Database - of - Systematic - Reviews in Varicose Vein

Cochrane Database of Systematic Reviews
Endovenous ablation (radiofrequency and laser) and foam

sclerotherapy versus open surgery for great saphenous vein
varices (Review)
Nesbitt C, Bedenis R, Bhattacharya V, Stansby G
Nesbitt C, Bedenis R, Bhattacharya V, Stansby G.

Endovenous ablation (radiofrequency and laser) and foam sclerotherapy versus open surgery for great saphenous vein varices.
Cochrane Database of Systematic Reviews 2014, Issue 7. Art. No.: CD005624.
DOI: 10.1002/14651858.CD005624.pub3.
www.cochranelibrary.com
Endovenous ablation (radiofrequency and laser) and foam sclerotherapy versus open surgery for great saphenous vein varices (Review)
Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
TABLE OF CONTENTS
HEADER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
ABSTRACT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
PLAIN LANGUAGE SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
BACKGROUND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
OBJECTIVES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
METHODS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Figure 1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Figure 2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Figure 3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
DISCUSSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
AUTHORS’ CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
ACKNOWLEDGEMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
CHARACTERISTICS OF STUDIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
DATA AND ANALYSES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
Analysis 1.1. Comparison 1 Foam sclerotherapy versus surgery, Outcome 1 Recurrence. . . . . . . . . . . 63
Analysis 1.2. Comparison 1 Foam sclerotherapy versus surgery, Outcome 2 Recanalisation. . . . . . . . . . 64
Analysis 1.3. Comparison 1 Foam sclerotherapy versus surgery, Outcome 3 Neovascularisation. . . . . . . . . 65
Analysis 1.4. Comparison 1 Foam sclerotherapy versus surgery, Outcome 4 Technical failure. . . . . . . . . . 65
Analysis 2.1. Comparison 2 Laser ablation versus surgery, Outcome 1 Recurrence. . . . . . . . . . . . . 66
Analysis 2.2. Comparison 2 Laser ablation versus surgery, Outcome 2 Recanalisation. . . . . . . . . . . . 67
Analysis 2.3. Comparison 2 Laser ablation versus surgery, Outcome 3 Neovascularisation. . . . . . . . . . . 68
Analysis 2.4. Comparison 2 Laser ablation versus surgery, Outcome 4 Technical failure. . . . . . . . . . . 69
Analysis 2.5. Comparison 2 Laser ablation versus surgery, Outcome 5 Long term recurrence (≥ 5 years). . . . . . 70
Analysis 2.6. Comparison 2 Laser ablation versus surgery, Outcome 6 Long term recanalisation (≥ 5 years). . . . 70
Analysis 2.7. Comparison 2 Laser ablation versus surgery, Outcome 7 Long term technical failure (≥ 5 years). . . . 71
Analysis 3.1. Comparison 3 Radiofrequency ablation versus surgery, Outcome 1 Recurrence. . . . . . . . . . 71
Analysis 3.2. Comparison 3 Radiofrequency ablation versus surgery, Outcome 2 Recanalisation. . . . . . . . 72
Analysis 3.3. Comparison 3 Radiofrequency ablation versus surgery, Outcome 3 Neovascularisation. . . . . . . 73
Analysis 3.4. Comparison 3 Radiofrequency ablation versus surgery, Outcome 4 Technical failure. . . . . . . . 74
ADDITIONAL TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
APPENDICES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
WHAT’S NEW . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
CONTRIBUTIONS OF AUTHORS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
DECLARATIONS OF INTEREST . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
SOURCES OF SUPPORT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
DIFFERENCES BETWEEN PROTOCOL AND REVIEW . . . . . . . . . . . . . . . . . . . . . 94
INDEX TERMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
Endovenous ablation (radiofrequency and laser) and foam sclerotherapy versus open surgery for great saphenous vein varices (Review) i
[Intervention Review]
Endovenous ablation (radiofrequency and laser) and foam

sclerotherapy versus open surgery for great saphenous vein
varices
Craig Nesbitt1 , Rachel Bedenis2 , Vish Bhattacharya3 , Gerard Stansby4

1 SunderlandRoyal Hospital, Sunderland, UK. 2 Centre for Population Health Sciences, University of Edinburgh, Edinburgh, UK.
3 Department of General and Vascular Surgery, Queen Elizabeth Hospital, Gateshead, UK. 4 Northern Vascular Centre, Freeman
Hospital, Newcastle upon Tyne, UK
Contact address: Gerard Stansby, Northern Vascular Centre, Freeman Hospital, Newcastle upon Tyne, NE7 7DN, UK.
Gerard.Stansby@nuth.nhs.uk.
Editorial group: Cochrane Vascular Group.

Publication status and date: New search for studies and content updated (no change to conclusions), published in Issue 7, 2014.
Citation: Nesbitt C, Bedenis R, Bhattacharya V, Stansby G. Endovenous ablation (radiofrequency and laser) and foam sclerotherapy
versus open surgery for great saphenous vein varices. Cochrane Database of Systematic Reviews 2014, Issue 7. Art. No.: CD005624.
DOI: 10.1002/14651858.CD005624.pub3.
ABSTRACT
Background
Minimally invasive techniques to treat great saphenous varicose veins include ultrasound-guided foam sclerotherapy (UGFS), radiofre-
quency ablation (RFA) and endovenous laser therapy (EVLT). Compared with flush saphenofemoral ligation with stripping, also
referred to as open surgery or high ligation and stripping (HL/S), proposed benefits include fewer complications, quicker return to
work, improved quality of life (QoL) scores, reduced need for general anaesthesia and equivalent recurrence rates. This is an update of
a review first published in 2011.
Objectives
To determine whether endovenous ablation (radiofrequency and laser) and foam sclerotherapy have any advantages or disadvantages
in comparison with open surgical saphenofemoral ligation and stripping of great saphenous vein varices.
Search methods
For this update the Cochrane Peripheral Vascular Diseases Group Trials Search Co-ordinator searched the Specialised Register (last
searched January 2014) and CENTRAL (2013, Issue 12). Clinical trials databases were also searched for details of ongoing or unpublished
studies.
Selection criteria
All randomised controlled trials (RCTs) of UGFS, EVLT, RFA and HL/S were considered for inclusion. Primary outcomes were
recurrent varicosities, recanalisation, neovascularisation, technical procedure failure, patient QoL scores and associated complications.
Data collection and analysis
CN and RB independently reviewed, assessed and selected trials which met the inclusion criteria. CN and RB extracted data and used
the Cochrane Collaboration’s tool for assessing risk of bias. CN and RB contacted trial authors to clarify details as needed.
Endovenous ablation (radiofrequency and laser) and foam sclerotherapy versus open surgery for great saphenous vein varices (Review) 1
Main results
For this update, eight additional studies were included making a total of 13 included studies with a combined total of 3081 randomised
patients. Three studies compared UGFS with surgery, eight compared EVLT with surgery and five compared RFA with surgery (two
studies had two or more comparisons with surgery). Study quality, evaluated through the six domains of risk of bias, was generally
moderate for all included studies, however no study blinded participants, researchers and clinicians or outcome assessors. Also, nearly
all included studies had other sources of bias. The overall quality of the evidence was moderate due to the variations in the reporting
of results, which limited meaningful meta-analyses for the majority of proposed outcome measures. For the comparison UGFS versus
surgery, the findings may have indicated no difference in the rate of recurrences in the surgical group when measured by clinicians, and
no difference between the groups for symptomatic recurrence (odds ratio (OR) 1.74, 95% confidence interval (CI) 0.97 to 3.12; P =
0.06 and OR 1.28, 95% CI 0.66 to 2.49, respectively). Recanalisation and neovascularisation were only evaluated in a single study.
Recanalisation at < 4 months had an OR of 0.66 (95% CI 0.20 to 2.12), recanalisation > 4 months an OR of 5.05 (95% CI 1.67 to
15.28) and for neovascularisation an OR of 0.05 (95% CI 0.00 to 0.94). There was no difference in the rate of technical failure between
the two groups (OR 0.44, 95% CI 0.12 to 1.57). For EVLT versus surgery, there were no differences between the treatment groups for
either clinician noted or symptomatic recurrence (OR 0.72, 95% CI 0.43 to 1.22; P = 0.22 and OR 0.87, 95% CI 0.47 to 1.62; P =
0.67, respectively). Both early and late recanalisation were no different between the two treatment groups (OR 1.05, 95% CI 0.09 to
12.77; P = 0.97 and OR 4.14, 95% CI 0.76 to 22.65; P = 0.10). Neovascularisation and technical failure were both statistically reduced
in the laser treatment group (OR 0.05, 95% CI 0.01 to 0.22; P < 0.0001 and OR 0.29, 95% CI 0.14 to 0.60; P = 0.0009, respectively).
Long-term (five-year) outcomes were evaluated in one study so no association could be derived,but it appeared that EVLT and surgery
maintained similar findings. Comparing RFA versus surgery, there were no differences in clinician noted recurrence (OR 0.82, 95%
CI 0.49 to 1.39; P = 0.47); symptomatic noted recurrence was only evaluated in a single study. There were also no differences between
the treatment groups for recanalisation (early or late) (OR 0.68, 95% CI 0.01 to 81.18; P = 0.87 and OR 1.09, 95% CI 0.39 to 3.04;
P = 0.87, respectively), neovascularisation (OR 0.31, 95% CI 0.06 to 1.65; P = 0.17) or technical failure (OR 0.82, 95% CI 0.07 to
10.10; P = 0.88).
QoL scores, operative complications and pain were not amenable to meta-analysis, however quality of life generally increased similarly
in all treatment groups and complications were generally low, especially major complications. Pain reporting varied greatly between the
studies but in general pain was similar between the treatment groups.
Authors’ conclusions
Currently available clinical trial evidence suggests that UGFS, EVLT and RFA are at least as effective as surgery in the treatment of
great saphenous varicose veins. Due to large incompatibilities between trials and different time point measurements for outcomes, the
evidence is lacking in robustness. Further randomised trials are needed, which should aim to report and analyse results in a congruent
manner to facilitate future meta-analysis.
PLAIN LANGUAGE SUMMARY

Endovenous ablation (radiofrequency and laser) and foam sclerotherapy versus open surgery for varicose veins
Varicose veins are dilated, tortuous superficial veins. When they are in the legs they can be painful, itchy or unsightly, especially when
patients are standing and walking. Varicose veins are conventionally treated with surgery to remove the veins, by stripping them to the
level of the knee (so-called high ligation and stripping). New less invasive treatments seal the main leaking vein in the thigh using foam
sclerotherapy, laser (endovenous laser therapy) or radiofrequency ablation. These techniques may result in less pain after the procedure,
fewer complications, and a quicker return to work and normal activities with improved patient quality of life, as well as avoiding the
need for a general anaesthetic.
Our review brought together all available randomised controlled trials that compared the new techniques to surgery in the treatment
of varicosities in the great saphenous vein. We found 13 trials, with a combined total of 3081 randomised patients, which met our
inclusion criteria. Three trials compared foam sclerotherapy with surgery, eight trials compared endovenous laser therapy with surgery
and five compared radiofrequency ablation with surgery (two studies had two or more comparisons with surgery). Overall the quality
of the studies was acceptable, however none of the studies tried to conceal the treatment type from the participants, researchers and
clinicians, or those who measured the outcomes. Most of the studies also had other biases.
For foam compared with surgery, there was no difference between the treatment groups in the rate of recurrence as measured by a
clinician and the rate of recurrence that was noted by patient symptoms. There was also no difference between the treatment groups
for technical failure. Comparing laser therapy and surgery, there was no difference between the recurrence rates (either clinician noted
or by symptoms) or for reopening of the treated vein (recanalisation). New vein growth (neovascularisation) and technical failure
were both higher in the surgery group than in the laser group. For the comparison between radiofrequency ablation and surgery there
were no differences between the treatment groups for recurrence, recanalisation, neovascularisation or technical failure. Outcomes that
measure changes in patients’ quality of life, operative complications and pain were not able to be compared directly, however quality
of life generally increased similarly in all treatment groups and complications were generally low, especially major complications. Pain
reporting varied greatly between the studies but in general pain was similar between the treatment groups.
The limited evidence that is available supports that foam sclerotherapy, endovenous laser therapy and radiofrequency ablation are no
worse than open surgery. However, it should be noted that there were large differences between the way the studies reported their
outcomes, which included definitions and collection time points. These differences limited the findings of our review. We need more
data from randomised controlled trials comparing these novel therapies to surgery before we really know their true potential.
Description of the intervention

BACKGROUND Three newer techniques have been developed as alternatives to
open surgery to treat GSV incompetence. They have been advo-
cated to offer benefits over open surgical stripping of the above-
knee portion of the GSV in eliminating saphenofemoral reflux
(Bergan 2002; Navarro 2001; Rautio 2002) and in reducing com-
Description of the condition
plication rates, patient discomfort and the length of time before
Varicose veins are dilated, tortuous superficial veins (varices) which return to work (EVOLVeS Study; Rautio 2002). These three tech-
usually affect the lower extremities. There are two main superficial niques are foam injection sclerotherapy, endovenous laser therapy
venous systems, the great (GSV) and small (SSV) saphenous veins and radiofrequency ablation.
and their tributaries. Endovenous laser therapy and RFA require duplex ultrasound lo-
The surgical management of GSV varices has traditionally con- calisation of the GSV for endovenous insertion of the device, iden-
sisted of saphenofemoral ligation and stripping of the GSV to the tification of the device at the saphenofemoral junction and direct
knee (HL/S) (Rutgers 1994; Sarin 1992). Visible truncal and non- visualisation of the device during treatment. Foam sclerotherapy
saphenous varices in the calf as well as varicose GSV branches in is also carried out under ultrasound visualisation.
the thigh are treated with hook phlebectomies (stab avulsions), 1. Foam sclerotherapy relies on the basic principle of inducing
sclerotherapy or (more recently) transilluminated powered phle- fibrosis of the vein and obliteration of the lumen by causing in-
bectomy. These procedures have often been performed under gen- flammation in the endothelial and subendothelial layers of the
eral anaesthetic (Dwerryhouse 1999; Rautio 2002a), however in- vein wall. The GSV is cannulated under ultrasound guidance and
creasingly no anaesthesia is required or local block techniques have sclerosant foam injected. The foam replaces blood in the vein,
been adopted (Hamel-Desnos 2008; Wright 2006). The recur- which enhances the efficacy of the sclerosing agent by reducing
rence rate at five years following saphenofemoral junction ligation the volume of sclerosant required for treatment and increasing the
and GSV HL/S is reported to be between 20% and 28% (Darke effective surface area of the sclerosant in contact with the vein wall.
1992; Fischer 2002) and in the absence of HL/S (saphenofemoral Ultrasound monitoring during the procedure is necessary to pre-
ligation only) the recurrence rate is doubled (Sarin 1992). For pri- vent spread of the foam into the femoral vein. A major proposed
mary GSV surgery the overall complication rate is between 17% advantage of foam sclerotherapy over conventional sclerotherapy
and 20% (Critchley 1997; Wigger 1998) while the rate for recur- is that larger veins appear to be successfully treated with the foam
rent varicose veins is 40% (Hayden 2001). These complications technique compared with liquid sclerotherapy (Baccaglini 1997;
include wound haematoma, wound infections, lymphatic leaks, Tessari 2001). The success rate varies between 88% and 93% de-
common femoral vein and artery injuries and neurological com- pending on the technique employed to produce the foam (Frullini
plications. Patients are encouraged to mobilise immediately after 2002) and the types of complications and the complication rate
surgery and are usually advised to return to work after 10 to 14 are reported to be the same as with liquid sclerotherapy.
days. Bruising along the tract of the avulsed GSV is common and 2. EVLT involves insertion of a bare tipped optical laser fibre into
can last for up to six weeks following surgery.
the GSV under ultrasound guidance with partial compression ap- in comparison with open surgical saphenofemoral ligation and
plied to the vein as some intravascular blood is necessary for heat- stripping of great saphenous vein (GSV) varices.
ing of the vein; intravascular coagulation therefore occurs with this
method. The optical fibre is then withdrawn along the length of
the vein from just below the saphenofemoral junction to the knee, METHODS
producing obliteration of the vein lumen. This procedure can be
performed under local or general anaesthetic with the addition of
tumescent anaesthesia. Tumescent anaesthesia is the injection of
local anaesthetic around and along the vein length to form a cuff Criteria for considering studies for this review
that protects the skin from burns and also protects the adjacent
cutaneous nerves, especially the saphenous nerve, from damage in
addition to providing local anaesthesia. Patients are encouraged Types of studies
to mobilise immediately after the procedure. The recurrence rate Randomised controlled trials (RCTs) comparing ultrasound-
is reported as 7% after two years (Min 2003), with most recur- guided foam sclerotherapy (UGFS) of the GSV, radiofrequency
rences occurring within the first three months. A low incidence ablation (RFA) and endovenous laser therapy (EVLT) GSV abla-
of bruising and induration around the site of local (tumescent) tion with open GSV HL/S were used.
anaesthetic infiltration has been reported with only few instances
of reports of either superficial or deep phlebitis or thermal skin
and nerve injuries. Types of participants
3. RFA uses radiofrequency energy supplied via a catheter electrode Study participants were males and females of any age with varicose
inserted into the GSV under ultrasound guidance. Compression veins affecting the GSV system, confirmed on duplex ultrasound
of the vein around the catheter is maintained as the catheter is imaging, who were suitable for any of the treatment options. Pa-
withdrawn from just below the saphenofemoral junction to the tients who had undergone previous high ligation but had proven
knee, resulting in heating of the vein wall which is kept at a con- recurrent saphenofemoral junction (SFJ) incompetence and a pre-
stant temperature by a thermocouple. The procedure can be per- served and refluxing GSV to the groin were also included. Stud-
formed under local or general anaesthetic with tumescent anaes- ies that included combined treatments consisting of endovenous
thesia. Immediate mobilisation is encouraged following the pro- treatment or foam sclerotherapy with surgical ligation of the SFJ
cedure. Wound haematoma and infection are reported to be in- were not included. Studies looking at the treatment of varicose
frequent following this procedure, however superficial phlebitis, veins affecting the small saphenous vein (SSV) system were not
thermal skin injuries and nerve injuries have been reported. included. Studies that included patients who had undergone treat-
ment of both GSVs and SSVs and did not provide any subanalysis
of these groups were excluded.
How the intervention might work
Jones 1996 have proposed that open surgery groin haematoma Types of interventions
and disrupted superficial pudendal venous drainage are stimulants 1. Foam sclerotherapy
for neovascularisation (development of new blood vessels) in the 2. Laser endovenous ablation (EVLT)
groin, which are a source of venous reflux recurrence. The newer 3. Radiofrequency endovenous ablation (RFA)
techniques preserve the pudendal and lower abdominal wall ve- 4. Saphenofemoral junction ligation and stripping of the great
nous drainage, do not involve a surgical incision, and prevent saphenous vein (GSV)
haematoma formation in the groin thus potentially removing the
stimulus for neovascularisation and perhaps reducing longer-term
recurrence (Pichot 2004). Additionally, as they are less invasive Types of outcome measures
they may be advantageous in relation to issues such as pain and
recovery times compared with open surgery.
Primary outcomes
1. Recurrence or recanalisation
2. Ultrasound evidence of neovascularisation
OBJECTIVES 3. Technical failure
4. Patient satisfaction determined by quality of life (QoL)
To determine whether endovenous ablation (radiofrequency and scores pre and post-procedure
laser) and foam sclerotherapy have any advantages or disadvantages 5. Post-operative complications
Secondary outcomes Data extraction and management
1. Length of the procedures CN and RB independently extracted data from the included stud-
2. Hospital stay ies. Any disagreements on study selection were resolved through
3. Procedural costs discussion.
Assessment of risk of bias in included studies

