Journal Logo


Role of Ultrasound in the Assessment of Chronic Venous Insufficiency

Allan, P. L. M.D.

Author Information
  • Free


Objectives: After reading this article and completing the posttest, the learner should be able to:

  • Identify how to perform ultrasound examinations of the lower limb veins to assess chronic venous insufficiency and identify structural and functional abnormalities, which may be associated with CVI.
  • Describe how to elicit venous reflux and how to assess its severity and clinical significance.

Problems associated with chronic venous disease are often given a low priority by clinicians because they are not seen to be particularly important or life-threatening. However, when looked at on a national basis, they are responsible for significant morbidity and expenditure on health care. The National Health Service in the United Kingdom spends approximately £400 to £600 million per year on health care arising from chronic venous disease. 1

In general terms, chronic venous insufficiency (CVI) can be defined as changes developing in the tissues of the lower limb as a result of inadequate function of the veins arising from impaired valvular function. This may be primary failure or may occur as a result of previous deep vein thrombosis/thrombophlebitis affecting the deep and/or superficial veins in the leg. This, in turn, results in impaired venous return and increased hydrostatic pressure on the tissues of the lower limb with associated hypoxic, inflammatory, and toxic damage to the cells.


The clinical assessment of patients with CVI is based on examination of the legs for signs of abnormal venous structures, such as hyphen web and reticular varices, varicosities and varicose veins, edema, and pigmentation. Some information concerning venous function can be obtained using hand-held continuous wave (CW) Doppler, but the lack of precise anatomic information may make this difficult to interpret accurately.

The Widmer Classification 2 was developed in the 1970s as part of a large epidemiologic study performed in Basle, Switzerland and is based on the appearances of the veins and the associated skin changes. The main categories in this classification are shown in Table 1. This classification deals with the appearances but does not take into account the underlying pathophysiologic changes.

Widmer classification of CVI2

The clinical, etiologic, anatomic, and pathophysiologic (more commonly known as CEAP), or Hawaii, classification, 3 has been developed in an attempt to improve the classification of CVI. This classification looks at the clinical, etiologic, anatomic, and pathophysiologic factors producing signs of chronic venous disease; the main features are shown in Table 2. Each of these main categories can be subdivided for greater precision. However, the disadvantage of this classification is that it can result in quite complex classifications such as C 3s , E P , A 1R,2R, 3R, 4R, 17R, 18R , P RO . Although this conveys a lot of information concerning the status of the leg in question, it does not fit well with the clinical categories of CVI on which treatment decisions are based. However, it can be considered a basis from which to go forward, and subsequent versions should be more useful and clinically relevant. 4

CEAP classification of CVI3


A good clinical history and examination will provide much information, but to define the treatment options more accurately, it is usually necessary to find out more about the venous system in the affected leg. There are a variety of techniques available, but over the course of the last 12 to 15 years, Doppler ultrasound has become the mainstay of lower limb venous assessment in patients with CVI. It is noninvasive and can therefore be repeated, if necessary, without discomfort. There are three components to a Doppler examination of the lower limb veins in patients with CVI (Fig. 1): 1) the imaging component provides anatomic information. This allows the identification of the venous segment being examined and will show any structural abnormalities, such as internal echoes in a thrombosed vein, or thickened walls in patients who have had previous episodes of thrombophlebitis; 2) the color Doppler component provides information on the presence, or absence, of blood flow in the relevant segment, which way the blood is moving, and how fast it is moving (Fig. 2); and 3) the spectral Doppler component allows the measurement of the duration of reflux and other calculations relating to the severity of the valvular dysfunction, including the velocity of reflux and an estimate of the volume of reflux. These findings allow the operator to build an overall picture of the structure and function of the venous segments in the leg being examined. One of the disadvantages of ultrasound is the time required to perform an examination in complex cases, this can be quite tiring for some elderly patients.

FIG. 1.
FIG. 1.:
A color duplex examination of an incompetent vein showing the placement of the sample volume within the vein lumen. The reversed direction of flow in the color map and the duration of reversed flow on the spectral trace can be measured against the scale at the bottom of the image or by using the system's calipers.
FIG. 2.
FIG. 2.:
Color Doppler images of the same segment of superficial femoral vein showing flow in opposite directions.

