© 2002 by the American Society for Dermatologic Surgery, Inc. • Published by Blackwell Publishing, Inc.ISSN: 1076-0512/02/$15.00/0 • Dermatol Surg 2002;28:937–941

Quantitative Venous Severity Scoring Using the Venous Arterial Flow Index by Duplex Sonography

Birgit Kahle, MD, Felix Hennies, MD, Steffen Hummel, MD, and Detlef Petzoldt, MD

Division of Phlebology, Phlebological and Dermatological Surgery, Dermatological Medical Center,

University of Heidelberg, Heidelberg, Germany

background.

Volume flow, as the product of the mean bloodflow velocity by the cross sectional area, means an interestinghemodynamic pattern that can be calculated by duplex.

objective.

To quantify the severity of venous insufficiency us-ing the correlation between the volume flow in the commonfemoral vein (VFV) and artery (VFA), called the venous arterialflow index (VAFI).

methods.

A total of 163 consecutive patients were included:46 patients with postthrombotic syndrome (PTS, group 1), 38patients with complete varicosity of the greater saphenous vein(group 2), 40 patients with only segmental or truncal varicosity(group 3), and 39 patients with competent veins (group 4). Un-der standardized conditions, duplex sonography was performedto calculate volume flow in the common femoral vein and ar-

tery as a product of mean blood flow velocity (

v

m

) and precisediameter (

d

�

2

�

r

) of the vessel due to the formula VF

�

v

m

��

r

2

(L/min). Division of the venous and arterial volume flowdata calculated the VAFI.

results.

Significant differentiation of VFV (

P

�

.001) andVAFI (

P

�

.0001) between varicose veins and healthy limbswere found. In PTS the mean VFV was 0.50 L/min and themean VAFI was 1.465. In the complete varicosity group, meanVFV was 0.46 L/min and mean VAFI was 1.48. In group 3, themean VFV was 0.41 L/min and the mean VAFI was 1.31. Inhealthy persons, mean VFV was 0.36 L/min and mean VAFIwas 0.87.

conclusion.

The VAFI can be used to quantify the hemody-namic severity in venous insufficiency.

B. KAHLE, MD, F. HENNIES, MD, S. HUMMEL, MD, AND D. PETZOLDT, MD HAVE INDICATED NO SIGNIFICANT INTEREST WITH COMMERCIAL SUPPORTERS.

QUANTITATIVE VENOUS severity scoring is neededfor both the planning of therapeutic procedures and out-come assessments.

1

In regard to venous hypertension asthe common cause for development of severe venousdisease, grading according to hemodynamic patterns isof interest. Numerous data on venous hemodynamicscan be obtained using plethysmographic methods orambulatory venous pressure measurement, which is ac-cepted as a standard test of venous insufficiency.

2–6

Butthe extent and restricted repeatability are limiting fac-tors. Photoplethysmography, strain gauge, or air pleth-ysmography only allow an approximate calculation ofthe competence of the venous system

7–9

because ofwidely dispersed standard and pathologic values.

6,10

Duplex sonography allows hemodynamic evaluation ofvenous insufficiency.

11–18

The common femoral vein(CFV) can be considered as representative of the totalvenous cross section of the limb. Therefore the hemody-namic evaluation of the CFV, especially the calculation ofvenous flow volume (VFV) as a product of mean bloodflow velocity and the cross-sectional area of the vessel

19

is

of great interest. Significantly increased VFVs (

P

�

.005)were found in varicose veins compared to competent veinsaccording to the clinical grade of venous insufficiency.

20,21

Because venous blood flow is considered to be in-fluenced by arterial blood flow, we correlated VFVwith the volume flow in the common femoral artery(VFA). The relation was expressed as a quotient ofvenous and arterial volume flow data in the commonfemoral vessels and is called the venous arterial flowindex (VAFI). The following study examined whetherthe calculation of VAFI is a suitable method to quan-tify the hemodynamic situation in patients with super-ficial and deep venous incompetence compared to pa-tients without venous insufficiency.

Patients and Methods

The experimental protocol and informed consent were ap-proved by the institutional review board (Committee of Med-ical Ethics of the University Hospital of Heidelberg). All in-cluded patients gave informed consent to the examination.