Search methods for identification of studies
CN and RB independently assessed the methodological quality of
the included trials using the Cochrane Collaboration’s tool for risk
of bias (Higgins 2011). Any disagreements were resolved through
Electronic searches
discussion.
For this update, the Cochrane Peripheral Vascular Diseases Group
Trials Search Co-ordinator (TSC) searched the Specialised Register
(last searched January 2014) and the Cochrane Central Register of Measures of treatment effect
Controlled Trials (CENTRAL) (2013, Issue 12) in The Cochrane
Statistical analysis was performed according to the Cochrane PVD
Library (www.thecochranelibrary.com). See Appendix 1 for de-
Group’s statistical guidelines for review authors. Odds ratios with
tails of the search strategy used to search CENTRAL. The Spe-
95% confidence intervals were used as the measure of effect for
cialised Register is maintained by the TSC and is constructed from
each dichotomous outcome.
weekly electronic searches of MEDLINE, EMBASE, CINAHL,
For outcomes with continuous scales of measurement we planned
AMED, and through handsearching relevant journals. The full list
to analyse the data in continuous form (that is using mean dif-
of the databases, journals and conference proceedings which have
ference and standard deviation). If different scales were used in
been searched, as well as the search strategies used are described in
different studies, where possible the results were standardised and
the Specialised Register section of the Cochrane Peripheral Vas-
then combined (that is using standardised mean difference and
cular Diseases (PVD) Group module in The Cochrane Library (
standard deviation).
www.thecochranelibrary.com).
The following trial databases were searched by the TSC (January
2014) for details of ongoing and unpublished studies using the
Unit of analysis issues
terms varicose veins and surgery:
• World Health Organization International Clinical Trials Due to the nature of the disease and interventions, the studies
Registry (http://apps.who.int/trialsearch/); relating to this subject matter often have mixed units of analysis
• ClinicalTrials.gov (http://clinicaltrials.gov/); (that is some studies may randomise by patient and some ran-
• ISRCTN register (http://www.controlled-trials.com/ domise by legs, including patients with both uni- and bilateral
isrctn/). venous incompetence). For the purposes of the analyses we tried
to use patients as our unit of analysis, but in some cases this was
not possible and the number of limbs was used. Despite enquiries
Searching other resources to relevant authors for data per patient, we were unable to include
only per patient data so analyses should be interpreted with this
Reference lists from relevant studies and reviews were checked to
in mind.
make sure all possible studies were evaluated for inclusion.
Dealing with missing data

Data collection and analysis Where possible, study authors were contacted to provide missing
The Cochrane Collaboration’s tool for assessing risk of bias was data. Meta-analyses were performed using available or provided
used when assessing the risk of bias in the included studies (Higgins data.
2011).
Assessment of heterogeneity
Selection of studies Heterogeneity in the data was noted and explored using previously
CN and RB independently reviewed the trials and selected those identified characteristics of the studies, particularly assessments of
which met the above criteria. Any disagreements on study selection quality. The I2 statistic was used to determine heterogeneity. A
were resolved through discussion or by consulting with a third random-effects model was considered for I2 values greater than
author (GS). 50%.
Assessment of reporting biases random-effects model was considered for I2 values greater than
To evaluate reporting bias, we planned to construct funnel plots 50%.
for meta-analyses that have 10 or more studies included.
Sensitivity analysis
Data synthesis Sensitivity analyses were performed within meta-analyses for stud-
Where there were sufficient data, a summary statistic for each ies that had a higher risk of bias in four or more bias categories,
outcome was calculated using a fixed-effect model or a random- which included EVOLVeS Study and HELP-1, in order to deter-
effects model, depending on the heterogeneity in the data between mine the impact of these studies on the overall findings.
each study.
Subgroup analysis and investigation of heterogeneity

RESULTS
We planned to undertake subgroup analyses to examine the sta-
bility of the results in relation to a number of factors including
patient type. However, due to a lack of outcome data reported by
Description of studies
categories of interest, subgroup analysis was not performed at this
time.
Heterogeneity in the data was noted and explored using previously
identified characteristics of the studies, particularly assessments of Results of the search
quality. The I2 statistic was used to determine heterogeneity. A See Figure 1.
Figure 1. Study flow diagram.
publications. Rasmussen 2007 has published five-year data, which
Included studies
has been added to the update of this review. The only exception
For this update eight additional studies were included ( is Subramonia 2010, which had no results beyond five weeks. A
Flessenkaemper 2013; FOAM-Study; Helmy ElKaffas 2011; summary of the outcome measures of the trials is provided in Table
HELP-1; Magna 2007; Pronk 2010; Rasmussen 2011; RELACS 4.
Study) making a total of 13 included studies (Darwood
2008; EVOLVeS Study; Flessenkaemper 2013; FOAM-Study;
Helmy ElKaffas 2011; HELP-1; Magna 2007; Pronk 2010; Excluded studies
Rasmussen 2007; Rasmussen 2011; Rautio 2002; RELACS Study; For this update, there were 12 additional studies excluded
Subramonia 2010). (Compagna 2010; Disselhoff 2011; Goode 2010; HELP-2; Jia
See Characteristics of included studies for full details. 2010; Kalodiki 2012; Lattimer 2012; Lin 2009; Liu 2011;
All 13 trials were prospective RCTs, varying between single-, dou- Rasmussen 2010; Shepherd 2009; Yang 2013) making a total of
ble- and multicentre settings. The included studies were performed 75 excluded studies (Abela 2008; Almeida 2007; Alos 2006; Anon
in centres in the UK, Austria, USA, France, Denmark, Finland, 2008; Belcaro 2000; Blaise 2010; BLARA Trial; Bountouroglou
the Netherlands, Germany and Egypt, with most in public hospi- 2006; Bush 2008; Ceulen 2007; Chant 1972; Christenson 2010;
tals. Compagna 2010; Demagny 2002; De Medeiros 2006; Desmyttere
In order to achieve congruity, sample sizes were considered in terms 2005; Disselhoff 2008; Disselhoff 2011; Doran 1975; Duffy
of ’number of patients’ rather than ’number of limbs’, although 2005; Einarsson 1993; Figueiredo 2009; Gale 2009; Goode 2010;
in some cases this was not possible. Sample sizes by number of Hamel-Desnos 2003; Hamel-Desnos 2005; Hamel-Desnos 2008;
patients randomised ranged from 33 patients (Rautio 2002) to Hamel-Desnos 2009; Hayes 2008; HELP-2; Hinchliffe 2006;
500 patients (Rasmussen 2011). The total number of patients Hobbs 1968; Jia 2010; Kabnick 2003; Kalodiki 2012; Kalteis
randomised was 3081 and results from 2489 were analysed. Table 1 2008; Kern 2005; Kuznetsov 2005; Lattimer 2012; Lin 2007;
summarises the study sample sizes. Participant ages ranged from as Lin 2009; Liu 2011; Lugli 2009; Lupton 2002; Martimbeau
low as 18 years (Rasmussen 2011) to as high as 79 years (Rasmussen 2003; Maurins 2009; McDaniel 1999; Mekako 2006; Mekako
2007). All included studies had more women than men, which 2007; NCT00841178; Neglen 1993; Ogawa 2008; Ouvry
reflects the natural epidemiology of the disease. See Table 2 for 2008; Rabe 2008; Rao 2005; Rasmussen 2010; REACTIV;
full details of the age and sex of the trial participants. RECOVERY Trial; Rutgers 1994; Rybak 2003; Sadoun 2003;
Three studies compared foam sclerotherapy with open surgery Seddon 1973; Selles 2008; Shepherd 2009; Sica 2006; Stotter
(FOAM-Study; Magna 2007; Rasmussen 2011). Eight studies 2005; Theivacumar 2008; Theivacumar 2009; VEDICO Trial;
compared endovenous laser ablation (EVLT) of the GSV with Viarengo 2007; Vuylsteke 2006; Wright 2006; Yamaki 2008; Yang
open surgery (Darwood 2008; Flessenkaemper 2013; HELP-1; 2013; Zeh 2003).
Magna 2007; Pronk 2010; Rasmussen 2007; Rasmussen 2011; See Characteristics of excluded studies for the full list of excluded
RELACS Study). See Table 3 for the laser types used in studies studies and the reasons for exclusion.
evaluating EVLT. Five studies compared radiofrequency ablation
(RFA) of the GSV with open surgery (EVOLVeS Study; Helmy
ElKaffas 2011; Rasmussen 2011; Rautio 2002; Subramonia
Risk of bias in included studies
2010). All trials followed up their participants within the first three Full details on risk of bias can be found in the tables within the
months. Follow-up continued for at least one or two years for Characteristics of included studies section as well as in Figure 2
most studies, and several expected five-year follow-ups in future and Figure 3.
Figure 2. Methodological quality graph: review authors’ judgements about each methodological quality
item presented as percentages across all included studies.
Figure 3. Methodological quality summary: review authors’ judgements about each methodological quality
item for each included study.
Allocation
which could lead to altered outcomes for the treatment groups.
Seven studies reported their random sequence generation thor- Also, there was no indication as to whether bilaterally treated pa-
oughly (Darwood 2008; Flessenkaemper 2013; FOAM-Study; tients were included. The authors were contacted for clarification
Magna 2007; Pronk 2010; Rasmussen 2007; Rautio 2002). Six but they did not respond.
studies had unclear selection bias for random sequence genera- Issues of recurrence were brought up in the FOAM-Study in a letter
tion because they did not adequately describe their generation at the back of the publication, which highlighted a much higher
method or it was unclear if the method would be truly random than expected recurrence rate in the surgical group, indicating
(EVOLVeS Study; Helmy ElKaffas 2011; HELP-1; Rasmussen possibly suboptimal surgical techniques. Also in this study, mini-
2011; RELACS Study; Subramonia 2010). All 13 studies had ad- phlebectomies were performed at the discretion of the surgeon,
equate allocation concealment techniques. which led to an imbalance between the treatment groups possibly
altering pain and other outcomes. Other studies with possible
bias relating to additional phlebectomies were Darwood 2008,
Blinding
Flessenkaemper 2013, FOAM-Study and RELACS Study. See
It was not possible for any of the study authors to blind either Table 5 for more information on which studies included additional
their patients or the operators, leading to high risk evaluations for phlebectomies.
all studies. Many subjective measures were made at subsequent Rasmussen 2007 and Rasmussen 2011 performed all procedures
follow-up visits, for example ’recurrence as noted by the observer’, in an office-based setting in a private clinic under tumescent anaes-
but no study blinded their post-operative assessors. thesia. Other studies that used local tumescent anaesthesia in all
treatment groups were Pronk 2010 and RELACS Study. All other
Incomplete outcome data studies used general anaesthesia in their surgical arm in a hospital-
based setting. It was possible that the private clinic setting and
All included studies reported their incomplete outcome data. local anaesthesia could confound results on absence from work,
The majority of studies had low risk of attrition bias, but
normal activity, QoL and pain scores.
Flessenkaemper 2013 did not mention the number of dropouts
Darwood 2008, EVOLVeS Study, Magna 2007, Pronk 2010,
and the Pronk 2010 study did not clarify the numbers used for
Rasmussen 2007, Rasmussen 2011 and Subramonia 2010 in-
analyses, leading to unclear risk for the two studies. The EVOLVeS
cluded patients who underwent treatment of bilateral vari-
Study had a high risk of attrition bias because the post-randomi- cose veins. Darwood 2008, Pronk 2010, Rasmussen 2007 and
sation exclusions led to clear imbalances, as well as a discrepancy Rasmussen 2011 randomised patients once to receive the same
of one patient when the results of their two-year follow-up were
treatment. Results were presented as ’number of limbs or legs’ and
scrutinised.
no stratification was made of those patients who underwent bilat-
eral treatment. Participants in the Magna 2007 study with bilateral
Selective reporting incompetence could have limbs randomised to different interven-
tions. Some outcome measures could be affected by bilateral treat-
Most of the included studies had low risk of reporting bias as
ment, for example pain, QoL measures, time to return to work,
all the predefined outcomes were reported on. However, Magna
therefore introducing a potential bias in the interpretation of these
2007 did not report on several complications that were outlined
measures. This was especially true in Darwood 2008 where 15
in their methods. Although minor, this resulted in unclear risk of
patients had bilateral laser ablation compared with just four surgi-
reporting bias.
cal patients who underwent bilateral stripping. Subramonia 2010
and the EVOLVeS Study waited for full recovery (> six weeks)
Other potential sources of bias before randomising a patient’s second limb thus minimising this
potential bias. Rautio 2002 excluded patients with bilateral veins.
Darwood 2008 declared that their sample sizes were insufficient
The effect of presenting results as ’limbs’ rather than ’patients’ is
to permit statistical testing for equivalence. Results were not anal-
discussed in Overall completeness and applicability of evidence.
ysed on an intention-to-treat basis since one patient was ran-
domised to surgery but underwent laser treatment and was anal- Subramonia 2010, Rasmussen 2007 and Rasmussen 2011 in-
cluded patients who had undergone previous ligation of the SFJ.
ysed with the laser patients. Other possibly underpowered stud-
They were included on the grounds that they had recanalised their
ies were Flessenkaemper 2013, HELP-1, Magna 2007 and Pronk
GSV and had a patent, refluxing SFJ and GSV. It was acknowl-
2010.
edged by the authors that ’surgery’ for such recurrent patients
In the Helmy ElKaffas 2011 study it was reported that two con-
would not be considered ’conventional’. This formed the basis for
sultant operators performed RFA and just one performed HL/S,
our exclusion of Hinchliffe 2006. Unlike Hinchliffe 2006, the in- 0.43 to 1.22; P = 0.22; I2 = 60% and OR 0.87, 95% CI 0.47 to
cluded trials had only small numbers of recurrent varicose veins. 1.62; P = 0.67; I2 = 0%, respectively) (Analysis 2.1).
CN contacted authors to see if any subgroup analysis had been Darwood 2008 presented two-year results on both recurrence and
carried out on these recurrent patients and in both instances it had neovascularisation but they were presented in conjunction with
not been done. Owing to both the small numbers and the equal all other patients who were treated (non-randomly) at the time of
distribution of the recurrent patients in the treatment cohorts it the original RCT. There was no stratification in these results of the
was decided to include these trials. Wound complications, pain randomised and non-randomised patients, therefore results could
scores and time to return to work or normal activities appeared to not be utilised in this review.
be no higher in these trials compared to the other included papers. For the RELACS Study, the publication reported recurrences in
several places, with different outcome numbers and possibly dif-
ferent definitions. For our analyses we used data from Table 3
Effects of interventions
of the RELACS Study publication. Also, the number of clinical
versus symptomatic recurrences was unclear, so only the REVAS
classified recurrences were used.
Primary outcomes Five-year data from the Rasmussen 2007 study were recently made
available. Because only a single study reported findings at this time
point no overall associations could be made, but the number of
recurrences was similar between the groups (OR 0.94, 95% CI
Recurrence
0.46 to 1.89) (Analysis 2.5).
Recurrence was evaluated in 11 studies with a variation of defini-
tions of recurrence as found in Characteristics of included studies
(EVOLVeS Study; Flessenkaemper 2013; FOAM-Study; Helmy (iii) Radiofrequency ablation versus surgery
ElKaffas 2011; HELP-1; Magna 2007; Pronk 2010; Rasmussen
Four studies evaluated clinician noted recurrences, comparing
2007; Rasmussen 2011; Rautio 2002; RELACS Study).
RFA and surgery (EVOLVeS Study; Helmy ElKaffas 2011;
Five studies separately reported clinical recurrences and symp-
Rasmussen 2011; Rautio 2002), of which only one also evaluated
tomatic recurrences noted by patients (Flessenkaemper 2013;
symptomatic recurrences (Rautio 2002). The fixed-effect model
FOAM-Study; Pronk 2010; Rasmussen 2007; Rautio 2002).
meta-analyses found no difference between the two treatment
Table 6 shows the results of recurrence and neovascularisation as
groups for either clinical or symptomatic recurrences (OR 0.82,
noted by clinicians and patients.
95% CI 0.49 to 1.39; P = 0.47 and OR 2.00, 95% CI 0.30 to
13.26, respectively) (Analysis 3.1).
(i) Foam versus surgery