Other techniques are available. Plethysmography provides information about the overall function and efficiency of the veins; data concerning the refilling time and ejection fraction may occasionally be useful, but this technique is rarely performed in our institution. Venography/varicography may also be of value on occasion, particularly for clarifying patterns of venous drainage from the upper thigh to the pelvic and perineal regions.

Finally, it is worth considering the developing role of magnetic resonance imaging. Initially, there were problems with both temporal and spatial resolution when imaging the leg veins. In addition, patients are normally imaged while lying horizontal, thereby reducing the effect of gravity in the production of reflux. However, many of these problems have been solved, or significantly reduced, and magnetic resonance imaging has the potential to provide significant information about the vein walls, venous blood flow, and the amount of edema in the tissues of the leg.


The anatomy of the venous system is more variable than the arterial anatomy. In the leg, the veins are divided into deep and superficial systems (Fig. 3). These are linked by a variable number of perforator veins, which normally carry blood from the superficial to the deep veins (Fig. 4)

FIG. 3.
FIG. 3.:
Diagram of the anterior and posterior aspects of the leg showing the main deep and superficial veins. From: Allan PL, Dubbins PA, Pozniak MA, et al, eds. Clinical Doppler Ultrasound London: Churchill Livingstone, 2000:91, with permission.
FIG. 4.
FIG. 4.:
Diagram of the leg showing the major perforating vein sites. From: Allan PL, Dubbins PA, Pozniak MA, et al., eds. Clinical Doppler Ultrasound London: Churchill Livingstone, 2000:92, with permission.

The deep veins generally correspond with the arterial branches. In the calf, there are usually two, occasionally three, veins running with the three main calf arteries. These join to form the popliteal and superficial femoral veins; in 25% of individuals, there are significant segments of dual popliteal and superficial femoral veins. 5 At the groin, the profunda femoris vein joins the superficial femoral vein to form the common femoral vein, which, in turn, becomes the external iliac vein.

The superficial veins are the long and short saphenous systems. The long saphenous vein drains into the common femoral vein at the groin and the anatomy is fairly standard; although, there may be communications with other veins in the region, even in legs that have not had surgery. There may be a prominent anterior thigh division of the long saphenous vein, which the unwary sonographer, or surgeon, may mistake for the main trunk unless a careful search for this is performed. After surgery, recurrent varicose veins may develop and the pattern of these will depend on the type of operation performed.

In standard anatomy texts, the short saphenous vein drains into the popliteal vein at, or just above, the level of the posterior skin crease of the knee. However, the situation is often more complex than this. Giacomini 6 first described the continuation of the short saphenous vein to connect with upper thigh veins in 1873. More recently, Burihan and Baptista-Silva 7 dissected 200 adult cadaver legs and reported 20 different patterns of termination of the short saphenous vein. In 27.5% of legs, the short saphenous vein terminated in the principal deep vein of the leg (popliteal or lower superficial femoral vein); in 25%, of legs the short saphenous vein, or a branch arising from it, communicated with the long saphenous vein. In the remaining legs, there was a wide variety and combination of communications with other veins, including the deep femoral vein, the midthigh perforator vein, muscular veins, and even the inferior gluteal vein in three legs. The veins draining the gastrocnemius and soleus muscles also join the popliteal vein behind the knee joint; they can be distinguished from the short saphenous vein by their deeper location within the muscles and the presence of the arteries that normally run with these veins, whereas the short saphenous vein does not have a companion artery. The short and long saphenous systems are connected through superficial communicating veins, which may be quite large in patients with CVI.

The perforator veins normally drain blood from the superficial system to the deep system. In patients with CVI, they can become incompetent and enlarged, allowing blood to pass from the deep to the superficial veins. The main groups of perforator veins are shown in Figure 4. The important groups are those in the calf between the posterior division of the long saphenous vein and the deep calf veins, and those in the mid and lower thigh, that connect the long saphenous vein to the superficial femoral vein.


There are several indications for an ultrasound examination of the lower limb veins, but in patients with signs and symptoms of CVI, the main problems for which answers are required include identification of incompetent segments, clarification of uncertain anatomy, and identification of perforator veins. The main indications are given in Table 3.

Indications for ultrasound in the investigation of CVI


The examination of the lower limb veins may be either a general survey of the venous system of one or both lower limbs, or a focused examination to answer a specific question, such as locating the level of the sapheno-popliteal junction, or assessing the pattern of recurrence at the sapheno-femoral junction. A full survey of one or both limbs may require some considerable time to perform, especially in complex cases, whereas a focused examination to answer a specific question can be completed in a relatively short time and can contribute to a “one-stop” venous clinic.