Patients

A total of 163 consecutive patients of the department ofdermatology of the University Hospital of Heidelberg re-

Address correspondence and reprint requests to: Birgit Kahle, MD,University Hospital of Heidelberg, Voßstr. 2, D-69115 Heidelberg,Germany, or e-mail: Birgit_Kahle@med.uni-heidelberg.de.

938

kahle et al.: quantification of venous hemodynamics

Dermatol Surg 28:10:October 2002

ferred to the phlebology unit were included in the study.Only the hemodynamic data of one limb in each patientwere analyzed. Patients with peripheral arterial occlusivedisease, identified based on an ankle-brachial pressure in-dex of less than 0.8, were excluded from the study. A totalof 124 patients (124 limbs) with venous insufficiency and39 patients (39 limbs) of the dermatology departmentwithout any venous problems were included. These healthyvolunteers were referred to exclude a venous cause of theirinflammatory skin disorders in the lower limbs (dermatitisor psoriasis).

On the basis of Doppler and duplex ultrasonic examina-tion and photoplethysmographic measurements the subjectswere classified into four groups:

Group 1: 46 patients (mean age 62.85 years) with post-thrombotic syndrome. The patients of this group pre-sented incompetence of the saphenofemoral junctionwith secondary varicose veins combined with obstruc-tion and/or valvular incompetence in the deep veins(C

4,5,6

E

s

A

s,d,p

P

R

,

0

).Group 2: 38 patients (mean age 54.56 years) with complete

varicosity of the greater saphenous vein and competentdeep veins (C

4,5

E

p

A

s

P

R

).Group 3: 40 patients (mean age 60.0 years) with segmental su-

perficial varicosity or an incomplete form of varicosity ofthe greater saphenous vein with a competent saphenofemo-ral junction and competent deep veins (C

2,3

E

p

A

s

P

R

).Group 4: 39 patients (mean age 42.59 years) with skin disor-

ders but without any venous disorder and competent su-perficial and deep venous systems (C

0

).

Duplex Sonography

A color-coded duplex scanner with a 5–7.5 MHz transducerwas used for this study. Duplex sonography was performedunder standardized conditions. The common femoral veinproximal to the saphenofemoral junction and common fem-oral artery proximal to the bifurcation were insonated withthe patient in a relaxed horizontal supine position with theupper body at a 45-degree angle. In both vessels the precisediameter was measured in the cross section. In the longitudi-nal section the mean blood flow velocity in the vessels wasdetermined. Both VFV and VFA were calculated by themean blood flow velocity (

v

m

) and the diameter (

d

�

2

�

r

) ofthe vessels with the formula VF

�

v

m

�

�

r.

2

Division ofVFV and VFA calculates the VAFI (Figures 1 and 2).

VAFI measurement was performed at a constant ambienttemperature. Before the examination the patients had a 30-minute waiting period to avoid the influences of recent exer-cise on venous and arterial flow.

Statistics

Statistical analysis was performed using Student’s

t

-test forunpaired random samples. Values are expressed as mean

�

standard deviation (SD). A significant difference was re-corded when

P

�

.05.

Results

General Results

In all groups the calculated data of VFV, VFA, andVAFI were normally distributed.

Volume Flow in the Common Femoral Vein (VFV)

The highest values were measured in postthromboticsyndrome (PTS) (group 1) followed by primary vari-

Figure 1. Duplex sonography determines the venous volume flowin the common femoral vein (VFV) proximal to the saphenofemo-ral junction in a patient with primary varicosity of the greatersaphenous vein. VFV � 0.39 L/min.

Figure 2. Duplex sonography in the same patient as in Figure 1 de-termines the arterial volume flow (0.29 L/min) in the commonfemoral artery (VFA). The quotient of VFV and VFA determinesthe VAFI. In this patient VAFI � 1.34, showing extensive varicosityof the greater saphenous vein.