Three studies evaluated clinician noted recurrence, comparing Recanalisation
UGFS and surgery (FOAM-Study; Magna 2007; Rasmussen We defined obliteration as complete anatomical obliteration on
2011) and the random-effects model meta-analysis found no sta- duplex ultrasound. Partial obliteration without reflux was consid-
tistically significant difference between the treatment groups (OR ered obliteration failure. The standard criterion for reflux was used:
1.74, 95% CI 0.97 to 3.12; P = 0.06; I2 = 55%). Of the three at least 0.5 s of reflux. Recanalisation in the surgical group was
studies, only the FOAM-Study also evaluated symptomatic recur- defined as incompetent GSV refilling from any saphenofemoral
rences with an OR of 1.28 (95% CI 0.66 to 2.49). No association neovascularisation following stripping, which could occur if pri-
could be determined as only a single study was evaluated (Analysis mary surgery was a failure (incomplete stripping) or a duplicate
1.1). GSV was not identified or stripped during the primary operation.
Recanalisation could occur with or without recurrent varicosities.
Recanalisation was reported at varying time intervals across the
(ii) Laser versus surgery studies. We took four months as the cut off between early and late
Seven studies evaluated clinician noted recurrence, compar- recanalisation.
ing EVLT and surgery (Flessenkaemper 2013; HELP-1; Magna
2007; Pronk 2010; Rasmussen 2007; Rasmussen 2011; RELACS
Study), of which three also evaluated symptomatic recurrences (i) Foam versus surgery
(Flessenkaemper 2013; Pronk 2010; Rasmussen 2007). For both Only a single study (Rasmussen 2011) that compared foam with
clinical recurrence determined by duplex ultrasound and symp- surgery reported on recanalisation, and they reported per limb and
tomatic recurrence the random-effects model meta-analyses found not per patient. Early recanalisations had an OR of 0.66 (95%
no difference between the treatment groups (OR 0.72, 95% CI CI 0.20 to 2.12) and late recanalisations an OR of 5.05 (95%
CI 1.67 to 15.28). With only a single study reporting, no overall (i) Foam versus surgery
association could be determined (Analysis 1.2). Only a single study comparing UGFS with surgery reported neo-
vascularisations (Magna 2007). This study had a very wide CI due
(ii) Laser versus surgery to no events in the foam group: OR of 0.05 (95% CI 0.00 to 0.94)
(Analysis 1.3).
Three studies evaluated early recanalisation (Darwood 2008;
Rasmussen 2007; Rasmussen 2011) and four evaluated late re-
canalisation (Darwood 2008; Rasmussen 2007; Rasmussen 2011;
RELACS Study). There were no differences between the treatment (ii) Laser versus surgery
groups in the random-effects model meta-analyses: early recanal- For the comparison between laser ablation and surgery, four stud-
isations had an OR of 1.05 (95% CI 0.09 to 12.77; P = 0.97; I2 ies reported neovascularisation (Darwood 2008; HELP-1; Magna
= 61%) and late recanalisations an OR of 4.14 (95% CI 0.76 to 2007; RELACS Study). The fixed-effect model analyses found
22.65; P = 0.10; I2 = 58%) (Analysis 2.2). a statistically significant association favouring EVLT (OR 0.05,
Five-year data from Rasmussen 2007 reported nine events in the 95% CI 0.01 to 0.22; P < 0.0001) but this was expected as neo-
HL/S group and no events in the EVLT group, with an OR of 0.05 vascularisation can truly only occur in the case of GSV removal
(95% CI 0.00 to 0.80). No overall association could be determined (therefore generally only in the surgical group) (Analysis 2.3).
from the single study (Analysis 2.6).
(iii) Radiofrequency ablation versus surgery (iii) Radiofrequency ablation versus surgery
Four studies comparing RFA to surgery evaluated early re- Neovascularisation was evaluated in two studies that compared
canalisation (EVOLVeS Study; Rasmussen 2011; Rautio 2002; RFA and surgery (EVOLVeS Study; Rautio 2002). There was no
Subramonia 2010) and three evaluated late recanalisation ( difference between the two treatment groups according to the
EVOLVeS Study; Rasmussen 2011; Rautio 2002). No differences fixed-effect model meta-analysis (OR 0.31, 95% CI 0.06 to 1.65;
were found between the treatment groups in the random-effects P = 0.17), although the point estimates from both studies favoured
model analyses: early recanalisation had an OR of 0.68 (95% CI radiofrequency ablation (Analysis 3.3).
0.01 to 81.18; P = 0.87; I2 = 81%) and late recanalisation had an
OR of 1.09 (95% CI 0.39 to 3.04; P = 0.87; I2 = 0%) (Analysis
3.2). Technical failure and re-intervention
Re-intervention was only fully evaluated in three studies, therefore
Neovascularisation we chose to use technical failure for analysis. However, Table 7
Magna 2007 used a two level classification of neovascularisation: does present the available re-intervention data.
1) tiny new veins up to 3 mm in diameter not connecting with
any superficial vein and 2) tortuous new veins with a diameter ≥
4 mm with pathological reflux and connecting with thigh varicose (i) Foam versus surgery
veins. All of the documented neovascularisations from the Magna Both Magna 2007 and Rasmussen 2011 reported technical fail-
2007 study were level one. Darwood 2008 mentioned neovascu- ure and the fixed-effect model meta-analysis found no difference
larisation as a cause of SFJ incompetence and GSV reflux in one between the groups with an OR of 0.44 (95% CI 0.12 to 1.57;
patient, but only followed up 12 of the 32 patients randomised P = 0.20); although the point estimates of both studies favoured
to surgery. They did not mention neovascularisation as a finding foam sclerotherapy (Analysis 1.4).
in any of their patients who underwent laser treatment. HELP-1
defined neovascularisations as serpentine vessels emanating from
the SFJ that were not present on duplex imaging at one or six
weeks. RELACS Study did not specifically indicate how they de- (ii) Laser versus surgery
fined neovascularisation. EVOLVeS Study defined neovascularisa- Six studies reported technical failures (Darwood 2008; HELP-1;
tion as “multiple small vessels in the groin reconnecting a more Magna 2007; Rasmussen 2007; Rasmussen 2011; RELACS
proximal vein or its tributaries and the distal patent vein below Study) and the fixed-effect model meta-analysis found a statisti-
the site of interruption, surgery or RFA”, determined by duplex cally significant odds reduction favouring the laser treatment group
ultrasound examination. Rautio 2002 regarded neovascularisation (OR 0.29, 95% CI 0.14 to 0.60; P = 0.0009) (Analysis 2.4).
as “small superficial branches at the area of the SFJ”. Interestingly Data from the Rasmussen 2007 five-year report had similar tech-
Rautio 2002 commented that neither patient identified with neo- nical failure rates, with an OR of 0.66 (95% CI 0.11 to 4.06)
vascularisation had clinical recurrent varicose veins. (Analysis 2.7).
(iii) Radiofrequency ablation (RFA) versus surgery (ii) Laser versus surgery
Five studies reported technical failures (EVOLVeS Study; Helmy Darwood 2008 reported increases in both AVVSS and VCSS
ElKaffas 2011; Rasmussen 2011; Rautio 2002; Subramonia scores for all treatment groups (P < 0.001, both scores) with no
2010). With few events reported we used the random-effects significant differences between the groups.
model and found no difference between the groups with a wide CI For the Flessenkaemper 2013 study, no data were presented for the
(OR 0.82, 95% CI 0.07 to 10.10; P = 0.88; I2 = 70%) (Analysis VCSS but the authors stated “VCSS parameters for postoperative
3.4). pain after an interval of two months showed that the data for
all three groups became extensively more similar after this time,
and that retrospectively no appreciable difference existed”. For
Quality of life (QoL) measures the venous disability score (VDS) Flessenkaemper 2013 reported
an increase in the proportion of asymptomatic patients in both
All included studies, with the exception of Helmy ElKaffas 2011,
treatment groups, and decreases in symptomatic patients at all
evaluated changes in QoL or venous severity scores, or both, using
levels. They showed improvements in both treatment groups in
a variety of questionnaires including disease-specific severity scores
the C-classification of CEAP from baseline to two months, with no
and QoL questionnaires and generic health-related QoL question-
difference between the groups (P = 0.90). The Hach classification
naires. Due to the vast number of different measures as well as
scores were only reported as baseline measures.
the different ways of recording and time points of the results it
HELP-1 reported for AVVSS an initial worsening in both study
was inappropriate to combine the results in a meta-analysis. The
groups one week after treatment (P < 0.001) followed by an im-
different measures reported within the included studies can be see
provement in both groups for the rest of the follow-up period (P <
in Table 8.
0.001), with no difference between the groups at any time point.
For the SF-36 there were deteriorations within several domains
for both groups within the first week but after the first week most
(i) Foam versus surgery domains showed overall improvements for both groups and from
The FOAM-Study showed similar improvements in both treat- four weeks on there were no differences between the groups. There
ment groups in the EuroQol 5D (EQ-5D) utility score at two years were initially worse scores in the first week for EQ-5D and then
(P = 0.889) as well as similar improvements in the Venous Clinical improvements for the remainder of the study (P < 0.001), with no
Severity Score (VCSS) in both the foam and surgery groups from differences between groups. SF-6D (a variation of SF-36) scores
baseline to two years (P = 0.232). showed initial deterioration within the first week, only seen in
Magna 2007 reported improved Chronic Venous Insufficiency the surgery group (P < 0.001), with no difference for EVLT from
Quality of Life Questionnaire (CIVIQ2) and EQ-5D scores in baseline (P = 0.141). Through follow-up both treatment groups
all treatment groups (foam, EVLT and surgery) at three months, had significantly improved scores (P < 0.001) and the EVLT group
remaining stable until one year, with no significant differences be- had significantly better scores than the surgical group (P = 0.003).
tween the groups. For the C-classification in the Clinical severity, For the VCSS scoring, HELP-1 reported similar, statistically sig-
Etiology, Anatomy, Pathophysiology (CEAP) scoring the Magna nificant improvements by three months (P < 0.001) which were
2007 study reported that 47.6% of all patients showed improve- maintained at one year, with no differences between the groups.
ment of at least two categories, with no differences between the CEAP scores were only measured at baseline, with no follow-up
three treatment groups. This description applied for all three treat- comparisons.
ment groups in the Magna 2007 trial and not just the comparison Pronk 2010 reported for those who underwent EVLT significantly
between foam and surgery. worse scores on the EQ-5D for mobility on days 7 and 10 (P < 0.01,
Rasmussen 2011 reported deteriorations for all treatment groups P = 0.01, respectively), daily activity on day 7 (P = 0.03) as well as
by Medical Outcomes Study Short Form-36 (SF-36) scoring at self care on day 7, but better scores for daily activity on day 1 (P =
three days, followed by improved scores in all four treatment 0.01) compared with surgery. Improvements were seen in both the
groups by one year. The RFA and UGFS groups performed better EVLT and surgery groups after one year for symptoms of chronic
in the short term for the domains bodily pain and physical func- venous disease (C-classification) (P < 0.01) and in overall CEAP
tioning. There were improvements in the Aberdeen Varicose Vein classification (P < 0.01) but there were no differences between the
Symptom Severity Score (AVVSS) for all treatment groups from groups (P = 0.28).
one month after surgery, with no differences between groups. The Rasmussen 2007 reported improvements in both treatment groups
reported VCSS for all four treatment groups demonstrated statis- for AVVSS, SF-36 (bodily pain, vitality and social functioning
tically significant improvements (P < 0.001), with no differences domains) and severity scoring by the VCSS for measures observed
between the groups. This description applies for all four treatment from three months onwards (P < 0.01). The improvements in
groups within the Rasmussen 2011 study and not just foam versus all measures were still present after two years. Five-year follow-
surgery. up data suggested that improvements in QoL and disease severity
were maintained if not continuing to improve. tribution of the great saphenous nerve’, ’numbness’ and ’saphe-
RELACS Study reported similar improvements for both treatment nous nerve injury’ were reported separately. In an attempt to dis-
groups for the disease-specific Homburg Varicose Vein Severity play such a wide and varying collection of complications we have
Score (HVVSS) and CIVIQ2. The HVVSS significantly decreased constructed a table which combines many of these ’similar’ events
(improved) by three months (P < 0.001) and then further decreased (Table 9). Across the studies, adverse events were recorded at vary-
by 12 months (P < 0.001) with no differences between the groups. ing times post-operatively. We have divided them into early (within
A similar pattern of improvement was observed for the CIVIQ2 three months) and late (beyond three months), again for simplic-
measurements. ity. We have also characterised adverse events as major and minor
depending on whether or not these events required intervention:
minor requiring no intervention and major requiring interven-
(iii) Radiofrequency ablation versus surgery tion. When events were presented more than once within the first
EVOLVeS Study reported an initial decrease in the QoL CIVIQ2 three months, as in the EVOLVeS Study, which presented adverse
scores for the surgery group in the first week, which was not ob- events at 72 hours, one and three weeks post-operatively, the high-
served in the RFA group (which had an initial increase). After three est value for any given event was recorded in our summary table.
weeks the differences between the two groups were not signifi-
cant as they both saw increases in CIVIQ2 scores. After one year
Secondary outcomes
there was once again a statistically significant difference between
the two groups, favouring RFA, which was maintained at the two-
year follow-up. A significant difference in VCSS score between the
Length of the procedure or operative time
groups was observed at 72 h and one week after treatment favour-
ing RFA. These differences were lost after three weeks, when both Table 10 shows length of procedure for the six reporting studies
groups continued to see improvement. The C-classification in the (EVOLVeS Study; Helmy ElKaffas 2011; HELP-1; Rasmussen
CEAP score saw general improvements from baseline to two years 2011; Rautio 2002; Subramonia 2010).
with no differences between the treatment groups. All statistically Rasmussen 2011 reported the length of the surgeon’s time for the
significant differences were reported as P < 0.05. procedure, while Subramonia 2010 gave values for operative and
Rautio 2002 evaluated generic QoL measures with the RAND-36 theatre time, presenting median values and an interquartile ratio
(a variation of SF-36 validated for Finland) and found general im- (IQR). Rautio 2002 presented operating time, operating room
provements across all subgroups at week four, with no differences time and recovery time using the mean and standard deviation.
between the treatment groups. The average improvement of the EVOLVeS Study admitted in his results that venous access time,
VCSS at three years was similar between the treatment groups (P = treatment time and adjunctive procedure time were not collected
0.7). The pre-operatively evaluated VDS was improved in all but at uniform times, resulting in confusion with their overall figures.
two patients (one in each group) who occasionally needed com- They presented total treatment times only. No units of time were
pression stockings while working. The Venous Segmental Disease provided. The lack of congruity with the presented results pre-
Score (VSDS) decreased in all but three patients in the RFA group vented any meaningful meta-analysis.
and one patient in the surgical group (P = 0.6).
In the Subramonia 2010 study, both the Venous Insufficiency Epi-
demiological and Economics Study (VEINES)-QoL/Sym ques- Duration of hospital stay
tionnaire (V-Q/SymQ) and AVVSS (referred to as AVVQ) showed Eight studies (Darwood 2008; EVOLVeS Study; Flessenkaemper
a significant improvement in venous symptoms and QoL for both 2013; FOAM-Study; Helmy ElKaffas 2011; HELP-1; Pronk
treatment groups after intervention, with no statistically signif- 2010; Rasmussen 2007) reported whether patients were day pa-
icant differences between the groups. The Total Clinical Sever- tients or in-patients. Table 11 details the available data. The ma-
ity Score (TCSS) and VDS both showed improvements of symp- jority of patients were operated on as day cases.
toms in more than three-quarters of participants in each treatment
group.
Procedural costs
Cost analysis was provided by six of the studies (FOAM-Study;
Post-operative complications Helmy ElKaffas 2011; Rasmussen 2007; Rasmussen 2011; Rautio
All studies declared and reported post-operative complications. 2002; Subramonia 2010) but the costs involved for each study
Terminology was the main factor that prevented accurate meta- varied. Table 12 summarises the cost information that was pre-
analysis. For example, ’post-operative wound infection’, ’wound sented. Of some significance was the cost of additional procedures
discharge’ and ’wound breakdown’ were reported separately across for residual or recurrent varicosities and none of the studies pro-
the studies. Similarly, ’post-operative paraesthesia along the dis- vided any estimation of these costs.
Additional outcome measures was 12.9 tablets in laser patients versus 12 tablets in surgery pa-
tients within the six months post-procedure. There were no sta-
tistically significant differences in the visual analogue scale (VAS)
Time to return to work or normal activities scores for pain between the two groups, although pain was initially
The time to return to work or normal activities following the in- slightly higher after surgery. There were no differences between the
tervention was reported in most trials. Unfortunately the different laser and surgery treatment groups in the Rasmussen 2011 study,
studies reported these results as either parametric or non-paramet- and the number of phlebectomies did not influence pain scores (P
ric data (mean, median, range, IQR) precluding accurate meta- = 0.136). The RELACS Study reported significantly higher pain
analysis. Table 13 summarises the results that were available. in the laser group in the first post-operative week (P = 0.005) but
a significantly shorter duration of pain (by days) in the laser group
compared with the surgery group (P = 0.03).
Type of anaesthetic required From the above descriptions of the studies there are clearly con-
Table 14 summarises the type of anaesthetic used in each trial. flicting findings on pain for the comparison between ELVT and
In four studies local tumescent anaesthesia was used for the HL/ surgery: Darwood 2008, Flessenkaemper 2013 and Rasmussen
S treatment group as well as the alternative therapy, which is not 2011 showed no difference between the groups; in contrast
currently standard practice for surgical varicose vein treatment. HELP-1 showed less pain following EVLT compared to HL/S;
This trend emerged with the newer studies included in the update while Pronk 2010 and RELACS Study clearly showed that EVLT
and was only seen in one study in the older version of this review. was associated with significantly more early post-operative pain.
Differences in outcomes such as pain, return to work or normal Rasmussen 2007 had mixed results with the EVLT group using
activity and complications between the HL/S groups in different more pain medication but reporting initially higher levels of pain
studies could be due to the different anaesthesia regimens used. after surgery.
The type of anaesthetic was one explanation for these observed dif-
ferences as four studies (Pronk 2010; Rasmussen 2007; Rasmussen
Post-procedure pain
2011; RELACS Study) performed HL/S under tumescence and
There was a wide variation in how pain was analysed and reported, three studies (Darwood 2008; Flessenkaemper 2013; HELP-1)
which precluded any meaningful meta-analysis. Brief summaries used general or spinal anaesthesia. This suggested a trend towards
of each study’s findings are outlined below. less post-operative pain with tumescence anaesthesia, and there-
fore a benefit to HL/S over EVLT.
(i) Foam versus surgery
FOAM-Study determined that the treatment did not greatly in- (iii) Radiofrequency ablation versus surgery
fluence pain, with similar scoring for ’more’, ’stable’ or ’less’ pain In the EVOLVeS Study the authors reported statistically significant
between the three time points (3, 12 and 24 months) and no dif- differences in the pain scores recorded at 72 hours and one week
ferences between the groups. In the Rasmussen 2011 study, pa- post-procedure in favour of RFA (P < 0.001, both time points).
tients in the foam treatment group had less post-operative pain Rasmussen 2011 reported that patients in the RFA group had less
than those in the surgery group (P < 0.001) and the number of post-operative pain than those in the surgery group (P < 0.001),
phlebectomies did not influence pain scores (P = 0.136). and the number of phlebectomies did not influence pain scores
(P = 0.136). Rautio 2002 reported that the mean pain scores at
rest, standing and walking were lower in the RFA group, especially
(ii) Laser versus surgery
between the fifth to 14th post-operative days, but these differences
Darwood 2008 showed a decrease in pain for all treatment groups were not statistically significant (P = 0.36). The RFA patients
in the first week after treatment, with no significant differences used less ibuprofen analgesia compared to surgical patients (P =
between the groups. Flessenkaemper 2013 reported no statistically 0.001). Subramonia 2010 reported median values for the pain
significant difference in the experience of pain for more than five scores within the first post-procedure week to be statistically higher
days or for persisting pain after treatment between the treatment in the surgery patients compared with RFA patients (P = 0.001),
groups (P = 0.12). HELP-1 reported less pain in the laser group and the duration of analgesia use was significantly shorter for RFA
compared with the surgical group from day one (P = 0.004 to patients (P = 0.001).
P < 0.001). In the Pronk 2010 study it was reported that there
was significantly more pain during treatment in the surgery group
compared with laser (P = 0.02). For days 7, 10 and 14 the laser Reporting bias, subgroup analysis and sensitivity
group experienced significantly more pain than those patients who analysis
underwent surgery (P < 0.001, P < 0.001 and P = 0.01, respec- Reporting bias was not evaluated as none of the analyses included
tively). Rasmussen 2007 reported that the mean analgesic intake the 10 or more studies needed in order to construct funnel plots.
Subgroup analysis was not performed in this update as none of the two studies reporting costs, although they were measuring dif-
the studies presented outcome data by the predefined variables ferent specific costs both had decreased costs within the foam treat-
of interest. Sensitivity analyses performed by removing the lower ment group compared with surgical patients. Time to return to
quality studies (EVOLVeS Study; HELP-1) from the meta-analy- work and normal activity was only reported in a single study but
ses did not change any of the findings for any of the analyses they showed decreased recovery time for the foam group. Two studies
were present in. reported on pain, with one reporting no large change in pain for
either treatment group from baseline and the other study report-
ing less pain during the procedure in the foam group compared
with the surgical group.
DISCUSSION
(ii) Laser versus surgery

