-35. 7. Christopoulos D, Nicolaides AN, Galloway JMD, Wilkinson A. Objective noninvasiveevaluation of venous surgical results. J VASC SUP,G 1988;8:683-7. 8. Christopoulos DG, Nicolaides AN, Szendro G, Irvine AT, Bull M, Eastcott HHG. Air-plethysmography and the effect of elastic compression on venous hemodynamics of the leg. J V^sc SURG 1987;5:148-59.
JOURNAL OF VASCULARSURGERY Volume 18, Number 5
9. Christopoulos D, Nicolaides AN, Belcaro G, Duffy P. The effect of elastic compression on calf mnsde pump function. Phlebology 1990;5:13-9. 10. Welkie JF, Comerota AJ, Kerr RP, Katz ML, Jayheimer EC, Brigham RA. The hemodynamics of venous ulceration. Ann Vasc Surg 1992;6:1-4. 1I. Welkie1F, Kerr RP, Katz ML, Comerota AJ. Can noninvasive venous volume determinations accuratelypredict ambulatory venous pressure? J Vasc Technology 1991;15:186-90. 12. Cordts PR, Hartono C, LaMorte WW, Menzoian JO. Physiologic similarities between extremities with varicose veins and with chronic insufficiencyutilizing air plethysmography. Am J Surg 1992;164:260-4. 13. van Bemmelen PS, Bedford G, Beach K, Strandness DE. Quantitative segmental evaluation of venous valvular reflux with duplex ultrasound scanning. J Vasc Surg 1989;10:42531. 14. Vasdekis SN, Clarke GH, Nicolaides AN. Quantification of venous: reflux by means of duplex scanning. J VAsc SURG 1989;10:670-7. 15. Porter JM, Rutherford RB, Clagett GP, et al. Reporting standards in venous disease. J VASC St3RG 1988;8:172-81. 16. Nicolaides AN, Sumner DS. Investigation of patients with deep venous thrombosis and chronic venous insufficiency. London: Med-Orion Publishing Co., 1991:1-73. 17. CzeredarzukM, Branas CC, Weingarten MS. Duplex imaging and distal cuff deflation to measure venous reflux time. J Vasc Technol 1992;15:284-7. 18. van Bemmelen PS, Beach K, Bedford G, Strandness DE. The mechanism of venous valve closure. Arch Surg 1990;125: 617-9. 19. van Bemmelen PS, Bedford G, Beach K, Strandness DE Jr. Status of the valves in the superficialand deep venous system in chronic venous disease. Surgery 1991;109:730-4.
van Bemmelen et al.
807
20. van Bemmelen PS, Bedford G, Beach K, Strandness DE. Functional status of the deep venous system after an episode of deep venous thrombosis. Ann Vase Surg 1990;5: 455-9. 21. McMullin GM, Coleridge Smith PD, Scurr JH. A study of tourniquets in the investigation of venous insufficiency. Phlebology 1991;6:133-9. 22. van Bemmelen PS, van Ramshorst B, Eikelboom B. Photoplethysmography reexamined: lack of correlation with duplex scanning. Surgery 1992;112:544-8. 23. Pierce EC II, Chiang K, SchanzerH. Volume tests for chronic venous insufficiency: an appraisal. Surgery 1991;109:56774. 24. Morgan RH, Psaila JV, Stone J, Carolan G, Woodcock JP. Postural changes in femoral artery blood flow in normal subjects, patients with peripheral vascular occlusive disease and patients undergoing lumbar sympathectomy, measured by duplex ultrasound flowmetry. Eur J Vase Surg 1992;6: 408-15. 25. Cornwall JV, Dore CJ, Lewis JD. Leg ulcers: epidemiology and aetiology. Br J Surg 1986;73:693-6. 26. CallamMJ, Harper DR, Dale JJ, Ruckley CV. Arterial disease in chronic leg ulceration; an underestimated hazard? Lothian and Forth Valley leg ulcer study. BMJ 1987;294:929-31. 27. Nelzen O, Bergqvist D, Lindhagen A. Leg ulcer etiology-a cross sectional population study. }"Vase SuRG 1991;14:55764. 28. Bergan JJ. Clinical application of duplex testing in treatment of primary venous stasis,varicoseveins. In: van BemmelenPS, Bergan JJ, eds. Quantitative measurement of venous incompetence. Austin: RG Landis, 1992:78-104.
Submitted April 7, 1993; accepted July 21, 1993.