Dermatol Surg 28:10:October 2002

kahle et al.: quantification of venous hemodynamics

939

cosity of the greater saphenous vein (GSV) with com-petent deep veins (group 2), segmental varicose veins(group 3), and healthy veins (group 4) (Figure 3).There was a significant differentiation concerning theheight of VFV between all varicose groups (groups 1,2, and 3) and group 4 (

P

�

.05) and also between PTS(group 1) or extent primary varicose veins (group 2)and tributaries (group 3).

The levels of significance were

P

�

.001 betweengroups 1 and 2 compared to group 4;

P

�

.004 be-tween group 1 and group 3;

P

�

.05 between group 2and group 3. Only between groups 1 and 2 could no sig-nificant difference in VFV be found (

P

�

.25) (Table 1).In PTS the mean VFV was 0.505 L/min (0.22–0.92

L/min) and the SD was 0.15. In group 2 the meanVFV was 0.467 L/min (0.22–0.87 L/min)

�

0.14 SD.In segmental varicosities (group 3) the mean VFV was0.415 L/min (0.18–0.69 L/min)

�

0.11 SD. In healthyveins the mean VFV was 0.365 L/min (0.17–0.78L/min)

�

0.12 SD (Table 2 and Figure 3).

Volume Flow in the Common Femoral Artery (VFA)

No significant difference could be found between thethree varicose groups concerning the value of VFA.But between varicose veins (groups 1, 2, and 3) andgroup 4 (healthy limbs) a significant difference (

P

�

.02) (Table 2) was noted.In group 1 the mean VFA of 0.352 L/min (0.17–

0.69 L/min) and SD of 0.11 was found. In group 2 themean VFA was 0.32 L/min (0.17–0.49 L/min) with aSD of 0.09. In group 3 the mean VFA was 0.33 L/min(0.14–0.58 L/min) with a SD of 0.08. In healthy limbsthe mean VFA was 0.42 L/min with a SD of 0.13 (Ta-ble 2 and Figure 3).

VAFI

In every group there was a difference between the levelof VFV and VFA. In healthy limbs (group 4) VFA wasincreased compared to VFV, but in incompetent veins(groups 1, 2, and 3) that relation was reversed. So incompetent veins the VAFI was lower than 1.0 and invaricose veins the VAFI was higher than 1.2 (Figure4). Each of the “varicose groups” showed significantdifferences in VAFI (

P

�

.0001) compared to thehealthy group (Tables 2 and 3). Between the segmen-tal form of varicose veins (group 3, C

2,3

E

p

A

s

P

R

) andthe groups with more extensive varicosities (C

4–6

group 1 and 2) significant differences in VAFI werefound (

P

�

.018). Between varicosity of the greatersaphenous vein and secondary varicose veins in thepostthrombotic group, no significant difference couldbe found (

P

�

.67).In PTS (group 1) the mean VAFI was 1.461

�

0.26(1.0–2.04). In group 2 the mean VAFI was 1.48(0.85–2.0) with a SD of 0.27. In group 3 the meanVAFI was 1.31 (0.57–1.88) with a SD of 0.30. Inhealthy limbs the VAFI ranged from 0.57 to 1.14. Themean was 0.87 with a SD of 0.13 (Table 3 and Figure 4).

Discussion

Many investigations have demonstrated the use of du-plex sonography for the hemodynamic assessment ofvenous diseases.

8,12,14,20–23

Volume flow as the productof mean blood flow velocity and the cross sectionalarea of a blood vessel in a certain part of the vascularsystem characterizes the hemodynamic situation in thissegment and can be calculated by duplex sonogra-phy.

19–21

Previous investigations have demonstratedsignificantly increased VFV (

P

�

.005) in primary var-icose veins according to the length of reflux.

20

Theseresults were confirmed in our study (Figure 3 and Ta-ble 1) which presented significant differences in VFVbetween varicose and healthy veins (

P

�

.05). The re-sults related to the increased cross section of the CFVaccording to the increased total venous cross sectionin limbs with varicose veins. Under the assumptionthat venous hemodynamics are influenced by arterialblood flow characteristics, the relation between ve-nous and arterial volume flow as a quotient of VFV

Figure 3. Mean volume flow in the common femoral vein (VFV,dark columns) and in the common femoral artery (VFA, light col-umns) in postthrombotic syndrome (group 1), extensive varicosityof the greater saphenous vein (group 2), segmental superficialvaricose veins (group 3), and healthy veins (group 4). Significancelevels (P value) of varicose veins compared to healthy veins.