Summary of main results Of the outcomes that were amenable to meta-analysis, rates of
recurrence and recanalisation were no different between the two
A total of 13 studies were included in this systematic review, with
treatment groups. Neovascularisation and technical failure were
3081 randomised participants. Three studies with a total of 870
both less in the laser treated groups compared with surgery. QoL
randomised participants compared foam sclerotherapy to surgery,
and disease severity scoring was generally uniform throughout the
a total of eight studies with 1760 randomised participants com-
studies, with worsening within the first few days or weeks followed
pared endovascular laser ablation therapy (EVLT) to surgery, and
by an overall improvement over the follow-up period, with no
five studies compared radiofrequency ablation (RFA) to surgery
difference between the groups. Early and late complications were
with a total of 642 participants randomised. Some of the included
distributed between the groups and with very few major adverse
studies included several relevant comparison groups to surgery, and
events.
one study evaluated a comparison group not relevant to this review,
For treatment anaesthesia, four studies used general or spinal
therefore the individual comparison group participant numbers
anaesthesia for their surgical groups and local tumescent anaes-
do no add up to the overall number of randomised participants.
thesia for their laser groups. Four studies used local tumescent
anaesthesia in both treatment groups, some with light sedation
(i) Foam versus surgery administered at the discretion of the surgeon. Length of surgery
Within this comparison, rates of recurrence that were noted by was evaluated in two studies, with one reporting less time for the
clinicians and recurrence as noted by patients showed no differ- EVLT group and the other reporting more. It should be noted that
ence between the treatment groups. Recanalisation (early and late) these studies were using different definitions and metrics for mea-
and neovascularisation were only evaluated in a single study. Early surement, so any conclusion should be drawn with this in mind.
recanalisation showed no difference between the treatment groups Of the four studies reporting on this, nearly 100% of patients were
and late recanalisation favoured surgery. As no neovascularisation treated as day cases. Procedural costs were only evaluated in two
events occurred in the foam group, the analysis favoured ultra- studies and both found marginally higher costs associated with
sound-guided foam sclerotherapy (UGFS). There were no differ- laser ablation, although this is a highly subjective measure. Time
ences between the treatment groups for technical failure. Although to return to work or normal activity was reported for six studies
meta-analysis could not be carried out to compare quality of life within this comparison. Most studies showed less time needed in
(QoL) and severity scoring, both treatment groups showed very the laser group to return to work or normal activity, but as this was
similar QoL and disease improvements by the final follow-up time measured in the different studies with different metrics, and with
point. Complications were few and generally equal between the different definitions, this conclusion is tentative. The measures of
groups. However, early haematoma and saphenous nerve injury pain varied greatly between studies therefore no overall conclusion
were more frequent in the surgical group and the foam group had can be drawn at this time, but it is of importance that two robust
a higher rate of phlebitis. Major complications were very few. trials included in this comparison found increased post-operative
Two studies in this comparison reported that the surgical proce- pain in the EVLT treatment group, which is a new finding that
dures were performed under general or spinal anaesthesia, with no requires further evaluation.
indication of anaesthesia type for the foam treatment group. The
third study reported that all patients in both comparison groups
were given tumescent local anaesthesia, with light sedation at the (iii) Radiofrequency ablation (RFA) versus surgery
surgeon’s discretion. Length of procedure was only evaluated in Of the outcomes that were amenable to meta-analysis, recurrence
one study and therefore could not be compared; and only a single (noted by clinicians or patients), recanalisation (early and late),
study reported that their surgical group participants were treated neovascularisation and technical failure had no statistically signif-
as day patients, with no indication for the foam participants. For icant differences between the two treatment groups. As with the
previous treatment comparisons, QoL and disease severity scores ceived the same treatment, whereas in the Magna 2007 trial bi-
saw general improvements over the length of the follow-up for laterally treated patients could have different treatments for each
both treatment groups, with most studies reporting no overall dif- leg. EVOLVeS Study and Subramonia 2010 had a waiting period
ferences between the groups. For early complications, there were between leg treatments. It should be noted that Subramonia 2010
higher rates of haematoma and wound problems in the surgical did not end up including any bilaterally treated patients. The au-
group compared with RFA; and the increased rate of haematoma thors were contacted, where possible, but overall little could be
and saphenous vein injury in the surgical group were still evident done as this was such a widescale problem. Helmy ElKaffas 2011,
in later complications. Overall, the numbers of complications were Flessenkaemper 2013 and FOAM-Study did not indicate if bilat-
low, especially for major complications. eral patients were permitted, and did indicate the number of legs.
Three studies used general anaesthesia in surgical patients and lo- It is therefore assumed that bilaterally treated patients were not
cal tumescent anaesthesia for the RFA group. One study used local included (the authors were contacted for explanation, with no re-
tumescent anaesthesia in both treatment groups and a final study sponse). There were a total of 172 included bilateral limbs, which
had a mixture of both general and local tumescent anaesthesia in made up approximately 6% of the total randomised population.
both treatment groups. Length of procedure was evaluated in four Bilateral limb proportions ranged from 0% (Subramonia 2010)
studies with two studies reporting less time for the RFA procedure and 1% (EVOLVeS Study) up to 22% in the Magna 2007 study.
and two studies reporting more time for the RFA procedure. All High ligation and stripping (HL/S) under tumescence anaesthesia
four studies were most likely evaluating slightly different defini- was used in some of the later trials and although it is not common
tions of procedure time, which should be kept in mind as this is or routine practice it was associated with enhanced post-operative
a subjective measure. Two studies reported the comparison of in- comfort compared to EVLT in some of the trials. This may alter
patient stays versus day cases, with one reporting all surgical pro- the applicability of the evidence in regard to pain scores or other
cedures as in-patient stays and all RFA patients as day cases, while outcomes that could be altered by general versus local anaesthesia.
the other study reported over 80% of day cases in both treatment Rasmussen 2007, Rasmussen 2011 and Subramonia 2010 in-
groups. Procedural costs were similar for both treatment groups, cluded patients who had undergone previous high ligation of their
as evaluated in three studies, however one study reported slightly saphenofemoral junction (SFJ) with ultrasound evidence of resid-
higher costs in the RFA group and two reported slightly higher ual SFJ reflux. However, the number of such limbs was small and
costs in the surgical groups. All five studies reported either time was unlikely to affect the outcome variables.
to return to work or time to return to normal activities, all with An additional point that should not be forgotten is the strict in-
less time for the RFA group. It should be noted this outcome was clusion criteria used in all trials regarding patients’ suitability for
reported differently between studies, and this should be kept in endovenous treatment. When considering treatment options for
mind when drawing any conclusions. patients, if they are deemed unsuitable for endovenous manage-
ment (> 1.2 cm, tortuous veins etc) open surgery remains their
only viable option, for which there is robust, long-term evidence
Overall completeness and applicability of for its safety and efficacy.
evidence
Despite an apparent congruity in the outcome measures of the
studies (Table 4) there was a serious lack of compatible data, in- Quality of the evidence
cluding differences in the outcome definitions, metrics and follow- We were able to include 13 RCTs with a total of 3081 randomised
up time points, with which any meaningful meta-analysis could participants, with a reasonable number of studies within each treat-
be performed. This has seriously limited the overall effectiveness ment comparison. Individually the trials were on the whole well
of this Cochrane review. However, all of the included studies were designed and conducted. However, overall the quality of the avail-
considered applicable to this review and reported most if not all able evidence that was comparable was moderate due to huge vari-
of the review’s primary and secondary outcomes. The studies all ations in the reporting of results, which limited meaningful meta-
included relevant participants and interventions, and there is now analyses for the majority of proposed outcome measures.
the emergence of much needed longer-term data although cur-
rently only from a single study.
The inclusion and treatment of patients with bilateral great
saphenous varicose veins has also impeded accurate meta-anal-
Potential biases in the review process
ysis. Darwood 2008, EVOLVeS Study, Magna 2007, Pronk To limit potential bias, where uncertainty was encountered in
2010, Rasmussen 2007, Rasmussen 2011 and Subramonia 2010 aspects of the included studies, authors of the relevant papers
all permitted bilateral treatment. Darwood 2008, Pronk 2010, were contacted. Most authors who were contacted provided
Rasmussen 2007 and Rasmussen 2011 treated patients with bi- swift and full replies, which are documented in full within the
lateral veins on the same day and ensured that both limbs re- Characteristics of included studies tables for the trials. However,
some authors did not respond, which led to assumptions that could nine studies that were included, four were also included in our own
have created biases. review and five were excluded for reasons such as they were of non-
Using all patients that were randomised for the meta-analysis is conventional techniques, a non-randomised study and included
generally the objective of the intention-to-treat method. However veins other than the GSV. For the comparison between EVLT
for this review, due to the nature of the disease and intervention, it and surgery at three months there was no difference between the
was inappropriate to use all patients randomised as these numbers treatment groups but this was borderline (RR 2.19, 95% CI 0.99 to
were often very different from the numbers analysed. For this 4.85; P = 0.05). For RFA there was also no difference in recurrence
reason we used the analysed number of patients for most studies. between the two treatment groups, measured at two years (RR
Where trials included treatment of both the great saphenous vein 1.27, 95% CI 0.60 to 2.67; P = 0.53). These findings are consistent
(GSV) and small saphenous vein (SSV), the authors were con- with our own.
tacted to ascertain if any subgroup analysis had been performed A meta-analysis published in 2009 evaluated recurrence rates when
which would enable us to include the trial. Not all authors re- comparing endovenous ablation techniques versus ligation and
sponded. Of those that did, on no occasion had any author per- stripping (Xenos 2009). Differing from our review, Xenos 2009
formed such an analysis and subsequently these trials were ex- did not include any studies evaluating sclerotherapy, they required
cluded from the review. at least one year of follow-up and they allowed non-conventional
Two-year recurrence and neovascularisation figures were published techniques in both the surgical and endovenous treatment groups.
by Darwood 2008. Unfortunately this publication included both This meta-analysis included four studies that we also included
the randomised patients and patients who had refused randomi- and an additional four studies that we excluded for various rea-
sation. No subanalysis was made and subsequently no two-year sons. Instead of reporting the comparisons of EVLT and RFA ver-
follow-up data were available for the Darwood 2008 study. sus surgery separately, the primary outcome of the Xenos 2009
The selection criteria of all the included trials introduces a bias review was recurrence (by duplex ultrasound or symptoms, or
into any interpretation of these procedures for the population as a both) after endovenous treatment (EVLT or RFA) compared with
whole. Any potential advantage of these novel therapies is obviated surgery. Despite the differences they found an OR of recurrence
in patients with veins deemed ’unsuitable’ for endovascular tech- that showed no difference between the different treatment meth-
niques. In these patients surgery remains the only viable treatment ods, which does agree with our own results (OR 0.98, 95% CI
option. 0.49 to 1.96; P = 0.95).
The inclusion in some trials of patients with bilateral varicose veins The systematic review by the Society for Vascular Surgery (SVS) in
and the presentation of results by numbers of limbs introduce a collaboration with the American Venous Forum (AVF) reviewed
large potential bias. Clearly bilateral treatment has implications available evidence in 2009, comparing endovenous ablation (RFA
for length of hospital stay, pain scores, operative time and adverse and EVLT) and foam sclerotherapy versus open surgery (Eklof
events. Without stratification of results on a per patient basis this 2009). Regarding EVLT versus surgery, the findings concur with
remains a potential source of confounding bias. those summarised above and they considered that the RCTs did
not report bias protection measures and therefore that the available
evidence was of low quality. With regard to RFA versus surgery,
Agreements and disagreements with other they concluded that none of the RCTs reported allocation conceal-
studies or reviews ment or blinded outcomes and hence the quality of the evidence
was considered low.
A recent meta-analysis from 2012 evaluated the same three
Despite the lack of evidence within our review to support one
methods UGFS, EVLT and RFA compared with surgery (
method of treatment over another, the National Institute for
Siribumrungwong 2012). However, this study also evaluated com-
Health and Care Excellence (NICE) published their recommen-
parisons between UGFS, EVLT and RFA, studies that included
dation in July 2013 that endovenous treatment should constitute
non-conventional treatment methods, heterogenous populations
the first treatment of choice for people with confirmed varicose
and non-original work. The Siribumrungwong 2012 review in-
veins and truncal reflux (NICE (CG168) 2013): “If endothermal
cluded nine studies that we included and 12 studies that we
ablation is unsuitable, offer ultrasound-guided foam sclerother-
excluded. For the recurrence outcome Siribumrungwong 2012
found no difference between EVLT or RFA and surgery (EVLT apy”, “If ultrasound guided foam sclerotherapy is unsuitable, of-
versus surgery: RR 0.6, 95% CI 0.3 to 1.1; RFA versus surgery: fer surgery”. It should be noted that our Cochrane review does not
RR 0.9, 95% CI 0.6 to 1.4), which is consistent with our review. constitute a guideline nor does it offer any recommendations for
The authors were able to include complications and time to return clinicians, our conclusions are open to interpretation.
to work or normal activity for meta-analysis and found that EVLT
and RFA were associated with fewer haematomas and wound in-
fections, less pain and a quicker return to normal activities.
Brar 2010 compared EVLT and RFA techniques to surgery. Of the AUTHORS’ CONCLUSIONS
Implications for practice Regarding timing of additional mini-phlebectomies there remains
great variation in practice. We have acknowledged the limitation
Current data suggest that foam sclerotherapy and endovenous ab-
of this variation and its impact on outcomes in the included trials
lation (laser and radiofrequency) have similar overall outcomes as
and future research could focus attention on this issue.
open surgery involving high ligation and stripping (HL/S). How-
ever, these findings still lack robustness due to a paucity of com- Although we were able to include 13 RCTs in this review, the
patible data. ability to combine outcomes in meaningful ways was limited due
to incompatibility between studies. Future trials should aim to
standardise their outcomes and follow-up schedules to allow for
Implications for research
more accurate meta-analysis. This is especially true for outcomes
The long-term rates of recanalisation following foam sclerother- such as QoL, pain and procedural costs, and for clinical definitions.
apy, RFA and EVLT need to be assessed as there appears to be In addition, in future trials efforts could be made to blind the
more recanalisation in the second and third years of follow-up in post-operative procedure assessors to increase the quality of the
the included studies. While one study has reported five-year out- related outcome measures. Techniques to blind assessors could
come data, the issues of the longer-term effect of leaving a proxi- include patients wearing a dressing in the groin region, regardless
mal segment of GSV open and leaving the GSV tributaries intact of procedure type, and to restrict assessors from being able to
still remain unidentified. determine the presence or absence of a scar that would alert them
to the treatment type.
As previously discussed, there were several new studies that per-
formed HL/S under tumescent anaesthesia, with decreased post-
procedure pain in the HL/S group compared with EVLT. Further
investigation of this potential association could show a method of
ACKNOWLEDGEMENTS
maximizing post-operative comfort following this well established
therapy. It is also acknowledged that newer therapies have been CN and RB would like to thank the authors of all included and
developed that do not require the use of tumescence, for example many excluded papers who, when contacted, kindly gave up their
Clarivein™. Research comparing these therapies to surgery could time to provide details to facilitate this review. All authors would
be included in future updates of our Cochrane review, or could like to thank the staff from the Cochrane Peripheral Vascular Dis-
alternatively form the basis of a separate primary review. ease Group for their support.
REFERENCES
References to studies included in this review O, et al. Prospective randomised study of endovenous
radiofrequency obliteration (closure) versus ligation and
Darwood 2008 {published data only} vein stripping (EVOLVeS): two-year follow-up. European
Beale R, Theivacumar N, Mavor AID, Gough MJ. Journal of Vascular and Endovascular Surgery 2005;29(1):
Endovenous Laser Treatment (EVLT) or surgery of varicose 67–73.
veins? A randomised controlled trial in patients with ∗
Lurie F, Creton D, Eklof B, Kabnick LS, Kistner
saphenofemoral and long saphenous incompetence. The RL, Pichot O, et al. Prospective randomized study of
Vascular Society of Great Britain & Ireland Yearbook 2005: endovenous radiofrequency obliteration (closure procedure)
77. versus ligation and stripping in a selected patient population
∗
Darwood RJ, Theivacumar N, Dellagrammaticas D, (EVOLVeS Study). Journal of Vascular Surgery 2003;38(2):
Mavor AI, Gough MJ. Randomized clinical trial comparing 207–14.
endovenous laser ablation with surgery for the treatment of Lurie F, Creton D, Eklof B, et al. Prospective randomised
primary great saphenous varicose veins. British Journal of study of Endovenous Radiofrequency Obliteration
Surgery 2008;95(3):294–301. (Closure) Versus Ligation and Vein Stripping (EVOLVeS):
EVOLVeS Study {published data only} 2-year follow-up. Journal of Vascular Surgery 2005;42(1):
Lurie F, Creton D, Eklof B, Kabnick L, Kistner R, Pichot 178.
O, et al. Prospective randomised study of endovenous Flessenkaemper 2013 {published data only}
radiofrequency obliteration (closure) versus ligation and Flessenkaemper I, Hartmann M, Stenger D. RCT
stripping (evolves study): early results and one year follow- for differentiated open and endoluminal therapies for
up. European Society for Vascular Surgery, Programme degenerative disease of the great saphenous vein. Union
and Abstract Book, XVII Annual Meeting and Course on Internationale de Phlebologie World Congress 2009.
∗
Vascular Surgical Techniques. 2003:74–5. Flessenkamper I, Hartmann M, Stenger D, Roll S.
Lurie F, Creton D, Eklof B, Kabnick LS, Kistner RL, Pichot Endovenous laser ablation with and without high ligation
compared with high ligation and stripping in the treatment conventional surgery for great saphenous varicose veins.
of great saphenous varicose veins: Initial results of a Journal of Vascular Surgery 2013;58(3):727–34.
multicentre randomized controlled trial. Phlebology 2013; Biemans AAM, Kockaert M, Van Den Bos RR, Cuypers P,
28(1):16–23. Neumann HAM, Nijsten TEC. A randomized comparative
Flessenkamper I, Stenger D, Hartmann M, Roll S. study of the three most commonly performed treatments for
Endovenous laser therapy vs. high ligation/stripping varicose veins, results after one year. [Dutch]. Nederlands
for varicosity of the great saphenous vein: Clinical and Tijdschrift voor Dermatologie en Venereologie 2012;22(2):
sonographic findings. Phlebologie 2013;42(1):7–11. 78–84.
FOAM-Study {published data only} NCT00529672. Surgery or Noninvasive Therapy for
NCT01103258. FOAM-study, Cost Minimization Varicose Veins (Magna). http://clinicaltrials.gov/ct2/show/
Study Comparing Surgery Versus Duplex Guided Foam NCT00529672?term=NCT00529672&rank=1 (accessed
Sclerotherapy of Varicose Veins. http://clinicaltrials.gov/ June 2014) 2007.
show/NCT01103258 (accessed June 2014) 2010. Pronk 2010 {published data only}
∗
Shadid N, Ceulen R, Nelemans P, Dirksen C, Veraart Pronk P, Gauw SA, Mooij MC, Gaastra MTW, Lawson
J, Schurink GW, et al. Randomized clinical trial of JA, van Goethem AR, et al. Randomised controlled trial
ultrasound-guided foam sclerotherapy versus surgery for the comparing sapheno-femoral ligation and stripping of the
incompetent great saphenous vein. British Journal of Surgery great saphenous vein with endovenous laser ablation (980
2012;99:1062–70. nm) using local tumescent anaesthesia: One year results.
Shadid N, Nelemans P, Sommer A. Duplex guided European Journal of Vascular and Endovascular Surgery 2010;
foam sclerotherapy vs surgery for the incompetent great 40(5):649–656+e61.
saphenous vein: a randomised controlled trial. Phlebology Rasmussen 2007 {published data only}
2010;25:306–7. ISRCTN16747172. Randomized controlled study
Helmy ElKaffas 2011 {published data only} comparing endovenous laser ablation with stripping, in
Helmy ElKaffas K, ElKashef O, ElBaz. Great saphenous patients with varicose veins due to greater saphenous
vein radiofrequency ablation versus standard stripping in vein insufficiency. http://www.controlled-trials.com/
the management of primary varicose veins-a randomized ISRCTN16747172/rasmussen 2010 (accessed June 2014).
clinical trial. Angiology 2011;62(1):49–54. Lawaetz M, Bjoern L, Rasmussen L. Randomised clinical
HELP-1 {published data only} trial comparing endovenous laser ablation with stripping of
Carradice D, Mekako A, Hatfield J, Chetter I. Recurrent the great saphenous vein. Clinical outcome and recurrence
varicose veins are more common following surgery than after 5 years. Scientific Programme and Book of Abstracts
EVLT - results of a randomised controlled trial. The of the 14th Annual Meeting of the European Venous Forum
Vascular Society of Great Britain & Ireland Yearbook 2009: 2013:41.
86. Rasmussen L, Lawaetz M, Bjoern L, Blemings A, Eklof
Carradice D, Mekako AI, Hatfield J, Chetter IC. A B. Randomized clinical trial comparing endovenous laser
randomised trial of EVLT vs surgery for varicose veins. The ablation and stripping of the great saphenous vein with
Vascular Society of Great Britain & Ireland Yearbook 2008: clinical and duplex outcome after 5 years. Journal of Vascular
91. Surgery 2013;58(2):421–6.
Carradice D, Mekako AI, Mazari FA, Samuel N, Hatfield Rasmussen LH, Bjoern L, Lawaetz B, Blemings A, Eklof
J, Chetter IC. Clinical and technical outcomes from a B. Randomised clinical trial comparing endovenous laser
randomized clinical trial of endovenous laser ablation ablation with striping of the great saphenous vein: clinical
compared with conventional surgery for great saphenous outcome and recurrence after 2 years. European Journal of
varicose veins. British Journal of Surgery 2011;98(8): Vascular and Endovascular Surgery 2010;39(5):630–5.
∗
1117–23. Rasmussen LH, Bjoern L, Lawaetz M, Blemings
∗
Carradice D, Mekako AI, Mazari FA, Samuel N, Hatfield A, Lawaetz B, Eklof B. Randomized trial comparing
J, Chetter IC. Randomized clinical trial of endovenous endovenous laser ablation of the great saphenous vein with
laser ablation compared with conventional surgery for great high ligation and stripping in patients with varicose veins:
saphenous varicose veins. British Journal of Surgery 2011;98 short-term results. Journal of Vascular Surgery 2007;46(2):
(4):501–10. 308–15.
Carradice D, Wallace T, Samuel N, Gohil R, Chetter Rasmussen LH, Lawaetz M, Bjoern L, Lawaetz B, Blemings
I. A comparison of the effectiveness of treating those A, Eklof B. Medium-term follow-up of a randomised
with and without the complications of superficial venous trial comparing laser ablation with stripping of the great
insufficiency. Vascular Society of Great Britain & Ireland saphenous vein. Recurrence rate and pattern after two years.
Yearbook 2012:38. Phlebology 2009; Vol. 24:231.
Magna 2007 {published data only} Rasmussen 2011 {published data only}
∗
Biemans AA, Kockaert M, Akkersdijk GP, van den Bos ISRCTN08060326. Endovenous laser ablation (EVLA),
RR, de Maeseneer MG, Cuypers P, et al. Comparing radio frequency (RF), foam sclerotherapy and stripping
endovenous laser ablation, foam sclerotherapy, and for treatment of varicose veins. http://www.controlled-
trials.com/ISRCTN08060326/rasmussen 2009 (accessed prospective randomized trial. Phlebology. 20 2005; Vol.
June 2014). 20, issue 3:149.
Lawaetz M, Rasmussen LH, Bjoern L, Blemings A, Eklof B.
Almeida 2007 {published data only}
Randomized trial comparing RF, laser, foam sclerotherapy
Almeida J. We’ve “Got” Plenty of Data to Show that
and stripping in varicose veins. Phlebology 2010;25:307.
Endovenous Thermal Ablation is Superior to Open Surgery.
Rasmussen L. Randomised trial shows higher technical
http://www.prolibraries.com/veithsymposium/?select=
failure rate with foam. Vascular News 2010;May:33.
∗ error&conferenceID=7&sessionID=308. 2007.
Rasmussen LH, Lawaetz M, Bjoern L, Vennits B,
Blemings A, Eklof B. Randomized clinical trial comparing Alos 2006 {published data only}
endovenous laser ablation, radiofrequency ablation, foam Alos J, Carreno P, Lopez JA, Estadella B, Serra-Prat M,
sclerotherapy and surgical stripping for great saphenous Marinel-Lo J. Efficacy and safety of sclerotherapy using
varicose veins. British Journal of Surgery 2011;98(8): polidocanol foam: a controlled clinical trial. European
1079–87. Journal of Vascular and Endovascular Surgery 2006;31(1):
Rautio 2002 {published data only} 101–7.
Perala J, Rautio T, Biancari F, Ohtonen P, Wiik H,
Anon 2008 {published data only}
Heikkinen T, et al. Radiofrequency endovenous obliteration
Anon. The BRILLIANT Study. http://clinicaltrials.gov/
versus stripping of the long saphenous vein in the
ct2/show/NCT00618514?term=endovenous&rank=5
management of primary varicose veins: 3-year outcome of a
2008.
randomized study. Annals of Vascular Surgery 2005;19(5):
669–72. Belcaro 2000 {published data only}
∗
Rautio T, Ohinmaa A, Perala J, Ohtonen P, Heikkinen T, Belcaro G, Nicolaides AN, Ricci A, Dugall M, Errichi
Wiik H, et al. Endovenous obliteration versus conventional BM, Vasdekis S, et al. Endovascular sclerotherapy, surgery,
stripping operation in the treatment of primary varicose and surgery plus sclerotherapy in superficial venous
veins: a randomized controlled trial with comparison of the incompetence: a randomized, 10-year follow-up trial--final
costs. Journal of Vascular Surgery 2002;35(5):958–65. results. Angiology 2000;51(7):529–34.
RELACS Study {published data only} Blaise 2010 {published data only}
Rass K, Frings N, Glowacki P, Hamsch C, Graber S, Vogt T, Anon. ESGVS: Sclerotherapy With Lauromacrogol. Study
et al. Comparable effectiveness of endovenous laser ablation NCT00348764 Information provided by University
and high ligation with stripping of the great saphenous vein: Hospital, Grenoble 2006.
two-year results of a randomized clinical trial (RELACS ∗
Blaise S, Bosson JL, Diamand JM. Ultrasound-guided
study). Archives of Dermatology 2012;148(1):49–58. sclerotherapy of the great saphenous vein with 1% vs. 3%
Subramonia 2010 {published data only} polidocanol foam: a multicentre double-blind randomised
Balakrishnan A, Mylankal K, Nalachandran S, Subramonia trial with 3-year follow-up. European Journal of Vascular and
S, Lees T. A randomized controlled trial of radiofrequency Endovascular Surgery 2010;39(6):779–86.
ablation and conventional surgery for primary long
saphenous varicose veins. Phlebology 2008;23(4):198. BLARA Trial {published data only}
Subramonia S, Lees T. Radiofrequency ablation vs Goode S. A randomised study comparing endovenous laser
conventional surgery for varicose veins - a comparison of ablation and radiofrequency ablation for the treatment
treatment costs in a randomised trial. European Journal of of varicose veins. European Society for Vascular Surgery
Vascular and Endovascular Surgery 2010;39:104–11. Annual Meeting; 2008 Sep 4-7; Nice, France. 208:65.
∗
Subramonia S, Lees T. Randomized clinical trial of Goode S, Crockett M, Richards T, Braithwaite B.
radiofrequency ablation or conventional high ligation and Preliminary results for the bilateral laser and radiofrequency
stripping for great saphenous varicose veins. British Journal ablation study. Phlebology 2008;23(4):196–7.
∗
of Surgery 2010;97(3):328–36. Goode S, Crockett M, Richards T, Braithwaite B,
Chowdhury A. The bilateral laser and radiofrequency
References to studies excluded from this review ablation (BLARA) trial. Phlebology 2008;23(5):241.
Abela 2008 {published data only} Bountouroglou 2006 {published data only}
∗
Abela R, Liamis A, Prionidis I, Mathai J, Gorton L, Bountouroglou DG, Azzam M, Kakkos SK, Pathmarajah
Browne T, et al. Reverse foam sclerotherapy of the great M, Young P, Geroulakos G. Prospective randomized study
saphenous vein with sapheno-femoral ligation compared to of ultrasound-guided foam sclerotherapy and adjuvant high
standard and invagination stripping: a prospective clinical tie under local anaesthesia versus conventional surgery for
series. European Journal of Vascular and Endovascular Surgery primary varicose veins: early results. Phlebology. 19 2004;
2008;36(4):485–90. Vol. 19, issue 3:151.
∗
Liamis A, Prionidis I, Mathai J, Gorton L, Browne T, Bountouroglou DG, Azzam M, Kakkos SK, Pathmarajah
Panayiotopoulos YP. Long saphenous vein reverse foam M, Young P, Geroulakos G. Ultrasound-guided foam
sclerotherapy with saphenofemoral junction ligation sclerotherapy combined with sapheno-femoral ligation
compared with head and invagination stripping: a compared to surgical treatment of varicose veins: early
results of a randomised controlled trial. European Journal of Desmyttere 2005 {published data only}
Vascular and Endovascular Surgery 2006;31(1):93–100. Desmyttere J, Grard C, Mordon S. A 2 year follow-up
Kalodiki E, Azzam M, Kakkos SK, Zambas N, study of endovenous 980 nm laser treatment of the great
Bountouroglou D, Geroulakos G. Ultrasound-guided foam saphenous vein: Role of the blood content in the GSV.
sclerotherapy combined with saphenofemoral ligation versus Medical Laser Application 2005;20(4):283–9.
surgical treatment of varicose veins: intermediate results of a Disselhoff 2008 {published data only}
randomized controlled trial. Phlebology 2008;23(5):242–3. Disselhoff BC, Buskens E, Kelder JC, der Kinderen DJ,
Bush 2008 {published data only} Moll FL. Randomised comparison of costs and cost-
Bush RG, Shamma HN, Hammond K. Histological changes effectiveness of cryostripping and endovenous laser ablation
occurring after endoluminal ablation with two diode lasers for varicose veins: 2-year results. European Journal of
(940 and 1319 nm) from acute changes to 4 months. Lasers Vascular and Endovascular Surgery 2009;37(3):357–63.
in Surgery and Medicine 2008;40(10):676–9. Disselhoff BC, der Kinderen DJ, Kelder JC, Moll FL.
Five-year results of a randomized clinical trial comparing
Ceulen 2007 {published data only} endovenous laser ablation with cryostripping for great
Ceulen RP, Bullens-Goessens YI, Pi-Van de Venne SJ, saphenous varicose veins. British Journal of Surgery 2011;98
Nelemans PJ, Veraart JC, Sommer A. Outcomes and side (8):1107–11.
effects of duplex-guided sclerotherapy in the treatment of Disselhoff BC, der Kinderen DJ, Kelder JC, Moll FL.
great saphenous veins with 1% versus 3% polidocanol foam: Randomized clinical trial comparing endovenous laser
results of a randomized controlled trial with 1-year follow- ablation of the great saphenous vein with and without
up. Dermatologic Surgery 2007;33(3):276–81. ligation of the sapheno-femoral junction: 2-year results.
European Journal of Vascular and Endovascular Surgery 2008;
Chant 1972 {published data only}
36(6):713–8.
Beresford SA, Chant AD, Jones HO, Piachaud D, Weddell ∗
Disselhoff BC, der-Kinderen DJ, Kelder JC, Moll FL.
JM. Varicose veins: A comparison of surgery and infection/
Randomized clinical trial comparing endovenous laser with
compression sclerotherapy. Five-year follow-up. Lancet
cryostripping for great saphenous varicose veins. British
1978;1(8070):921–4.
∗ Journal of Surgery 2008;95(10):1232–8.
Chant AD, Jones HO, Weddell JM. Varicose veins:
a comparison of surgery and injection/compression Disselhoff 2011 {published data only}
sclerotherapy. Lancet 1972;2(7788):1188–91. Disselhoff BCVM, der Kinderen DJ, Kelder JC, Moll FL.
Five-year results of a randomised clinical trial of endovenous
Christenson 2010 {published data only} laser ablation of the great saphenous vein with and without
Christensen JT, Gueddi S, Gemayel G, Bounameaux ligation of the saphenofemoral junction. European Journal
H. Prospective randomized trial comparing endovenous of Vascular and Endovascular Surgery 2011;41(5):685–90.
laser ablation and surgery for treatment of primary great
saphenous varicose veins with a 2-year follow-up. Journal of Doran 1975 {published data only}
Vascular Surgery 2010;52(5):1234–41. Doran FS, White M. A clinical trial designed to discover if
the primary treatment of varicose veins should be by Fegan’s
Compagna 2010 {published data only} method or by an operation. British Journal of Surgery 1975;
Compagna R, De Vito D, Rossi R, Fappiano F, De Magistris 62(1):72–6.
L, Sodano L, et al. A prospective randomised controlled Duffy 2005 {published data only}
trial comparing foam sclerotherapy combined with sapheno- Duffy D. Double-blind prospective comparative trial
femoral ligation to surgical treatment of varicose veins. between foamed and liquid polidocanol and sodium
European Surgical Research 2010;45(3-4):267. tetradecyl sulfate in the treatment of varicose and
Demagny 2002 {published data only} telangiectatic leg veins: Commentary. Dermatologic Surgery
Demagny A. Comparative study into the efficacy of 2005;31(6):635.
a sclerosant product in the form of liquid or foam in Einarsson 1993 {published data only}
echo-guided sclerosis of the arches of the long and short Einarsson E, Eklof B, Neglen P. Sclerotherapy or surgery
saphenous veins. Phlébologie 2002;55(2):133–7. as treatment for varicose veins: a prospective randomized
study. Phlebology 1993;8:22–6.
De Medeiros 2006 {published data only}
∗
De Medeiros CA, Luccas GC. Comparison of endovenous Figueiredo 2009 {published data only}
treatment with an 810 nm laser versus conventional Figueiredo M, Araujo S, Barros N Jr, Miranda F Jr. Results
stripping of the great saphenous vein in patients with of surgical treatment compared with ultrasound-guided
primary varicose veins. Dermatologic Surgery 2005;31(12): foam sclerotherapy in patients with varicose veins: a
1685–94. prospective randomised study. European Journal of Vascular
De Medeiros CAF. Comparison of endovenous laser therapy and Endovascular Surgery 2009 Dec;38(6):758–63.
vs. conventional stripping of the great saphenous vein: Gale 2009 {published data only}
Midterm results. Jornal Vascular Brasileiro 2006;5(4): Gale SS, Lee JN, Walsh ME, Wojnarowksi DL, Comerota
277–87. AJ. A randomized trial of endovenous thermal ablation for
superficial venous insufficiency of the great saphenous vein: Hayes 2008 {published data only}
radiofrequency ablation versus endovenous laser therapy. Hayes CA, Kingsley JR, Hamby KR, Carlow J. The effect
Phlebology 2009;24:231. of endovenous laser ablation on restless legs syndrome.
Gale SS, Lee JN, Walsh ME, Wojnarowski DL, Comerota Phlebology 2008;23(3):112–7.
AJ. A randomized, controlled trial of endovenous thermal HELP-2 {published data only}
ablation using the 810-nm wavelength laser and the Samuel N, Carradice D, Mekako AI, Wallace T, Hatfield J,
ClosurePLUS radiofrequency ablation methods for Chetter IC. Randomised trial of endovenous laser ablation
superficial venous insufficiency of the great saphenous vein. versus surgery for small saphenous varicose veins. The
Journal of Vascular Surgery 2010;52(3):645–50. Vascular Society of Great Britain & Ireland Yearbook 2011:
82.
Goode 2010 {published data only}
Samuel N, Carradice D, Wallace T, Mekako A, Hatfield J,
Goode SD, Chowdhury A, Crockett M, Beech A, Simpson
Chetter I. Randomized clinical trial of endovenous laser
R, Richards T, et al. Laser and radiofrequency ablation
ablation versus conventional surgery for small saphenous
study (LARA study): a randomised study comparing
varicose veins. Annals of Surgery 2013;257(3):419–26.
radiofrequency ablation and endovenous laser ablation (810
Samuel N, Mazari F, Wallace T, Carradice D, Chetter I.
nm). European Journal of Vascular and Endovascular Surgery
Economic analysis of evla and surgery in the treatment
2010;40(2):246–53.
of small saphenous incompetence. Results from an RCT.
Hamel-Desnos 2003 {published data only} European Surgical Research 2013;50 Suppl 1:25.
∗
Hamel-Desnos C, Desnos P, Wollmann JC, Ouvry Hinchliffe 2006 {published data only}
P, Mako S, Allaert FA. Evaluation of the efficacy of Hinchliffe RJ, Ubhi J, Beech A, Ellison J, Braithwaite
polidocanol in the form of foam compared with liquid BD. A prospective randomised controlled trial of VNUS
form in sclerotherapy of the greater saphenous vein: initial closure versus surgery for the treatment of recurrent long
results. Dermatologic Surgery 2003;29(12):1170–5. saphenous varicose veins. European Journal of Vascular and
Hamel-Desnos C, Ouvry P, Desnos P, Mako S. Evaluation of Endovascular Surgery 2006 Sept;31(2):212–8.
the efficacy of polidocanol in the form of foam versus liquid Hobbs 1968 {published data only}
form in sclerotherapy of the long saphenous vein. 16th Hobbs JT. Surgery and sclerotherapy in the treatment of
Annual Congress of the American College of Phlebology, varicose veins. A random trial. Archives of Surgery 1974;
2002 Nov 7-10; Fort Lauderdale, Florida. 2002. 109:793–6.
Hamel-Desnos 2005 {published data only} Hobbs JT. Surgery or sclerotherapy for varicose veins. 10
∗
Hamel-Desnos C, Allaert FA, Benigni JP, Boitelle G, year results of a random trial. In: Tesi M, Dormandy JA
Chleir F, Ouvry P, et al. Study 3/1. Polidocanol foam editor(s). Superficial and Deep Venous Diseases of the Lower
3% versus 1% in the great saphenous vein: early results. Limbs/ Surgical Therapy. Edizioni Panminerva Medica,
Phlébologie 2005;58(2):165–73. 1984:243–8.
∗
Hamel-Desnos C, Ouvry P, Benigni J-P, Boitelle G, Hobbs JT. The treatment of varicose veins. A random
Schadeck M, Desnos P, et al. Comparison of 1% and 3% trial of injection-compression therapy versus surgery. British
polidocanol foam in ultrasound guided sclerotherapy of the Journal of Surgery 1968;55(10):777–80.
great saphenous vein: A randomised, double-blind trial Jia 2010 {published data only}
with 2 year-follow-up. “The 3/1 Study”. European Journal Jia X, Liu XP, Xiong J, Zhang HP, Liu M, DU X, et al.
of Vascular and Endovascular Surgery 2007;34(6):723–9. [Foam sclerotherapy of the great saphenous vein with
Hamel-Desnos C, Ouvry P, Desnos P, Allaert FA. 3% sapheno-femoral ligation compared to standard stripping:
versus 1% polidocanol foam in sclerotherapy of the great a prospective randomized controlled trial]. [Chinese].
saphenous vein: randomised double-blind controlled trial Chung-Hua Wai Ko Tsa Chih [Chinese Journal of Surgery]
with a 2-year follow-up. Study 3/1. Phlébologie 2008;61(1): 2010;48(22):1731–4.
103–9. Kabnick 2003 {published data only}
Kabnick L. Comparison of 980nm and 810nm diode
Hamel-Desnos 2008 {published data only}
endovenous ablation lasers for treatment of the great
Hamel-Desnos CM, Guias B, Quehe P, Mesgard A, Desnos
saphenous vein. Abstracts from the UIP World Congress
P, Ferre B, et al. Foam sclerotherapy of saphenous veins with
Chapter Meeting, 2003 Aug 27-31; San Diego, California.
or without compression stockings. Biological consequences
2003.
on circulating blood - a controlled randomized study. ∗
Kabnick LS. Outcome of different endovenous laser
Phlebology 2008;23(5):245–6.
wavelengths for great saphenous vein ablation. Journal of
Hamel-Desnos 2009 {published data only} Vascular Surgery 2006;43(1):88–93.
Hamel-Desnos CM, Gillet JL, Desnos PR, Allaert FA. Kalodiki 2012 {published data only}
Sclerotherapy of varicose veins in patients with documented Kalodiki E, Lattimer CR, Azzam M, Shawish E,
thrombophilia: a prospective controlled randomized study Bountouroglou D, Geroulakos G. Long-term results of a
of 105 cases. Phlebology 2009;24(4):176–82. randomized controlled trial on ultrasound-guided foam
sclerotherapy combined with saphenofemoral ligation vs endovenous laser ablation of the great saphenous vein: a
standard surgery for varicose veins. Journal of Vascular randomized study. Phlebology 2009;24(4):151–6.
Surgery 2012;55(2):451–7.
Lupton 2002 {published data only}
Kalteis 2008 {published data only} Lupton JR, Alster TS, Romero P, Lupton JR, Alster TS,
Kalteis M, Berger I, Messie-Werndl S, Pistrich R, Schimetta Romero P. Clinical comparison of sclerotherapy versus
W, Polz W, et al. High ligation combined with stripping long-pulsed Nd:YAG laser treatment for lower extremity
and endovenous laser ablation of the great saphenous vein: telangiectases. Dermatologic Surgery 2002;28(8):694–7.
Early results of a randomized controlled study. Journal of
Vascular Surgery 2008;47(4):822-9, discussion 829. Martimbeau 2003 {published data only}
Martimbeau PR. A randomized clinical trial comparing the
Kern 2005 {published data only}
sclerosing and side effects of foam vs. liquid formula for
Kern P. Foam sclerotherapy of telangiectases: a controlled
sclerotherapy of primary varicose veins. Abstracts from the
study. Phlébologie 2005;58(2):183–7.
∗
UIP World Congress Chapter Meeting, 2003 Aug 27-31;
Kern P, Ramelet AA, Wutschert R, Bounameaux H, Hayoz
San Diego, California. 2003.
D. Single-blind, randomized study comparing chromated
glycerin, polidocanol solution, and polidocanol foam for Maurins 2009 {published data only}
treatment of telangiectatic leg veins. Dermatologic Surgery Maurins U, Rabe E, Pannier F. Does laser power influence
2004;30(3):367–72. the results of endovenous laser ablation (EVLA) of
Kuznetsov 2005 {published data only} incompetent saphenous veins with the 1 470-nm diode
Kuznetsov NA, Barinov VE, Teleshov BV, Trepilets VE, laser? A prospective randomized study comparing 15 and
Zheltikov AN. [Surgical technologies in the treatment of 25 W. International Angiology 2009;28(1):32–7.
varicose veins of the lower extremities complicated with McDaniel 1999 {published data only}
trophic ulcer]. Khirurgiia 2005, (4):37–41. McDaniel DH, Ash K, Lord J, Newman J, Adrian
Lattimer 2012 {published data only} RM, Zukowski M. Laser therapy of spider leg veins:
Lattimer CR, Azzam M, Kalodiki E, Shawish E, Trueman clinical evaluation of a new long pulsed alexandrite laser.
P, Geroulakos G. Cost and effectiveness of laser with Dermatologic Surgery 1999;25(1):52–8.
phlebectomies compared with foam sclerotherapy in
Mekako 2006 {published data only}
superficial venous insufficiency. Early results of a
ISRCTN56883670. EVLT for Sapheno-Popliteal
randomised controlled trial. European Journal of Vascular
incompetence and SS Reflux: A RCT. http://
and Endovascular Surgery 2012;43(5):594–600.
www.controlled-trials.com/ISRCTN56883670/
Lattimer CR, Kalodiki E, Azzam M, Makris GC, et al.
EVLT+for+sapheno-popliteal (accessed June 2014) 2006.
Interim results on abolishing reflux alongside a randomized
clinical trial on laser ablation with phlebectomies versus Mekako 2007 {published data only}
foam sclerotherapy. International Angiology 2013;32(4): Carradice D, Mekako A, McCollum C, Chetter I.
394–403. Endovenous laser therapy with concomitant or sequential
Lin 2007 {published data only} phlebectomy: A randomised controlled trial. European
Lin Y, Ye CS, Huang XL, Ye JL, Yin HH, Wang SM. A Society for Vascular Surgery Annual Meeting, 2008 Sep 4-
random, comparative study on endovenous laser therapy 7; Nice, France. 2008:67.
and saphenous veins stripping for the treatment of great Carradice D, Mekako AI, Hatfield J, Chetter IC.
saphenous vein incompetence. Chung-Hua i Hsueh Tsa Randomized clinical trial of concomitant or sequential
Chih [Chinese Medical Journal] 2007;87(43):3043–6. phlebectomy after endovenous laser therapy for varicose
Lin 2009 {published data only} veins. British Journal of Surgery 2009;96(4):369–75.
∗
Lin SM, Zhang ZH, Yao YD, Xiao JB, Lin Shao Mang, Mekako AI, Hatfield J, Abdul Rahman MNA, Gulati S,
Zhang Zhi Hui, et al. Experience of endovenous McCollum PT, Chetter IC. Endovenous laser therapy with
radiofrequency combined with TriVex in treatment of concomitant or sequential phlebectomy: a randomized
chronic venous insufficiency in lower extremity. Chung- controlled trial. The Vascular Society of Great Britain &
Hua Wai Ko Tsa Chih [Chinese Journal of Surgery] 2009;47 Ireland Yearbook. 2007:72.
(4):271–4. NCT00841178 {published data only}
Liu 2011 {published data only} NCT00841178. Endovenous Laser Therapy (EVLT) for
Liu X, Jia X, Guo W, Xiong J, Zhang H, Liu M, et Sapheno-Popliteal Incompetence and Short Saphenous Vein
al. Ultrasound-guided foam sclerotherapy of the great (SSV) Reflux: A RCT. http://clinicaltrials.gov/ct2/show/
saphenous vein with sapheno-femoral ligation compared to NCT00841178 (accessed June 2014).
standard stripping: a prospective clinical study. International Neglen 1993 {published data only}
Angiology 2011;30(4):321–6. Neglen P, Einarsson E, Eklof B. The functional long-term
Lugli 2009 {published data only} value of different types of treatment for saphenous vein
Lugli M, Cogo A, Guerzoni S, Petti A, Maleti O. Effects incompetence. Journal of Cardiovascular Surgery 1993;34:
of eccentric compression by a crossed-tape technique after 295–301.
Ogawa 2008 {published data only} treatment of the incompetent greater saphenous vein.
Ogawa T, Hoshino S, Makimura S, Shigematsu H, Azuma American Journal of Surgery 1994;168(4):311–5.
N, Sasajima T, et al. Endovenous laser ablation compared
Rybak 2003 {published data only}
with stripping - multicenter RCT in Japan. Presented at
Rybak Z, Garcarek J. Aethoxysclerol foam obliteration of
the 2008 American Venous Forum’s annual meeting in
insufficient perforating veins in patients suffering from leg
Charleston, SC, USA,. 2008:130–P25.
ulcers: a clinical recommendation. Abstracts from the UIP
Ouvry 2008 {published data only} World Congress Chapter Meeting, 2003 Aug 27-31; San
Ouvry P, Allaert FA, Desnos P, Hamel-Desnos C. Efficacy of Diego, California. 2003.
polidocanol foam versus liquid in sclerotherapy of the great
saphenous vein: a multicentre randomised controlled trial Sadoun 2003 {published data only}
with a 2-year follow-up. European Journal of Vascular and Sadoun S. Long-term follow-up study of ultrasound
Endovascular Surgery 2008;36(3):366–70. findings in varicose greater saphenous veins treated with
foam. Abstracts from the UIP World Congress Chapter
Rabe 2008 {published data only} Meeting, 2003 Aug 27-31; San Diego, California. 2003.
Rabe E, Otto J, Schliephake D, Pannier F. Efficacy
and safety of great saphenous vein sclerotherapy using Seddon 1973 {published data only}
standardised polidocanol foam (ESAF): a randomised Seddon J. The management of varicose veins. British
controlled multicentre clinical trial. European Journal of Journal of Surgery 1973;60(5):345–7.
Vascular and Endovascular Surgery 2008;35(2):238–45. Selles 2008 {published data only}
Rao 2005 {published data only} Selles DR, Arenas RJ, Ballester IC, Perez MJ, Gonzalez VS,
Rao J, Wildemore JK, Goldman MP. Double-blind Ruiz del CJ. [Phlebectomies or foam sclerosis for treatment
prospective comparative trial between foamed and liquid of the distal venous segment in the 3-S saphenectomy
polidocanol and sodium tetradecyl sulfate in the treatment technique]. [Spanish]. Cirugia Espanola 2008;84(2):92–9.
of varicose and telangiectatic leg veins. Dermatologic Surgery Shepherd 2009 {published data only}
2005;31(6):631–5. Shepherd AC, Gohel MS, Brown LC, Davies AH.
Rasmussen 2010 {published data only} Randomised clinical trial comparing VNUS Closure Fast
ISRCTN91735422. Comparing endovascular laser, versus laser for varicose veins (VALVV): Duplex and quality
ultrasound guided foam sclerotherapy, and conventional of life outcomes at 6 months. Phlebology 2010;25(2):305–6.
surgery for the treatment of small saphenous varicose veins. Shepherd AC, Gohel MS, Brown LC, Metcalfe MJ, Hamish
http://www.controlled-trials.com/ISRCTN91735422/ M, Davies AH. Early results of a randomised clinical trial
rasmussen 2010 (accessed June 2014). (RCT) comparing VNUS© ClosureFASTT Ablation and
Laser for Varicose Veins (VALVV). The Vascular Society of
REACTIV {published data only}
Great Britain & Ireland Yearbook 2009:87.
Michaels JA, Campbell WB, Brazier JE, MacIntyre JB,
Shepherd AC, Gohel MS, Brown LC, Metcalfe MJ, Hamish
Palfreyman SJ, Ratcliffe J, et al. Randomised clinical trial,
M, Davies AH. Randomized clinical trial of VNUS
observational study and assessment of cost-effectiveness of
ClosureFAST radiofrequency ablation versus laser for
the treatment of varicose veins (REACTIV trial). [Review]
varicose veins. British Journal of Surgery 2010;97(6):810–8.
[220 refs]. Health Technology Assessment (Winchester,
England) 2006;10(13):1-196, iii-iv. Sica 2006 {published data only}
Sica M, Biasi G, George E. A simplified method of
RECOVERY Trial {published data only}
∗ endovenous injection of 1% trombovar microfoam via
Almeida JI, Kaufman J, Gockeritz O, Chopra P, Evans
a short catheter (m.s. method): Lasting effectiveness
MT, Hoheim DF, et al. Radiofrequency endovenous
confirmed by 2 year results [Méthode simplifiée d’injection
ClosureFAST® versus laser ablation for the treatment
endoveineuse de micromousee de Trombovar® à 1%
of great saphenous reflux: a multicenter, single-blinded,
par cathéter court (méthode MS): une efficacité durable
randomized study (RECOVERY study). Journal of Vascular
confirmée par les résultats à 2 ans]. Phlébologie 2006;59(4):
and Interventional Radiology 2009;20(6):752–9.
339–42.
Anon. RECOVERY Trial; Final Results. http://vnus.fr/pdf/
Recovery-Data.pdf 2009. Stotter 2005 {published data only}
Kaufman JA, Williamson JA. Radiofrequency endovenous Stotter L, Schaaf I, Bockelbrink A, Baurecht HJ.
ClosureFAST® versus laser ablation for the treatment Radiofrequency obliteration, invagination or cryostripping:
of great saphenous reflux: A multicenter, single-blinded, Which is the best tolerated treatment by the patients?.
randomized study (RECOVERY). 34th Annual Scientific Phlébologie 2005;34(1):19–24.
Meeting of the Society of Interventional Radiology, 2009
Theivacumar 2008 {published data only}
Mar 7-12; San Diego, USA. 2009:Abstract no: 22. ∗
Theivacumar NS, Dellagrammaticas D, Mavor AI, Gough
Rutgers 1994 {published data only} MJ. Endovenous laser ablation: does standard above-knee
Rutgers PH, Kistlaar PJ. Randomized trial of stripping great saphenous vein ablation provide optimum results
versus high ligation combined with sclerotherapy in the in patients with both above- and below-knee reflux? A
randomized controlled trial. Journal of Vascular Surgery sclerotherapy for varicose veins. European Journal of Vascular
2008;48(1):173–8. and Endovascular Surgery 2009;37(3):343–8.
∗
Theivacumar SN, Dellagrammaticas D, Darwood RJ, Yamaki T, Nozaki M, Sakurai H, Takeuchi M, Soejima
Mavor AI, Gough MJ. Endovenous laser ablation (EVLA): K, Kono T. Prospective randomized efficacy of ultrasound-
is standard above-knee great saphenous vein (AK-GSV) guided foam sclerotherapy compared with ultrasound-
ablation sufficient? A randomised controlled trial. The guided liquid sclerotherapy in the treatment of symptomatic
Vascular Society of Great Britain & Ireland Yearbook 2007: venous malformations. Journal of Vascular Surgery 2008;47
71. (3):578–84.
Theivacumar 2009 {published data only} Yang 2013 {published data only}
Theivacumar NS, Darwood R, Gough MJ. Yang L, Wang XP, Su WJ, Zhang Y, Wang Y. Randomized
Neovascularisation and recurrence 2 years after varicose clinical trial of endovenous microwave ablation combined
vein treatment for sapheno-femoral and great saphenous with high ligation versus conventional surgery for varicose
vein reflux: a comparison of surgery and endovenous laser veins. European Journal of Vascular and Endovascular Surgery
ablation. European Journal of Vascular and Endovascular 2013;46(4):473–9.
Surgery 2009;38(2):203–7.
Zeh 2003 {published data only}
VEDICO Trial {published data only} Zeh RG. Expanding sclerosing properties of polidocanol
Belcaro G, Cesarone M, Di Renzo A, Brandolini R, foam with Gelofusine®. Abstracts from the UIP World
Coen L, Acerbi G, et al. Treatments for varicose veins: Congress Chapter Meeting, 2003 Aug 27-31; San Diego,
Surgery, sclerotherapy, foamsclerotherapy and combined California. 2003.
(surgery+sclerotherapy) options. A 10-year, prospective,
randomised, controlled, follow-up study. The VEDICO* References to ongoing studies
trial and EST (European Sclerotherapy Trial). Angeiologie
2003;55(1):29–36. CLASS {published data only}
∗
Belcaro G, Cesarone MR, Di Renzo A, Brandolini R, Anon. The CLASS trial. Comparison of laser, surgery
Coen L, Acerbi G, et al. Foam-sclerotherapy, surgery, and foam sclerotherapy. Research protocol. NIHR Health
sclerotherapy, and combined treatment for varicose veins: Technology Assessment Programme 2010.
a 10-year, prospective, randomized, controlled, trial ISRCTN51995477. Randomised controlled trial
(VEDICO trial). Angiology 2003;54(3):307–15. comparing foam sclerotherapy, alone or in combination with
endovenous laser therapy, with conventional surgery as a
Viarengo 2007 {published data only}
treatment for varicose veins. CLASS (Comparison of LAser,
Viarengo LM, Potério-Filho J, Potério GM, Menezes FH,
Surgery and foam Sclerotherapy). http://www.controlled-
Meirelles GV. Endovenous laser treatment for varicose veins
trials.com/ISRCTN51995477 (accessed June 2014).
in patients with active ulcers: measurement of intravenous
and perivenous temperatures during the procedure. Desai 2009 {published data only}
Dermatologic Surgery 2007;33(10):1234-42, discussion Desai DJ, Pedgoankar P, Sekhar R. Endovascular venous
1241-2. laser vs Oesch pin stripper in management of primary
Vuylsteke 2006 {published data only} varicosities of the great saphenous vein: a randomized
Vuylsteke M, Van den Bussche D, Audenaert E A, Lissens P. control trial. ANZ Journal of Surgery 2009; Vol. 79, issue
Endovenous laser obliteration for the treatment of primary S1:A89.
varicose veins. Phlebology 2006;21:80–7. RAFPELS Trial {published data only}
NCT00621062. Prospective Randomized Trial Comparing
Wright 2006 {published data only}
the New Endovenous Procedures Versus Conventional
Wright D. European randomized controlled trial of
Surgery for Varicose Veins Due to Great Saphenous Vein
Varisolve® PD microfoam compared with alternative
Incompetence (RAFPELS). http://clinicaltrials.gov/ct2/
therapy in management of moderate-to-severe varicose
show/NCT00621062 (accessed June 2014).
veins: preliminary results. Abstracts from the UIP World
Congress Chapter Meeting, 2003 Aug 27-31; San Diego, Venermo 2013 {published data only}
California. 2003. NCT01298908. Comparison of treatments in venous
∗
Wright D, Gobin JP, Bradbury AW, Coleridge-Smith insufficiency. http://www.clinicaltrials.gov/ct2/show/
P, Spoelstra H, Berridge D, et al. Varisolve polidocanol NCT01298908 (accessed May 2014).
microfoam compared with surgery or sclerotherapy in the Venermo M. A randomised trial of three VV treatment
management of varicose veins in the presence of trunk strategies. ESVS Budapest 2013.
vein incompetence: European randomized controlled trial. Wilhelmi 2009 {published data only}
Phlebology 2006;21(4):180–90.
Wilhelmi M, Hoch A, Hoch W, Pichlmaier AM, Teebken
Yamaki 2008 {published data only} OE, Bisdas T, et al. Effectiveness and clinical outcome
Yamaki T, Nozaki M, Sakurai H, Takeuchi M, Soejima following endovenous therapy of primary varicose veins:
K, Kono T. Multiple small-dose injections can reduce the first results of a randomised, prospective study comparing
passage of sclerosant foam into deep veins during foam the vnus closure fast system, 980nm and 1470nm lasers and
conventional surgery. XVI World Congress of the Union Jones 1996
Internationale de Phlebologie Abstract Book 2009. Jones L, Braithwaite BD, Selwyn D, Cooke S, Earnshaw
JJ. Neovascularisation is the principle cause of varicose
Additional references vein recurrence: results of a randomised trial of stripping
the long saphenous vein. European Journal of Vascular and
Baccaglini 1997
Endovascular Surgery 1996;12(4):442–5.
Baccaglini U, Spreafico G, Castoro C, Sorrentino P.
Consensus conference on sclerotherapy or varicose veins of Min 2003
the lower limbs. Phlebology 1997;12(1):2–16. Min RJ, Khilnani N, Zimmet SE. Endovenous laser
Bergan 2002 treatment of saphenous vein reflux: long-term results.
Bergan JJ, Kumins NH, Owens EL, Sparks SR. Surgical Journal of Vascular and Interventional Radiology 2003;14(8):
and endovascular treatment of lower extremity venous 991–6.
insufficiency. Journal of Vascular and Interventional Navarro 2001
Radiology 2002;13(6):563–8. Navarro L, Min RJ, Bone C. Endovenous laser: a new
Brar 2010 minimally invasive method of treatment for varicose veins-
Brar R, Nordon IM, Hinchliffe RJ, Loftus IM, Thompson -preliminary observations using an 810 nm diode laser.
MM. Surgical management of varicose veins: meta-analysis. Dermatologic Surgery 2001;27(2):117–22.
Vascular 2010;18(4):205–20. NICE (CG168) 2013
Critchley 1997 Anon. Varicose veins in the legs: The diagnosis and
Critchley G, Handa A, Maw A, Harvey A, Harvey MR, management of varicose veins. National Instititute
Corbett CR. Complications of varicose vein surgery. Annals for Health and Care Excellence (NICE) 2013 http://
of the Royal College of Surgeons of England 1997;79(2): www.nice.org.uk/guidance/CG168 (accessed 14 July 2014).
105–10.
Pichot 2004
Darke 1992
Pichot O, Kabnick LS, Creton D, Merchant RF, Schuller-
Darke SG. The morphology of recurrent varicose veins. Petroviae S, Chandler JG. Duplex ultrasound scan findings
European Journal of Vascular Surgery 1992;6(5):512–7.
two years after great saphenous vein radiofrequency
Dwerryhouse 1999 endovenous ablation. Journal of Vascular Surgery 2004;39
Dwerryhouse S, Davies B, Harradine K, Earnshaw JJ. (1):189–95.
Stripping the long saphenous vein reduces the rate of
Rautio 2002a
reoperation for recurrent varicose veins: five-year results of a
Rautio T, Ohinmaa A, Perala J, Ohtonen P, Heikkenen T,
randomized trial. Journal of Vascular Surgery 1999;29(4):
Wiik H, et al. Endovenous obliteration versus conventional
589–92.
stripping operation in the treatment of primary varicose
Eklof 2009 veins: a randomized controlled trial with comparison of the
Eklof B. Fire, foam, and knife for varicose veins: what costs. Journal of Vascular Surgery 2002;35(5):958–65.
have prospective randomized clinical trials taught us?.
VEITHsymposium 2009. Sarin 1992
Sarin S, Scurr JH, Coleridge Smith PD. Assessment of
Fischer 2002
stripping the long saphenous vein in treatment of primary
Fischer R, Chandler JG, De Maeseneer MG, Frings N,
varicose veins. British Journal of Surgery 1992;79(9):
Lefebvre-Vilarbedo M, Earnshaw JJ, et al. The unresolved
889–93.
problem of recurrent saphenofemoral reflux. Journal of the
American College of Surgeons 2002;195(1):80–94. Siribumrungwong 2012
Siribumrungwong B, Noorit P, Wilasrusmee C, Attia J,
Frullini 2002
Thakkinstian A. A systematic review and meta-analysis
Frullini A, Cavezzi A. Sclerosing foam in the treatment of
of randomised controlled trials comparing endovenous
varicose veins and telangiectases: history and analysis of
ablation and surgical intervention in patients with varicose
safety and complications. Dermatologic Surgery 2002;28(1):
vein. European Journal of Vascular and Endovascular Surgery
11–5.
2012;44(2):214–23.
Hayden 2001
Hayden A, Holdsworth J. Complications following re- Tessari 2001
exploration of the groin for recurrent varicose veins. Annals Tessari L, Cavezzi A, Frullini A. Preliminary experience with
of the Royal college of Surgeons of England 2001;83(4):272–3. a new sclerosing foam in the treatment of varicose veins.
Dermatologic Surgery 2001;27(1):58–60.
Higgins 2011
Higgins JPT, Green S, editors. Cochrane Handbook for Wigger 1998
Systematic Reviews of Interventions Version 5.1.0 [updated Wigger P. Surgical therapy of primary varicose veins.
March 2011]. The Cochrane Collaboration, 2011. Schweizerische Medizinische Wochenschrift. Journal Suisse de
Available from www.cochrane-handbook.org. Medecine 1998;128(45):1781–8.
Xenos 2009 versus conventional surgery for long saphenous vein varices.
Xenos ES, Bietz G, Minion DJ, Abedi NN, Sorial EE, Cochrane Database of Systematic Reviews 2006, Issue 1.
Karagiorgos N, et al. Endoluminal thermal ablation versus [DOI: 10.1002/14651858.CD005624
stripping of the saphenous vein: meta-analysis of recurrence
Nesbitt 2011
of reflux. International Journal of Angiology 2009;18(2):
Nesbitt C, Eifell RKG, Coyne P, Badri H, Bhattacharya
75–8.
V, Stansby G. Endovenous ablation (radiofrequency and
References to other published versions of this review laser) and foam sclerotherapy versus conventional surgery
for great saphenous vein varices. Cochrane Database
Eiffell 2006 of Systematic Reviews 2011, Issue 10. [DOI: 10.1002/
Eifell RKG, Bhattacharya V, Stansby GP. Endovenous 14651858.CD005624.pub2
∗
ablation (radiofrequency and laser) and foam sclerotherapy Indicates the major publication for the study
CHARACTERISTICS OF STUDIES
Characteristics of included studies [ordered by study ID]
Darwood 2008
Methods Study design: Prospective, randomised controlled trial