The most useful position for the patient is to be standing upright, when gravity and haemostatic pressure will produce dilatation of the lower limb veins and the maximum stimulus for reflux. It is better if patients stand on a raised plinth and have a frame or handle with which to support them. However, many elderly patients will find this tiring to maintain for a considerable time; in addition, standing still in the often warm and close atmosphere of the ultrasound room can result in vasovagal episodes, which are disconcerting for both the patient and the sonographer. Most departments undertaking these examinations on a regular basis will use a tilting couch, which can be raised to 60° and will give the patient support while allowing reasonable conditions for the detection of reflux.

If a tilting couch is not available, the patient can be asked to sit on a standard couch, with their torso elevated to produce some haemostatic pressure, and the thigh veins are then examined in this position. The patient is then asked to stand for a relatively short period, while the popliteal region is examined and any concerns relating to the thigh veins are clarified. Finally, the patient is asked to sit on the edge of the couch with their legs over the edge while the calf veins are examined.

The first stage of the examination assesses the common femoral vein, the sapheno-femoral junction, the long saphenous vein and any associated perforator veins, and the superficial femoral vein. The long saphenous and superficial femoral vein are conventionally divided into two (upper and lower) or three segments (upper, middle, and lower) for recording findings. The findings in the various segments are recorded as normal, reflux present, occlusion, or evidence of previous thrombosis. It is also necessary to look for any incompetent perforator veins, or superficial collateral veins, that may be connected to the main saphenous trunk. If incompetent perforator veins are demonstrated, their location should be related to anatomic landmarks, such as the groin, patella, or medial malleolus in the calf. The appearance of reflux on moving between two segments requires a careful examination for associated incompetent perforator or collateral veins to be undertaken.

The second stage of the examination assesses the popliteal region and short saphenous vein; the popliteal vein is usually divided into upper and lower segments. The sapheno-popliteal junction must be sought and examined with care, bearing in mind the multitude of possible short saphenous terminations noted earlier. It is useful to relate the level of the saphena-popliteal junction to the posterior skin crease at the knee. The short saphenous vein is then examined down the posterior aspect of the calf and attention is given to any significant perforator veins or superficial collaterals connecting to the long saphenous system.

The final part of the examination is the assessment of the calf veins. The deep veins should be examined to confirm patency and assess competence of the valves. In some patients with significant venous insufficiency, it can be quite difficult to generate cranial flow to test for possible reflux. This is because of the complete incompetence of the valvular apparatus, such that the veins in the calf act as a single container and there is therefore no net forward flow on compression. The long and short saphenous veins are examined and any incompetent perforator veins must be sought, particularly along the medial aspect of the calf. The level of any such vessel is related to either the medial malleolus or the patella. Some patients will have open ulcers on the calf and it will often be necessary to scan the region of the ulcer to get a full picture of the venous anatomy and function. Scanning in the region of the ulcer must be performed with a careful aseptic technique, ensuring that the minimum of contamination occurs. The ulcer itself can be covered with a thin, sterile, plastic membrane, through which the sound is able to pass. It is important that appropriately trained staff are available to reapply compression dressings properly after the examination. At the end of the examination, the findings can usefully be recorded on a diagram of the lower limb veins (Fig. 5), unless the examination is being performed in the immediate preoperative period, when the abnormal segments, perforator veins, or collateral channels can be marked on the skin with an appropriate marker pen.

FIG. 5.
FIG. 5.:
An example of a CVI report form showing the findings in the various vein segments.


The foregoing section deals with the basic technique for the examination of patients with chronic venous disease but does not deal with some important questions, such as how to elicit reflux and how to define significant reflux. The main abnormalities that may be found during an examination of a patient with CVI are given in Table 4.