Table 1. P Values of the Differences in VFV

Group 1 Group 2 Group 3 Group 4

Group 1 — .25 n.s. �.004 �.001Group 2 .25 n.s. — �.05 �.001Group 3 �.004 �.05 — �.05Group 4 �.001 �.001 �.05 —

940 kahle et al.: quantification of venous hemodynamics Dermatol Surg 28:10:October 2002

and VFA was calculated as a quantitative hemody-namic pattern for interindividual comparison. The re-sults of the present study confirmed this ratio (theVAFI) as a suitable means of grading the severity ofvenous insufficiency.

In healthy veins, VAFI is less than 1.0 (mean 0.87 �0.13), in contrast to varicose veins with values greaterthan 1.2 (P � .001), in accordance with the clinicalextent of the venous insufficiency. In limbs with inflam-matory skin disorders, such as eczema or psoriasis, butcompetent veins, in contrast to a previous investiga-tion,24 increased VFV and VFA data were measured buttheir quotient (VAFI) was always lower than 1.0 (0.87 �0.13 SD), confirming values recorded earlier (0.82 �0.14).24 So we think that VAFIs less than 1.0 in sub-jects without arterial occlusive disease represent venousblood flow not obstructed or disturbed by valvular in-competence. A VAFI less than 1.0 does not mean thatthe arterial inflow would be higher than the venousoutflow. Volume flow in blood vessels describes onlythe examined segment. In our opinion, the point ofequivalence between arterial inflow and venous out-flow is not located at a horizontal level of the commonfemoral vein and artery. In healthy limbs this pointseems to be located proximal to the region of our du-plex sonographic assessment, in contrast to varicoseveins where it is supposed to be located distally. For

the duplex sonographic examination and calculationof hemodynamic data, both common femoral vesselsare optimal because of their anatomic location. Signifi-cantly increased VAFI was found in each group withincompetent or insufficient veins compared to compe-tent veins (P � .0001) (Figure 4 and Tables 2 and 3).

The highest VAFI values were measured in subjectswith extensive varicosity of the greater saphenous veinwithout affection of the deep vein system (C4,5EpAsPR,group 2). Mean VAFI in this group was 1.48 � 0.27SD. In limbs with PTS and secondary varicose veins(C4,5,6EsAs,d,pPR,0, group 1), mean VAFI was 1.46 � 0.26SD. The decreased level of VAFI in PTS compared to pri-mary varicose veins (group 2) could be interpreted as apattern of obstruction in the deep veins. It can be sup-posed that low VAFI values in extensive secondary vari-cose veins represent relevant obstruction of the deep veinsinsufficient to be compensated by superficial veins. Thedifference between group 1 and 2 was not significant(P � .67). This could be referred to as the main hemody-namic situation of the deep vein system in group 1, wherereflux in the deep veins was the predominant symptom incontrast to obstruction. Further studies are needed to dis-tinguish venous and arterial volume flow in postthrom-botic syndrome according to the extent of deep vein ob-struction and/or reflux. In tributaries (C2,3EpAsPR, group3), mean VAFI was 1.31 � 0.3 (Figure 4 and Tables 2and 3). The difference between the VAFI of group 3 andthe groups with more severe varicose veins (groups 1 and2) was significant (P � .018 and P � .011).

Our results suggest that VAFI or the ratio of arterialto venous volume flow measured by duplex sonographyis a significant indicator of severity of venous disease.The calculation of VAFI, which is relatively easy toperform, enables the noninvasive assessment of the he-modynamic relevance of venous diseases and the out-

Table 2. Mean VFV, VFA, and VAFI in the Four Groups With Standard Deviation

VFV(L/min) P value

VFA (L/min) P value

VAFI � VFV/VFA P value

Group 1 .505�/�.15 � .001 .352 �/� .11 � .003 1.46 �/� .26 � .00001Group 2 .467�/�.14 � .001 .32 �/� .09 � .0003 1.48 �/� .27 � .00001Group 3 .415�/�.11 � .05 .33 �/� .08 � .02 1.31 �/� .30 � .00001Group 4 .365�/�.12 .42 �/� .13 0.87 �/� .13

Significance levels (P value) between varicose veins compared to healthy veins.