Country: UK
Setting/Location: Hospital
Source of funding: Promed (Bluntisham, UK) - sponsor had no input in study design,
data collection, data analysis/interpretation or preparation of the manuscript
Intention-to-treat analysis: no (one surgery patient had EVLT 1 and was followed up
in the laser cohort)
Participants No of patients randomised: Total n = 118 patients (136 legs) (EVLT1 49 legs; EVLT2
42 legs; HL/S 45 legs)
No of patients analysed: Total n = 95 patients (114 legs) (EVLT1 42 legs; EVLT2 29
legs; HL/S 32 legs)
Exclusions post-randomisation: Seven patients (11 legs) withdrew from the study as
not happy with their treatment allocation. Six patients were treated outside the study
interval and were also excluded
Losses to follow-up: Total n = 11 patients (EVLT1 5 legs; EVLT2 4 legs; HL/S 2 legs)
Age - median years (IQR): EVLT1: 42 (30.5 - 54.5); EVLT2: 52 (35 - 59); HL/S: 49
(38.5 - 57.5)
Sex - F/M : EVLT1: 22/16; EVLT2: 16/11; HL/S: 16/14
No. bilateral limbs randomised: EVLT1 9, EVLT2 6, HL/S 4
Inclusion criteria: >18 years of age; symptomatic varicose veins and primary SFJ in-
competence (confirmed on duplex ultrasonography)
Exclusion criteria: On warfarin; unsuitable for EVLT (tortuous GSV, large incompetent
anterior accessory saphenous vein)
Interventions Treatment(s): 2 EVLT techniques:

EVLT1- 12W power with 1s laser pulses and 1s intervals between pulses; laser fibre
withdrawn 2-3mm during intervals
EVLT2- 14W continuous power and continuous laser withdrawal
Both EVLT procedures performed with EVLT®; Diomed, Andover, Massachusetts, USA
Control: HL/S - open surgery; SFJ ligation, GSV stripping to knee level + multiple
phlebectomies of varicosities
Duration: Follow-up at 1, 6, 12 weeks and 1 year after treatment
Outcomes Primary outcomes: Abolition of reflux in the treated segment of GSV and improvement
in disease-specific quality of life three months after treatment
Secondary outcomes: Post-procedure pain, time to return to normal activity + work,
cosmesis, overall satisfaction at three months
Recurrence definition: Authors state “This short-term study was not designed to assess
recurrence rates”
Notes Patients with bilateral veins were randomised once and received the same treatment
simultaneously on each leg
Authors reported difficulty recruiting patients to the study. They did not meet the sample
Darwood 2008 (Continued)
sizes for their study groups to make their desired power calculations Statistical tests for
equivalence were therefore not performed
Risk of bias
Bias Authors’ judgement Support for judgement
Random sequence generation (selection Low risk Block randomisation using sealed en-
bias) velopes. Randomisation was stratified by
consultant ’to allow for any minor varia-
tions in technique’. No clear details on how
this stratification was achieved
Allocation concealment (selection bias) Low risk Used sealed envelopes
Blinding (performance bias and detection High risk Not possible to blind investigators or pa-
bias) tients. Assessors at follow-up were also not
All outcomes blinded
Incomplete outcome data (attrition bias) Low risk Missing data are balanced across the
All outcomes groups, with similar reasons given for the
missing data
Selective reporting (reporting bias) Low risk The pre-specified outcomes in the study
protocol were reported in the pre-specified
way
Authors declared that their sample sizes
were insufficient to permit statistical test-
ing for equivalence
Analysis was not on an intention-to-treat
basis
Other bias Unclear risk Authors reported difficulty recruiting pa-