Abnormalities on ultrasound in patients with CVI

Elicitation of Reflux

There has been much discussion on the best way to elicit reflux reliably. In general clinical practice, many experienced operators use a manual squeeze on the calf veins to promote forward venous flow, using color Doppler to identify incompetent segments. With practice and experience, this is adequate for routine clinical examinations. A more standardized method for provoking reflux is the use of pneumatic cuffs to produce a reproducible stimulus. Cuffs can be used to squeeze the limb below the point of assessment to empty the veins, the cuff is then rapidly deflated and color Doppler is used to assess any reverse venous flow down into the “empty” vein segment. 8 Alternatively, the cuff can be inflated rapidly to replicate, in a more standardized fashion, a hand squeeze. 9

A Valsalva maneuver may also be used to elicit reflux, particularly in the more proximal veins. The sudden increase of intraabdominal pressure will produce cessation of flow in competent veins, but reversal of flow in incompetent veins. There are several problems associated with the Valsalva: firstly, it only results in reversed flow down to the first competent valve; therefore, any incompetent segments below a competent valve will not be identified. Secondly, there will be significant variations in the pressure change generated by different patients. 10 Thirdly, explaining the requirements for the maneuver in terms that patients may understand can be difficult. One way around this last problem is to ask the patient to blow into a high-resistance spirometer circuit, which produces the required increase of abdominal pressure. It is important to realize that the Valsalva maneuver produces reflux through a different mechanism from a distal squeeze or pressure cuff release. 10 The Valsalva maneuver increases the intraabdominal pressure and reverse flow will last as long as a pressure gradient persists across an incompetent valve; in the vertical position, there will also be a component of pressure from the hydrostatic pressure of the blood. When a pressure cuff is released, or the distal limb is squeezed and released, the stimulus for reflux is the negative pressure gradient between the upper, full segment and the lower, empty segment. In addition, there is the hydrostatic pressure from the volume of blood in the vessels above the point of assessment. This pressure difference will only last for as long as it takes blood to redistribute.

A further complication of the Valsalva technique for producing reflux is that the reflux times vary between upright and supine positions because there is no associated increase in venous capacitance in the upright position, whereas this will occur in the supine position, resulting in longer reflux times. 11,12 Reflux times in the common femoral vein in normal patients have been noted to vary from 0.69 (±0.83) seconds on Valsalva in the upright position to 1.77 (± 0.96) seconds in the supine position. 12 Variations in reflux times have also been reported to be associated with body mass index, gender, and familial tendency to varicose veins. 10

Definition of Significant Reflux

There is also much debate on the definition of significant reflux. Should we assess reflux by its duration, its velocity, or its volume? In general clinical practice, reflux is elicited by a distal manual squeeze and release, and the duration of the reflux is measured. A useful, subjective categorization for routine work is given in Table 5. For more objective assessments, some workers suggest that reflux lasting longer than 1 second is significant (but this lowers sensitivity), whereas others will take 0.5 seconds as the cut-off point (which reduces specificity). Evans et al. 13 in the Edinburgh Vein Study compared the value of 0.5 seconds with 1.0 seconds and concluded that calculations of sensitivity and specificity tended to support the use of 0.5 seconds. Van Bemmelen et al. noted that 95% of valve closure times in normal legs were less than 0.5 seconds. 12

Subjective assessment of the severity of reflux

However, there is evidence that the duration of physiological reflux, which occurs as a normal valve is closing, and also the diameter increase in a vein segment can vary between segments in both normal patients and patients with CVI, with the common femoral vein segment reflux time being longer and having a larger percentage increase in diameter on Valsalva than more distal segments. 10 The reflux times for individual segments can be added together to obtain a “total limb reflux time,” which is a measure of the overall severity of reflux in a limb and correlates with the venous filling index measured at air plethysmography. 14 However, the relationship between the duration of reflux, the amount of reflux, and the severity of the changes of CVI is not particularly clear. Rodriguez et al. 15 showed that the duration of reflux did not correlate with the volume of reflux as measured by air plethysmography; although, there was some correlation between total limb reflux times and total limb reflux volume.

The peak velocity of reflux can be measured and this has been related to the severity of reflux. It has also been shown that it is necessary to develop a threshold velocity of reverse flow to close the venous valves. 16 However, a high velocity may be caused by a small volume of blood crossing a partially incompetent valve or a much larger volume with reversed flow down a completely incompetent segment. Measurement of the volume of reflux takes into account both the velocity of the reflux and the diameter of the vein, and it has therefore been suggested that this will provide a more accurate assessment of the severity of reflux, with a reflux volume greater than 10 ml/sec being more frequently associated with skin changes of CVI. 17 A similar assessment can be made by calculating the area under the reflux segment of the spectral Doppler trace, which will be related to the volume of blood in terms of the velocity profile over the course of time and the duration of reflux. 18

Turbulence, particularly in the larger veins, should not be confused with reflux. Turbulence is usually recognizable on color Doppler, appearing as an area of reverse flow occurring simultaneously with forward flow. On spectral Doppler, the same finding of simultaneous forward and reversed flow is seen. Reflux is seen as reversed flow occurring after the period of forward flow (Fig. 6).