Figure 4. Mean VFV (black columns) and VFA (light columns) withthe mean VAFI (gray columns) in complete varicosity of thegreater saphenous vein (group 1), segmental superficial varicoseveins (group 2), complete varicosity of the greater saphenous veinwith deep venous insufficiency (group 3), and healthy veins(group 4). Standard deviation of VAFI is noted as black lines.

Table 3. P Values of the Differences in VAFI

Group 1 Group 2 Group 3 Group 4

Group 1 — .6771 � .018 � .0001Group 2 .6771 — � .011 � .0001Group 3 � .018 � .011 — � .0001Group 4 � .0001 � .0001 � .0001 —

Dermatol Surg 28:10:October 2002 kahle et al.: quantification of venous hemodynamics 941

come measurement of any therapeutic procedure. Incomparison with plethysmographic methods, VAFImeasurement is significantly less time consuming andcan be repeated as often as necessary without discom-fort for the patient. In an actual investigation we ex-amine whether VAFI data correlate with strain gaugeplethysmography-derived patterns. The first results dem-onstrate a positive correlation of VAFI and venous ca-pacity and venous outflow.25

The results suggest further that in varicose veins thelevel of equivalence between arterial inflow and venousoutflow is distally displaced according to the increase ofthe total venous cross section and the total length ofvenous blood flow in the venous system. In extensivevaricosity of the greater saphenous vein (C4–6), values ofmore than 1.3 can be expected, with indices less than1.0 in healthy limbs. In regard to the length of insuffi-cient veins and the total venous cross section of thelimb as relevant patterns for the value of VAFI, it is ofgreat interest whether VAFI decreases according venoussurgery. In a few subjects with primary varicose veinswe could object a decrease of VAFI to levels, as theywere found in healthy subjects after ligation of saphe-nofemoral junction and stripping of the greater saphen-ous vein.25 Further investigations have to object the in-fluence of therapeutic procedures like venous surgery orcompression therapy on the VAFI.

In conclusion, we think that the VAFI is a valid meansof quantifying the hemodynamic severity in venous in-sufficiency that could allow the quantification of theoutcome of therapeutic procedures. The sonographiccalculation of VAFI must be performed accurately inconjunction with an adequate examination of the ar-teries. Further studies are needed to determine the in-fluence of peripheral arterial occlusive disease or ob-structive post-thrombotic disorders on the index.

References1. Rutherford RB, Padberg FT, Comerota AJ, Kistner RL, Meissner

MH, Moneta GL. Venous severity scoring: an adjunct to venousoutcome assessment. J Vasc Surg 2000;31:1307–12.

2. Payne SPK, London NJM, Newland CJ, Trush AJ, Barrie WW, BellRF. Ambulatory venous pressure: correlation and role in identifyingsurgically correctable disease. Eur J Endovasc Surg 1996;11:195–200.

3. Partsch H. Zur Pathogenese des venösen Ulcus cruris. Hautarzt1985;36:196–202.

4. Nicolaides AN, Hussein MK, Szendro G, Christopoulos D, Vas-dekis S, Clarke H. The relation of venous ulceration with ambula-tory venous pressure measurements. J Vasc Surg 1993;17:414–9.

5. Lees TA, Lambert D. Patterns of venous reflux in limbs with skinchanges associated with chronic venous insufficiency. Br J Surg1993;80:725–8.

6. Payne SPK, Thrush AJ, London NJM, Bell PRF, Barrie WW.

Venous assessment using air plethysmography: a comparison withclinical examination, ambulatory venous pressure measurementand duplex scanning. Br J Surg 1993;80:967–70.

7. Kerner J, Schultz-Ehrenburg U, Lechner W. Quantitative Photo-plethysmographie bei gesunden Erwachsenen, Kindern und Schwan-geren und bei Varizenpatienten. Phlebologie 1992;21:134–9.