tients to the study. They declared that their
sample sizes were insufficient to permit sta-
tistical testing for equivalence
One patient randomised to surgery under-
went laser, and was followed up in the laser
cohort showing no analysis of intention to
treat
Some patients received bilateral treatment.
Authors stated these patient would receive
the same treatment on both limbs. These
patients who received bilateral treatment
were not stratified in the results. Some out-
come measures can be affected by bilateral
treatment e.g. QoL, pain, time to return to
work etc. therefore introducing a bias
Patients who underwent laser did not have
Darwood 2008 (Continued)
concomitant mini-phlebectomies. This

adds a potential confounding bias when
analysing results of post-operative pain,
time to return to work etc
EVOLVeS Study
Methods Study design: Multicentre, prospective, randomised controlled trial

Country: France, Austria and USA
Setting/Location: Hospitals
Source of funding: VNUS medical technologies provided financial support for data
collection, clinical monitors and disposable catheters (RFA) free of charge
Intention-to-treat analysis: No
Participants No of patients randomised: Total n = 85 patients (86 limbs) (RFA n = 45 (46 limbs);
HL/S n = 40 (40 limbs)
No of patients analysed:
At 72 hours, total 80 legs (RFA 44 leg; HL/S 36 legs)
At 4 months, total n = 79 (77 legs) (RFA 43 legs; HL/S 34 legs)
At 2 years, total n = 65 (65 legs) (RFA 36 legs; HL/S 29 legs)
Exclusions post-randomisation: three patients refused surgery, one patient repeatedly
DNA’d, two patients excluded from RFA due to inclusion criteria violation
Losses to follow-up: Yes: 1) Clinical Ex: 2 surgery + 1 RFA no follow up at 4 months /
2) QoL questionnaires: Surgery: 1 at 72 hours, 4 at 4 months not completed. RFA: 1 at
72hr, 1 wk, 3 wk + 4 months not completed / at 1 year 19 limbs in RFA and 16 limbs
in HL/S were lost but at 2 years it improved with only 8 RFA and 7 HL/S losses
Age - mean years (SD): RFA 49 (4); HL/S 47 (4)
Sex - F/M: RFA 32/12; HL/S 26/10
No. bilateral limbs randomised: RFA 1, HL/S 0
Inclusion criteria: Reverse flow in GSV lasting > 0.5s in standing position / Age 21 - 80
/ CEAP classification C2, C3, C4 / Ambulatory status / Segmental deep reflux allowable
/ Saphenous vein diameter ≤ 1.2 cm in supine position / Availability for follow up visits
- 72 hours, 1 wk, 3 wk, 4 months
Exclusion criteria: Saphenous vein diameter > 1.2 cm or < 0.2 cm / Duplication of
saphenous trunk or incompetent accessory branch / Small saphenous vein reflux / Varices
of the thigh / Previous DVT / ABPI < 0.9 / Axial deep venous reflux from groin through
popliteal vein / Tortuosity of GSV segment to be treated on basis of appearance and USS
as unsuitable for catheterisation
Interventions Treatment(s): GSV obliteration with RFA without high ligation of SFJ- used the Closure
catheter and system (VNUS Medical Technologies)
Control: HL/S - vein stripping (from knee or upper calf to the SFJ) with high ligation
of SFJ
Duration: Follow-up was at 72 hours, 1 and 3 weeks, 4 months, 1 and 2 years
Outcomes Primary outcomes: It is not clear from the paper what the specific primary or secondary
measures were. EVOLVeS was designed to compare procedure-related complications,
patient recuperation and QoL outcomes
Secondary outcomes: Although it was not initially declared, the EVOLVeS trials later
EVOLVeS Study (Continued)
presented rates of neovascularisation in the groin and recurrence at 2 years

Recurrence definition: New varicose veins below the knee
Notes Two investigators audited the study’s raw data handling and storage methods, data pro-
cessing accuracy, and presentation of specific results. They reported all was in order and
that the raw data reflected the results accurately. This was done at four months and two
years post-data collection
Risk of bias
Random sequence generation (selection Unclear risk ’Randomisation was allocated via Internet’
bias) - no further details were given
Allocation concealment (selection bias) Low risk Allocation performed via the Internet
Blinding (performance bias and detection High risk Not possible to blind patients or operators;
bias) assessors were also not blinded, not appli-
All outcomes cable in this study
Incomplete outcome data (attrition bias) High risk Details were provided on all missing out-
All outcomes come data, however it led to an imbalance
in the study treatment group
There is also discrepancy with the miss-
ing outcome data and explanations of these
missing data compared to the published
two-year follow-up
way
Other bias Unclear risk The RFA treatment cohort included one
patient who underwent treatment of both
limbs with a three-month gap between
treatments. The patient was only ran-
domised once and each limb treated as a
separate episode
All centres were established centres in the
use of RFA and the company funded the
research. No subjective data were reported.
However as in all of these studies surgical
technique and ultrasonographic results are
operator dependent
Flessenkaemper 2013
Methods Study design: Multicentre, prospective, randomised controlled trial

Country: Germany
Setting/Location: Three hospital centres: 1) Centre for Vascular Medicine, Helios
Klinikum Emil von Behring, Berlin 2) Centre for Venous Diseases, Frieburg 3) Centre
for Venous Diseases, Saarlois
Source of funding: Sponsored by Deutschen Gessellschaft für Phlebologie (DGP)
Intention-to-treat analysis: Not indicated
Participants No of patients randomised: Total n = 449 (EVLT n = 142; EVLT+HL n = 148; HL/
S n = 159). Details of the EVLT+HL group are reported here but were not used in this
review
No of patients analysed: 100% at 2 months; 86% at 6 months Total n = 385 (EVLT n
= 127; EVLT+HL n = 133; HL/S n = 128)
Exclusions post-randomisation: Not indicated
Losses to follow-up: At 6 months EVLT n = 15; EVLT+HL n = 15; HL/S n = 39
Age - mean years (SD): EVLT 47.4 (12.9); EVLT+HL 48.7 (12.0); HL/S 47.7 (11.5)
Sex - M/F: EVLT 45/97; EVLT+HL 37/111; HL/S 47/112
Inclusion criteria: Patients between 18 and 72 years old with clinical signs or symptoms
of superficial venous insufficiency with proven reflux into GSV, with a life expectancy of
more than 5 years; all patients suitable for open and endoluminal therapy with diameter
of GSV not exceeding 16 mm at a point 5 cm distal to the SFJ
Exclusion criteria: Previous surgery of the GSV - only reported exclusion criteria
Interventions Treatment(s): EVLT- laser therapy with a 980-nm diode laser, used local tumescent and
general anaesthesia
EVLT with high ligation (EVLT+HL)- EVLT performed combined with high ligation,
under general anaesthesia
Both EVLT procedures performed with instruments from Biolitec Jena, Germany (30W)
Control: HL/S- resection of all branches down to the dorsal level of the femoral vein;
under general anaesthesia
Duration: Follow-up at two, six, 12 and 24 months for re-examination, then followed
participants as long as possible
Outcomes Primary outcomes: Inguinal venous reflux after 2 years

Secondary outcomes: Peri-operative technical success rate, rate of hyperpigmentation
and matting, neurological compilations, duration of compression therapy and lym-
phoedema, complications, postop ecchymosis, pain (visual analogue scale 1 -10) or dis-
comfort, duration of disability, patient satisfaction, clinical severity (CEAP, VCSS, Hach
classification, VDS)
Recurrence definition: Any reflux more than 0.5 s from the SFJ into the GSV, which
was assessed by physicians by duplex ultrasound at two-year follow-up
Notes May 2005 to July 2009

Reflux was defined as retrograde flow of > 0.5 s duration after Valsalva manoeuvre or
manual compression and decompression of distal vein
Nearly all patients were treated as in-patients
All three groups had simultaneous mini-phlebectomies, as required
Risk of bias
Flessenkaemper 2013 (Continued)
Random sequence generation (selection Low risk Used lottery ticket box at central office and
bias) telephone randomisation
Allocation concealment (selection bias) Low risk Used central office and telephone randomi-
sation
Blinding (performance bias and detection High risk Described as ’open’, and “Because of the
bias) scars, blinding for the follow-up was not
All outcomes possible”
Incomplete outcome data (attrition bias) Unclear risk There was no mention of dropouts, in-
All outcomes tention-to-treat analysis or technical failure
rates, but the numbers analysed were simi-
lar between the groups
Selective reporting (reporting bias) Low risk All outcomes reported on, but only to six
months; two-year data is expected in future
publications
Other bias Unclear risk Possibly underpowered; power calculation

described a need of 469 participants, but
only 449 were randomised
Number of patients needing additional
phlebectomies was not recorded, which
could effect post-operative pain
FOAM-Study
Methods Study design: Multicentre, prospective randomised controlled trial

Country: The Netherlands
Setting/Location: Three hospital outpatient dermatology and surgery departments
Source of funding: The Netherlands Organization for Health Research and Develop-
ment (ZonMw); sponsor had no input in study design, data collection, data analysis/
interpretation or preparation of the manuscript
Intention-to-treat analysis: No - “Only patients who underwent the allocated inter-
vention were included in the analysis”
Participants No of patients randomised: Total n = 460 (UGFS n = 233; HL/S n = 227)

No of patients analysed: Total n = 390 (UGFS n = 213; HL/S n = 177)
Exclusions post-randomisation: UGFS n = 3; Surgery n = 27
Losses to follow-up: UGFS n = 17; HL/S n = 23
Age - mean years (SD): UGFS 55.8 (13.4); HL/S 54.6 (13.4)
Sex (F/M): UGFS 175/58; HL/S 162/65
Inclusion criteria: Patients with primary GSV incompetence, presence of one or more
venous symptoms in combination with incompetence of the saphenofemoral junction
FOAM-Study (Continued)
and GSV; reflux time of more than 0.5 s; normal deep venous system on duplex imaging
Exclusion criteria: Patients with an incompetent deep venous system; sign of a previous
deep venous thrombosis on duplex imaging; an active ulcer; contraindication to the use
of polidocanol
Interventions Treatment(s): UGFS - sclerosing foam was prepared with the double-syringe technique,
applying a 1:4 ratio of sclerosant:air; the treatment was considered successful when the
proximal GSV was completely filled with foam and maximal venospasm was achieved
Control: HL/S - performed as day-case procedure under general or spinal anaesthesia;
saphenofemoral junction was ligated and the GSV divided and stripped to just below
the knee
Duration: Follow-up at 3 months, 1 year and 2 years
Outcomes Primary outcomes: Recurrence

Secondary outcomes: Recurrent reflux (irrespective of symptoms), reduction of symp-
toms, QoL (EQ-5D), adverse events, direct hospital costs, patient satisfaction
Recurrence definition: defined as reflux longer than 0.5 s by duplex scans, combined
with the presence of one or more venous symptoms
Notes October 2005 to December 2007

Phlebectomies: UGFS - as needed; HL/S - at discretion of the surgeon (UGFS n = 26;
HL/S n = 87)
Risk of bias
Random sequence generation (selection Low risk “… assigned randomly to UGFS or surgery
bias) using a computer-generated randomiza-
tion scheme with random permuted blocks
of eight”
Allocation concealment (selection bias) Low risk Used computer-generated randomisation
Blinding (performance bias and detection High risk No indication of blinding of patients, re-
bias) search/medical staff or outcome assessors
All outcomes
Incomplete outcome data (attrition bias) Low risk Dropouts and reasons were thoroughly re-
All outcomes ported
Selective reporting (reporting bias) Low risk All outcomes reported on
Other bias Unclear risk Mini-phlebectomies were performed at the

operating surgeon’s discretion and 26 pa-
tients in the UGFS group received phlebec-
tomies and only seven did in the surgery
treatment group; this could alter the pain
and other outcomes
FOAM-Study (Continued)
As seen in the commentary letter from MJ

Gough at the end of the British Journal
of Surgery publication, there are concerns
with the high incidence of recurrence in
the surgery treatment group and the defini-
tion of recurrence solely as reflux; caution
should be taken when interpreting data
from this study
Helmy ElKaffas 2011
Methods Study design: Prospective, randomised controlled trial

Country: Egypt
Setting/Location: Not indicated
Source of funding: Not indicated
Participants No of patients randomised: Total n = 180 (RFA n = 90; HL/S n = 90)

No of patients analysed: At 24 months Total n = 162 (RFA n = 81; HL/S n = 81)
Exclusions post-randomisation: It appears two were excluded from the RFA group, but
no explanation, none excluded from HL/S
Losses to follow-up: RFA n = 7; HL/S n = 9
Age - mean years (SD): RFA 33.1 (2.6); HL/S 34.9 (3.7)
Sex - M/F: RFA 42/48; HL/S 45/45
Inclusion criteria: Patients with saphenofemoral junction and great saphenous reflux
on duplex ultrasound, either in response to Valsalva manoeuvre or with standing manual
compression and release
Exclusion criteria: Patients with deep or superficial venous thrombosis; patiens on
anticoagulants; those with concomitant peripheral arterial disease, pacemakers or serious
systemic disease; pregnant women; patiens with GSV lumen more than 18 mm in the
thigh or extremely tortuous veins
Interventions Treatment(s): Ultrasound guided radiofrequency ablation- RFA Closuresystem, using

local (tumescent) anaesthesia, managed as day patients; ClosureSystem VNUS Medical
Technologies Inc
Control: Standard surgical treatment (HL/S)- saphenofemoral high ligation and great
saphenous stripping at ankle in 40 patients and at knee level in 50 patients, using general
anaesthesia, managed as in-patients
Duration: Followed up after 1 week, 1 month and then 6 month intervals for 24 months
Outcomes Primary outcomes: Operative time, hospital stay, costs, short-term and mid-term com-
plications, recurrence
Recurrence definition: not provided
Notes Conducted between May 2006 and January 2009

No information was given on the inclusion or exclusion of participants with bilateral
treatment - authors have been contacted, but no response received
Adjuvant stab phlebectomies were performed in n = 15 patients in RFA and n = 39 in
Helmy ElKaffas 2011 (Continued)
the surgical group; all phlebectomies took place at the primary intervention. In addition
n = 24 patients required foam sclerotherapy for persistent veins following RFA; n = 0
required foam following HL/S
Risk of bias
Random sequence generation (selection Unclear risk No description of sequence generation

bias) methods
Allocation concealment (selection bias) Low risk “Patients were asked to blindly choose an
assignment card that would place them in
either group”
All outcomes
Incomplete outcome data (attrition bias) Low risk Dropouts were reported and similar be-
All outcomes tween groups, although reasons were not
given
Selective reporting (reporting bias) Low risk All outcomes were reported on
Other bias Unclear risk Two operators performed RFA and just one
performed HL/S; this could have led to en-
hanced outcomes; no indication was given
about whether bilaterally treated patients
were included or excluded, and how many
between groups (authors were contacted)
HELP-1
Methods Study design: Single-centre, prospective, randomised controlled trial

Country: UK
Setting/Location: Tertiary referral vascular surgical department
Source of funding: Internal university funding; Diomed/Angiodynamics provided 50%
of a research nurse’s salary over 12 months; sponsor had no input in study design, data
collection, data analysis/interpretation or preparation of the manuscript
Intention-to-treat analysis: Yes
Participants No of patients randomised: Total n = 280 (EVLT n = 140; HL/S n = 140)

No of patients analysed: Total n = 237 (EVLT n = 124; HL/S n = 113)
Exclusions post-randomisation: EVLT n = 1; HL/S n = 3
Losses to follow-up: EVLT n = 15; HL/S n = 24
Age - mean years (SD): EVLT 49 (14); HL/S 49 (13)
Sex - F/M: EVLT 85/54; HL/S 90/47
HELP-1 (Continued)
Inclusion criteria: Primary, symptomatic unilateral varicose veins with isolated saphe-
nofemoral junction incompetence, leading to reflux into the GSV; incompetence was
defined as reflux of at least one s on spectral Doppler analysis; both surgeon and patient
had to occupy position of equipoise of either procedure
Exclusion criteria: Previous treatment for ipsilateral varicose veins; deep venous incom-
petence or obstruction; age less than 18 years; pregnancy; impalpable foot pulses; inabil-
ity to give informed consent
Interventions Treatment(s): EVLT (810nm, bare tipped) - Performed under local tumescent anaesthe-
sia within an outpatient department; GSV was cannulated percutaneously; cannulation
was performed laterally at the lowest point of demonstrable reflux; catheter positioned
at the SFJ, aiming for a flush occlusion; bar-tipped 600-nm laser fibre was introduced
and delivered energy using an 810-nm diode laser generator set to 14 W; Diomed/An-
giodynamics, Cambridge UK
Control: HL/S- All patients received general anaesthesia; flush SFJ ligation followed by
ligation of all tributaries to second branch; inversion stripping of the GSV to the knee
Duration: Assessed at 1 week, 6 weeks, 3 months and 1 year
Outcomes Primary outcomes: QoL (UK SF-36 V1); recurrence

Secondary outcomes: QoL (EQ-5D), AVVQ, severity of venous disease by CEAP and
VCSS, post-op pain scores (0-10 VAS scale), time to return to normal activity and work,
patient satisfaction (0-10 scale)
Recurrence definition: Clinically evident varicose veins at least 3 mm in diameter and
not present at one or six weeks
Notes September 2004 to March 2009