FIG. 6.
FIG. 6.:
Turbulence in a popliteal vein. The color Doppler image shows simultaneous forward and reversed flow, with similar findings on the spectral trace below.

Dilatation of Veins

Whereas normally functioning veins may show reflux, 20 in patients with CVI, the affected veins in both the superficial and deep segments have a larger caliber than normal. The calf veins are more variable in size and number, but, even so, an impression of dilatation can be obtained. Often, however, it is difficult to elicit reflux in these veins because the blood is effectively in a large single space and local pressure changes do not produce any significant directional flow. In the superficial veins, transition from an upper segment of normal caliber to a lower dilated segment should prompt a search for incompetent perforator veins.

Abnormal Perforator Veins

The detection of perforator veins has been significantly improved by the advent of color Doppler ultrasound. The detection rates reported are variable and depend on the size of the perforators and diligence with which they are sought by the operator. Careful scanning along the saphenous veins, looking specifically for vessels penetrating the superficial fascia, can be improved by inducing deep venous flow as described previously and by using color Doppler to identify regions of outward flow consistent with incompetent perforator veins (Fig. 7). Perforator veins may take a relatively long and oblique course between the superficial and deep veins, and they should not be confused with muscle branch veins or superficial communicating channels between saphenous branches, neither of which penetrate the superficial fascia. Normally, perforator veins are of narrow caliber and veins >4 mm are strongly associated with incompetence. 20 Any reproducible, outward flow in a perforator vein is abnormal and, if the vein is dilated, likely to be clinically significant. However, there is continuing discussion as to whether there is any value in ligation of incompetent perforator veins with, or without, associated saphenous ligation in terms of the treatment of CVI. 21

FIG. 7.
FIG. 7.:
Outward flow in an incompetent perforator vein situated on the medial aspect of the calf 16 cm above the medial malleolus.

Postthrombotic Changes

Signs of a previous deep vein thrombosis may be identified on ultrasound. These include an irregular venous lumen, thickened vein walls, occluded segments, and collateral venous channels. In addition, reflux can result from damage to the venous valves in the affected segment. The development of CVI can take some years. In one study, 48% of patients had incompetence in previously thrombosed segments when scanned approximately 3 years after the thrombotic episode, increasing to 60% when scanned approximately 7 years after the deep venous thrombosis. 22

Previous Venous Surgery

Patients with CVI may have had one or several previous operations on their leg veins. Often, the exact details will not be available to the sonographer and the possibility of previous surgery should be kept in mind when trying to clarify unusual patterns of venous drainage or reflux. In patients who have had a high tie and resection of the upper few centimeters of the long saphenous vein, three patterns of recurrence may be recognized:23 1) recurrence through collateral channels between the upper end of the long saphenous vein and the region of the sapheno-femoral junction (Fig. 8); 2) drainage through incompetent perforator veins in the thigh; 3) collateral channels connecting with pelvic, perineal, vulval, or other veins around the upper thigh and pelvic region. Patients who have undergone stab avulsions, or sclerotherapy, will have segments of superficial vein irregularity, or occlusion.

FIG. 8.
FIG. 8.:
An area of collateral veins that has developed at the site of surgery to the upper long saphenous vein.


Color Doppler ultrasound has transformed techniques for the assessment of CVI. It can show normal veins, segments of reflux or occlusion, and postthrombotic changes. This information can be used to plan treatment options. However, there is still much work to be performed to clarify those changes that are abnormal and clinically significant. There is a need to further define the nature of clinically significant reflux and identify the patterns of reflux that are associated with the development of CVI. The definition of incompetent perforator veins must also be explored further and their relevance to CVI progression and treatment need to be more clearly understood. At the pathophysiologic level, there is a need to elucidate the processes by which incompetent veins produce the changes of CVI in the tissues of the lower limb. Finally, more work is required to clarify the principles underlying conservative and surgical treatment strategies.