8. Stemmer R, Blazek V. Digitale Photoplethysmographie (D-PPG) inder Phlebologie. Methodik und klinische Anwendung. Phlebologie1990;43:551–2.

9. Wienert V. Anwendungsfehler und Fehlinterpretationen bei der Licht-reflexionsrheographie. Phlebologie 1991;21:126–30.

10. Lees TA, Lambert D. A comparative study of air plethysmographyand Doppler colour flow imaging with ambulatory venous pressuremeasurements in the diagnosis of venous reflux in the lower limb.In: Raymont-Martimbeau P, Prescott RM, eds. Phlebology 92.Paris: John Libbey Eurotext, 1992:594–6.

11. Araki CT, Back TL, Padberg FT, Thompson PN, Duran WN, Hob-son RW. Refinements in the ultrasonic detection of popliteal veinreflux. J Vasc Surg 1993;18:742–8.

12. Bays RA, Healy DA, Atnip RG, Neumeyer M, Thiele BL. Valida-tion of air plethysmography, photoplethysmography and duplexultrasonography in the evaluation of severe venous stasis. J VascSurg 1994;20:721–7.

13. Labropoulos N, Leon M, Nicolaides AN, et al. Venous reflux inpatients with previous deep venous thrombosis: correlation with ul-ceration and other symptoms. J Vasc Surg 1994;20:20–26.

14. Labropoulos N, Mansour MA, Kang SS, Gloviczki P, Baker WH.New insights into perforator vein incompetence. Eur J EndovascSurg 1999;18:228–34.

15. Masuda M, Kistner RL, Eklof B. Prospective study of duplex scan-ning for venous reflux: comparison of Valsalva and pneumatic cufftechniques in reverse Trendelenburg and standing positions. J VascSurg 1994;20:711–20.

16. Yamaki T, Nozaki M, Sasaki K. Quantitative assessment of super-ficial venous insufficiency using duplex ultrasound and air plethys-mography. Dermatol Surg 2000;26:644–8.

17. Sandri JL, Barros FS, Pontes S, Jacques C, Salles-Cunha SX. Diam-eter relationship in perforating veins of patients with varicose veins.J Vasc Surg 1999;30:867–75.

18. Sakurai T, Matsushita M, Nishikimi N, Nimura Y. Hemodynamicassessment of femoropopliteal reflux in patients with primary vari-cose veins. J Vasc Surg 1997;26:260–64.

19. Ruland WO, Borkenhagen N. Technische Grundlagen der Doppler-sonographie. In: Ruland WO, ed. Dopplersonographische Diagnos-tik. Köln: Deutscher Ärzte-Verlag, 1993:37–9.

20. Kahle B, Köhler P, Utermann S. Duplexsonographische Unter-suchungen zum venösen Stromzeitvolume n in der Vena femoralisbei CVI. Phlebologie 1999;28:81–6.

21. Kahle B, Hummel S, Deichmann M, Petzoldt D. Bestimmung desSchweregrads einer venösen. Insuffizienz Mittels Duplexsonogra-phie. Hautarzt 2000;51:70–74.

22. Weingarten MS, Branas CC, Czeredarczuk M, Schmidt JD, Wolf-erth CC. Distribution and quantification of venous reflux in lowerextremity chronic venous stasis disease with duplex scanning. JVasc Surg 1993;18:752–9.

23. Weingarten MS, Czeredarczuk M, Scovell S, Branas CC, MigngnaGM, Wolferth CC Jr. A correlation of air plethysmography andcolor-flow-assisted duplex scanning in the quantification of chronicvenous insufficiency. J Vasc Surg 1996;24:750–54.

24. Kahle B, Hummel S, Kirchner P. Korrelationsindex des Strom-zeitvolume ns in der Arteria und Vena femoralis communis “SV-VA-Index” bei verschiedenen Stadien der Varikosis. Phlebologie2000;29:27–31.

25. Kahle B, Hummel S, Kirchner P. The correlation index of the vol-ume flow in the common femoral vein and artery (VAFI) allows thequantification of the severity of venous insufficiency. Oral presen-tation at the 2nd Meeting of the European Venous Forum, Rome,Italy, September 13–14, 2001.