1. Bosanquet N. Costs of venous ulcers: from maintenance therapy to investment programmes. Phlebology 1992; 7(S1):44–6.
2. Widmer LK, Käch K, Veader G, et al. Risk factors for varicosity. In: Widmer LK, ed. Peripheral Venous Disorders. Bern, Switzerland: Hans Huber, 1978:58–66.
3. Porter JM, Moneta GL, International Consensus Committee on Chronic Venous Disease. Reporting standards in venous disease: an update. J Vasc Surg 1995; 21:635–45.
4. Bergan JJ. How should venous disease be classified? Venous Disease: Epidemiology, Management and Delivery of Care. In: Ruckley CV, Fowkes FGR, Bradbury AW, eds. London: Springer-Verlag, 1999:73–9.
5. Screaton NJ, Gillard JH, Berman LH, et al. Duplicated superficial femoral veins: a source of error in the sonographic investigation of deep vein thrombosis. Radiology 1998; 206:397–401.
6. Giacomini C. Osservazioni Aatomische per Service Allo Studio Della Circulazioni Venosa Delle Extremitá Inferiori. Torino, Italy: Tip V Vercillino, 1873.
7. Burihan E, Baptista-Silva JCC. Anatomical study of the small saphenous vein (saphena parva): types of termination. Phlebology 1995; 1(S):57–60.
8. van Bemmelen PS, Bedford G, Beach K, et al. Quantitative segmental evaluation of venous valvular reflux with duplex scanning. J Vasc Surg 1989; 10:425–31.
9. Evans CJ, Leng GC, Stonebridge P, et al. Reproducibility of duplex ultrasound in the measurement of venous reflux. Phlebology 1995; 10:149–54.
10. Jeanneret C, Labs KH, Aschwanden M, et al. Physiological reflux and venous diameter change in the proximal lower limb veins during a standardised Valsalva manoeuvre. Eur J Vasc Endovasc Surg 1999; 17:398–403.
11. Kistner R, Eckloff B. Prospective study of duplex scanning for venous reflux: comparison of Valsalva and pneumatic cuff techniques in the reverse Trendelenberg and standing positions. J Vasc Surg 1994; 20:711–20.
12. van Bemmelen PS, Bedford G, Beach K, et al. Quantitative segmental evaluation of venous valvular reflux with duplex scanning. J Vasc Surg 1989; 10:425–31.
13. Evans CJ, Allan PL, Lee AJ, et al. Prevalence of venous reflux in the general population on duplex scanning: Edinburgh Vein Study. J Vasc Surg 1998; 28:767–76.
14. Weingarten MS, Czeredarczuk M, Scovell S, et al. A correlation of air plethysmoghraphy and color-flow-assisted duplex scanning in the quantification of chronic venous insufficiency. J Vasc Surg 1996; 24:750–4.
15. Rodriguez AA, Whitehead CM, McLaughlin RL, et al. Duplex-derived valve closure times fail to correlate with reflux flow volumes in patients with chronic venous insufficiency. J Vasc Surg 1996; 23:606–10.
16. van Bemmelen P, Beach K, Bedford G, et al. The mechanism of venous valve closure. Its relationship to the velocity of reverse flow. Arch Surg 1990; 125:617–9.
17. Vasdekis SN, Clarke GH, Nicolaides AN. Quantification of venous reflux by means of duplex scanning. J Vasc Surg 1989; 10:670–7.
18. Beckwith Tc, Richardson GD, Sheldon M, et al. A correlation between blood flow volume and ultrasonic Doppler waveforms in the study of valve efficiency. Phlebology 1993; 8:12–6.
19. Lagatolla NRF, Donald A, Lockhart S, et al. Retrograde flow in the deep veins of subjects with normal venous function. Br J Surg 1997; 84:36–39.
20. Phillips GW, Cheng LS. The value of ultrasound in the assessment of incompetent perforating veins. Australas Radiol 1996; 40:15–8.
21. Darke SG. Tailoring surgery to the venous abnormality. In: Ruckley CV, Fowkes FGR, Bradbury AW, eds. Venous Disease: Epidemiology, Management and Delivery of Care. London: Springer, 1999.
22. van Haarst EP, Liasis N, van Ramshorst B, et al. The development of valvular incompetence after deep vein thrombosis: a seven year follow-up study with duplex scanning. Eur J Vasc Endovasc Surg 1996; 12:295–9.
23. Stonebridge PA, Chalmers N, Beggs I, et al. Recurrent varicose veins: a varicographic analysis leading to a new, practical classification. Br J Surg 1995; 82:60–2.

Lower limb; Veins; Venous incompetence; Ultrasound; Doppler ultrasound; Color Doppler

© 2001 Lippincott Williams & Wilkins, Inc.