Concomitant phlebectomies were performed via stab incisions (EVLT n = 7; HL/S n =
10)
Risk of bias
Random sequence generation (selection Unclear risk “Patients were randomized equally into two
bias) groups by means of sealed opaque en-
velopes, receiving either surgery or EVLA.
Patients selected their own envelope in the
clinic under the supervision of a research
nurse”. Does not adequately describe se-
quence generation
Allocation concealment (selection bias) Low risk Used sealed opaque envelopes
Blinding (performance bias and detection High risk Described as an ’unblinded’ trial and no
bias) discussion of blinding of assessors
All outcomes
All outcomes ported
HELP-1 (Continued)

described a need of 120 participants in each
group, but only 113 were available for fol-
low-up in the surgery group
Magna 2007
Methods Study design: SIngle-centre*, prospective, randomised controlled trial

* A second centre was added in May 2009 due to slow inclusion rate
Country: Netherlands
Setting/Location: Departments of Dermatology and Vascular Surgery in two hospitals
Source of funding: Erasmus Medical Centre listed under Sponsors and Collaborators
Participants No of patients randomised: Total n = 240 legs (EVLT n = 80 legs; UGFS n = 80 legs;
HL/S n = 80 legs)
No of patients analysed: Total n = 223 legs (EVLT n = 78 legs; UGFS n = 77 legs; HL/
S n = 68 legs)
Losses to follow-up: Total n = 1 (EVLT n = 0; UGFS n = 1; HL/S n = 0)
Age - mean years (SD): EVLT 49 (15.03); UGFS 56 (13.30); HL/S 52 (15.59)
Sex - M/F: EVLT 24/54; UGFS 25/52; HL/S 22/46
No. bilateral limbs randomised: EVLT 16, UGFS 19, HL/S 17
Inclusion criteria: Adult patients with symptomatic primary incompetent GSV at least
above the knee with a diameter ≥ 0.5 cm; with an incompetent SFJ (incompetence
defined as reflux ≥ 0.5 s at colour duplex ultrasound
Exclusion criteria: Previous treatment of the ipsilateral GSV; deep venous incompetence
or obstruction; agenesis of the deep system; vascular malformations; use of anticoagulant;
pregnancy; heart failure; contraindication for one of the treatments; immobility; arterial
insufficiency; age under 18 years; inability to provide written informed consent
Interventions Treatment(s): EVLT (940 nm diode laser)- performed under ultrasound guided tumes-
cent anaesthetic
UGFS- prepared foam made with 1 cc aethoxysclerol 3%, 3 cc air; if considered necessary
procedure could be repeated after three months; no manufacturer information given
Control: HL/S- high ligation with short (above knee) stripping; performed under spinal
or general anaesthesia
Duration: Evaluated at three and 12 months
Outcomes Primary outcomes: Anatomic success according to DUS, neovascularisation

Secondary outcomes: C of the CEAP classification; type and frequency of complications;
QoL (CIVIQ and EuroQol-5D)
Recurrence definition: For the UGFS and EVLT groups - flow or reflux of the GSV
at midthigh; for surgery - presence of the GSV in the saphenous compartment at thigh
level (both groups evaluated by clinical examination and duplex ultrasound)
Magna 2007 (Continued)
Notes January 2007 to May 2010

Intention for additional phlebectomies was to perform during initial treatment, but in
several cases were performed after three months (during initial treatment: EVLT n = 15;
UGFS n = 0; HL/S n = 18 - after three months: EVLT n = 12; UGFS n = 15; HL/S n =
11)
Risk of bias
Random sequence generation (selection Low risk “...randomized using a computerized list by
bias) an independent research nurse.”
Allocation concealment (selection bias) Low risk “...randomized using a computerized list by
an independent research nurse.”
All outcomes
All outcomes ported
Selective reporting (reporting bias) Unclear risk For complications, authors stated they
would report on migraine, skin burns, skin
necrosis, and anaphylactic shock. No data
were presented for these outcomes

described a need of 240 participants, but
only 223 were analysed
Also unclear how QoL was evaluated for
bilaterally treated patients - authors do not
clarify
Pronk 2010
Methods Study design: Single-centre, prospective, non-blinded, randomised controlled trial

Country: Netherlands
Setting/Location: Outpatient clinic specialising in venous disease
Source of funding: In article as ’None’
Intention-to-treat analysis: Unclear - but most likely as analyses 68 in the HL/S group,
although two were lost to follow-up
Participants No of patients randomised: Total n = 122; legs = 130 (EVLT legs = 62; HL/S legs = 68)
No of patients analysed: Total n = 122; legs = 130 (EVLT legs = 62; HL/S legs = 68)
Losses to follow-up: Total = 2 (EVLT n = 0; HL/S n = 2)
Pronk 2010 (Continued)
Age - mean years (SD): EVLT 49 (11.0); HL/S 50 (10.5)

Sex - M/F: EVLT 16/46; HL/S 15/53
No. bilateral limbs randomised: EVLT 8, HL/S 0
Inclusion criteria: EVLT 16/46; HL/S 15/53
Exclusion criteria: Previous surgical treatment of the GSV; intrafascial GSV reflux length
≤ 15 cm measure from SFJ downwards; GSV diameter ≤ 0.3 or ≥ 1.5 cm; pregnancy;
immobility; intolerance of lidocaine; active superficial phlebitis; previous or active deep
venous thrombosis; deep venous insufficiency
Interventions Treatment(s): EVLT (980 nm diode laser; Biolitec)- DUS-guided; perivenous tumescent
anaesthesia under ultrasonographic guidance
Control: HL/S - Sapheno-femoral ligation and stripping of the great saphenous vein;
perivenous tumescent anaesthesia; ligation of GSV followed by ligation of all tributaries
then stripping by Pin stripper through small incision just below or above the knee
Duration: Followed up at one and six weeks, six and 12 months
Outcomes Primary outcomes: Recurrent varicose veins in follow-up of 10 years (current publication
only focuses on 1-year results)
Secondary outcomes: QoL (EQ-5D), post-operative pain (visual analogue scale from 0
to 10) and complications
Recurrence definition: Visible, palpable varicosities in the area of the treated GSV,
classified as CEAP greater than or equal to C2; after surgery a new refluxing vein less
than 3 mm and clinically visible was also considered recurrent; after EVLT a recurrent
varicose vein on DUS was defined as the ability to compress the GSV, or as reflux > 0.5
s in a vein originating in the groin and connected with the femoral vein
Notes June 2007 to December 2008

“Patients with bilateral GSV incompetence were randomised only once”
“Directly after SFL/S and EVLA treatment, sclerotherapy (Aethoxysclerol 0.5-3.0%,
Kreussler) of residual superficial varicose veins was performed by a phlebologist”
Risk of bias
Random sequence generation (selection Low risk Used computer randomisation, per patient
bias)
Allocation concealment (selection bias) Low risk Used computer randomisation
Blinding (performance bias and detection High risk Described as ’non-blinding’, with no indi-
bias) cation of blinding of assessors
All outcomes
Incomplete outcome data (attrition bias) Unclear risk Although it was stated that two participants
All outcomes were lost to follow-up at six weeks; there
is no explanation of the numbers used to
analyse the one-year outcomes or the pa-
tient satisfaction outcome
Pronk 2010 (Continued)
Other bias Unclear risk Unclear: Possibly underpowered; power

calculation described a need of 120 partic-
ipants in each group, but only 113 were
available for follow-up in the surgery group.
Also, authors say patients with dual incom-
petencies were only randomised once, but
the number randomised (130) is legs, and
not patients, which is n = 122. This is
confusing and possibly misleading; all pro-
cedures performed under local tumescent
anaesthesia
Rasmussen 2007
Methods Study design: Randomised controlled trial

Country: Denmark
Setting/Location: Office-based setting, private clinic
Source of funding: grant from the Public Health Insurance Research Foundation of
Denmark; EVLT catheters provided, in part, by Biolitec AG (Bonn, Germany) and
Micronmed (Kristianstad, Sweden)
Participants No of patients randomised: Total n = 121 patients (137 legs) (EVLT n = 62 (69 legs);
HL/S n = 59 (68 legs)
No of patients analysed: At 6 months: total n = 88 (EVLT n = 47; HL/S n = 41) (all
meta-analyses performed on an intention to treat basis by all legs randomised)
Exclusions post-randomisation: none
Losses to follow-up: 12 days - EVLT 2, HL/S 0; 1 month - EVLT 4, HL/S 2; 3 months
- EVLT 6, HL/S 5; 6 months - EVLT 15, HL/S 18
Age - mean years (range): EVLT 53 (26 - 79); HL/S 54 (22 - 78)
Sex - (M/F): EVLT 21/41; HL/S 16/43
No. bilateral limbs randomised: EVLT 7, HL/S 9
Inclusion criteria: CEAP C2 - 4, Ep, As, Pr; informed consent; age 18 - 80; GSV
incompetence confirmed by > 0.5 s reflux on duplex imaging
Exclusion criteria: Duplication of GSV or incompetent anterior accessory GSV; SSV
reflux (or < 3 months since surgery for SSV incompetence); previous DVT; ABPI < 0.9
or Hx arterial disease; femoral or popliteal insufficiency; tortuous GSV
Interventions Treatment(s): EVLT (duplex guided) 980 nm diode laser, 1.5 sec pulses, 1.5 sec pause,
12 W energy; EVLT Ceralas D 980 Biolitec, Bonn Germany
Control: High tie strip and multiple stab avulsion (HL/S)
Duration: Follow-up 12 days, 1, 3, and 6 months, 2 years post procedure; 5 year data
are now available
Outcomes Primary outcomes: It is not clear what their specific primary or secondary measures
were. Rasmussen et al set out to assess safety, efficacy, post-operative morbidity, sick leave,
Rasmussen 2007 (Continued)
QoL and costs. They reported results on:

Absence from work and normal activity; AVVSS; SF-36 score; VVSS; pain VAS; com-
plications (minor e.g. required no treatment versus major e.g. required treatment, hos-
pitalisation, permanent sequelae or death); cost (procedure and days off sick from work)
Recurrence definition: Veins which had not been observed before or not previously
marked by the patient on the Aberdeen varicose vein severity score (AVVSS) form
Notes Eight patients in each group had previous high ligation i.e. were recurrent. They were
permitted as they had a patent refluxing SFJ and GSV
Author contacted and further information on randomisation process given: “A block of
10 envelopes would ensure that a sufficient number of each treatments were available, i.
e. 5 of each all the time. This is like tossing a coin but easier to document. The envelopes
were kept in a basket, but the basket was filled by a research nurse when the patients
were not present. All envelopes were alike. There was no chance of bias.”
Risk of bias
Random sequence generation (selection Low risk After contacting the author, further details
bias) on the random sequence generation were
confirmed:
’Blocks of 10 envelopes kept in a basket, the
basket was filled by a research nurse when
the patients were not present. All envelopes
were alike.’
Allocation concealment (selection bias) Low risk After contacting the author, further details
on allocation concealment were confirmed:
’The envelopes were kept in a basket, but
the basket was filled by a research nurse
when the patient were not present. All en-
velopes were alike.’
Blinding (performance bias and detection High risk Impossible to blind operator or patients to
bias) treatment; No mention that assessors post-
All outcomes operatively were blinded
Incomplete outcome data (attrition bias) Low risk Losses to follow-up did not have an impact
All outcomes on the outcome measures; the two treat-
ment groups remained similar in numbers
despite losses
way
Additional outcome measures were re-
ported in a subsequent publication (2-year
results) reporting recurrence rates, which
were not a pre-specified outcome measure.

However this does not introduce any bias
or inaccuracy into the trial
Other bias Unclear risk 121 patients (137 limbs): Included 16 pa-
tients with bilateral varicose veins; no strat-
ification of these patients in the results;
all bilaterally treated patients received the
same treatment on both legs
Rasmussen 2011
Methods Study design: Two-centre, prospective, randomised controlled trial

Country: Denmark
Setting/Location: Two private surgical centres under contract to the national healthcare
system in Denmark: Danish Vein Centre, Naestved, Surgical Centre Roskilde, Denmark
Source of funding: Financed by a grant from the Public Health Insurance Research
Foundation of Denmark. Radiofrequency equipment was provided by VNUS Medical
Technologies
Participants No of patients randomised: Total n = 500 (580 legs) (EVLT n = 125 (144 legs); RFA n
= 125 (148 legs); UGFS n = 125 (145 legs); HL/S n = 125 (143 legs))
No of patients analysed:
At 3 days: Total n = 494 (573 legs) (EVLT n = 124 (143 legs); RFA n = 124 (146 legs);
UGFS n = 123 (143 legs); HL/S n = 123 (141 legs)
At 1 month: Total n = 489 (564 legs) (EVLT n = 125 (144 legs); RFA n = 121 (141 legs)
; UGFS n = 124 (144 legs); HL/S n = 119 (135 legs)
At 1 year: Total n = 417 (476 legs) (EVLT n = 107 (121 legs); RFA n = 106 (124 legs);
UGFS n = 107 (123 legs); HL/S n = 97 (108 legs)
Exclusions post-randomisation: Total n = 2 (EVLT n = 0; RFA n = 0; UGFS n = 1;
HL/S n = 1)
Losses to follow-up: At 3 days 4 losses; (1 in EVLT, 1 in RFA (2 legs), 1 in UGFS, 1 in
HLS group), at 3 months 9 losses; (4 (7 legs) from RFA, 5 (7 legs) from HLS). At 1 year
81 losses; (18 (23 legs) from EVLT, 19 (24 legs) RFA, 17 (21 legs) UGFS, 27 (34 legs)
HLS groups)
Age - mean years (range): EVLT 52 (18 - 74); RFA 51 (23 - 75); UGFS 51 (18 - 75);
HL/S 50 (19 - 72)
Sex - percent women: EVLT 72%; RFA 70%; UGFS 76%; HL/S 77%
No. bilateral limbs randomised: EVLT 19, RFA 23, UGFS 20, HL/S 18
Inclusion criteria: Age 18-75; symptomatic varicose veins; CEAP class C2-C4E pAsPr;
GSV incompetence defined by reflux time of more than 05.s on duplex imaging; in-
formed consent provided
Exclusion criteria: Duplication of the saphenous trunk or an incompetent anterior
accessory saphenous vein; small saphenous vein reflux (until 3 months after removal of
such a vein); previous DVT; history of arterial insufficiency or ABPI <0.9, or both; axial
deep venous insufficiency (femoral, popliteal or both); tortuous GSV rendering the vein
unsuitable for endovenous treatment
Interventions Treatment(s): All performed under tumescent anaesthesia ’most’ with a light sedative
1) EVLT - Duplex guidance - 980 nm diode for the first 17 patients, 1470 nm diode for
the rest, in one centre (Roskilde) pulse mode was used and continuous mode was used
in the other centre; Ceralas D 980 Biolitec, Jeno, Germany & 1470 Ceralas D
2) RFA - performed according to the manufacturers recommendations; VNUS Medical
Technologies Inc
3) USGF - patient in reverse Trendelenburg position, GSV cannulated (5-Fr) just above
the knee, foam was 3% polidocanol (2 ml and 8 ml air mix), before injection the table
was put into the Trendelenburg position, foam was injected under USS guidance -
Retreatment was permitted within one month
Control: HLS - Under tumescent anaesthesia and ’most with sedation’. Standard groin
incision, flush ligation of GSV, division of all tributaries, GSV stripped with a pin stripper
to below the knee
Duration: Follow-up at 3 days, 1 month and 1 year after treatment. 5 year follow-up is
planned
Outcomes Primary outcomes: GSV closure (closed or absent GSV with lack of flow)
Secondary outcomes: Pain, absence from work and normal activity, QoLs (SF-36,
AVVSS) and VCSS and recurrence rates, costs
Notes Randomisation took place between February 2007 to July 2009

Bilateral treatment permitted, but both legs received same treatment
Mini-phlebectomies performed in all treatment groups to remove varicose veins (mean
(range)): EVLT 14 (1 - 43); RFA 16 (10 - 80); UGFS 15 (1-43); HL/S 15 (1-48))
Risk of bias
Random sequence generation (selection Unclear risk “Consecutive patients referred for varicose
bias) vein treatment by the family physician were
randomized in the two sites in blocks of 12
sealed envelopes to one of the four treat-
ments”. Insufficient description of random
sequence generation
Allocation concealment (selection bias) Low risk Used sealed envelopes
All outcomes
All outcomes ported
Selective reporting (reporting bias) Low risk All outcomes were reported on
Other bias Unclear risk Different methods, energies and diodes

used for EVLT in the two centres, the pro-
cedure technique was not uniform; no re-
port on what happened to the technical fail-
ures in the HL/S group; used limbs and
not patients in analysis; patients with ‘re-
current varicose veins’ were also included
if the GSV was preserved to the groin on
duplex imaging: no report on number of
recurrent varicose veins in each group; all
procedures performed under local tumes-
cent anaesthesia
Rautio 2002
Methods Study design: RCT

Country: Finland
Setting/Location: University of Oulu
Source of funding: Grant from University of Oulu, Finland
Intention-to-treat analysis: One patient was excluded after randomisation but not
withdrawn from the study indicating some intention to treat process, but which group
this patient retired from is not made clear

No of patients analysed: Total n = 28 (RFA n = 15; HL/S n = 13)
Exclusions post-randomisation: three patients left as found schedule unsuitable, further
four refused to chosen treatment and one excluded due to pregnancy
Losses to follow-up: No
Age - mean years (SD): RFA 33 (6.7); HL/S 38 (6.8)
Sex - M/F: RFA 1/14; HL/S 1/12
Inclusion criteria: Confirmed reflux (USS > 2 s GSV reflux); pt suitable for day case;
symptomatic previously untreated uncomplicated GSV tributary varicosities and isolated
unilateral SFJ incompetence
Exclusion criteria: Coagulopathies; pregnancy; multiple, tortuous (> 90 degree bend)
large-diameter GSV trunks; bilateral varicose veins; concomitant SSV varicosities
Interventions Treatment(s): RFA - VNUS® Closure® system, inserted into GSV at ankle level; no
ligation of SFJ
Control: HL/S - open surgery; SFJ ligation of all tributaries and stripping of GSV to
just below knee
Duration: Follow-up for three years
Outcomes Primary outcomes: It is not clear what their specific primary or secondary measures
were, aimed to evaluate outcome in terms of pain, sick leave, health-related QoL and
cost
Secondary outcomes: assessed further outcomes at three years including recurrence, sat-
isfaction, VCSS, VSDS and the VDS, patency of GSV and presence of neovascularisa-
tion was also assessed
Rautio 2002 (Continued)
Notes Author contacted 2/2/2010

Replied 8/2/2010:
“The 36 patients had their preoperative diagnostic done in an earlier trial (Accuracy of
HHD in planning the operating for primary varicose veins. Eur J Vasc Endovasc Surg
2002). After examining these patients and ensuring their suitability to the study they
were included.
The patients were given the study information and after getting informed consent from
all of them, I put 36 named tags to identical envelopes, which were sealed. After shuffling
the envelopes I numbered them randomly. List of numbers for randomization was done
earlier according to instructions of the biostatistician of our department. I opened the
envelopes in numerical order.
Randomisation was done this way, because our strict schedule. Resource allocations
(operating theatres, angiography suites etc.) forced us to perform the operations and
procedures during a period of two weeks.
We also thought, that it was better to inform the result of randomisation to patients in
good time beforehand. Four patients withdrew because of the disappointment of having
been assigned to the stripping group. Three patients discontinued the study because of
an unsuitable schedule. One patient was excluded because of pregnancy. As a result we
missed three patients from the RFA group and five patients from the stripping group.
I do not see any chance of bias because of selection process itself. Withdrawal of eight
patients might have had some influence to the results.”
Risk of bias
Random sequence generation (selection Low risk After contacting the author, the sequence
bias) generation details were clarified:
“36 named tags to identical envelopes,
which were sealed. After shuffling the en-
velopes they were numbered randomly. List
of numbers for randomization was done
earlier according to instructions of the bio-
statistician of our department. The en-
velopes were opened in numerical order”
Allocation concealment (selection bias) Low risk As above
Blinding (performance bias and detection High risk Impossible to blind operator or patients to
bias) treatment. No mention that assessors post-
All outcomes operatively were blinded
Incomplete outcome data (attrition bias) Low risk In the later publication (three year outcome
All outcomes measures), the authors claim ’all patients
also underwent 3 year follow up’; They re-
port no long-term losses to follow-up
Rautio 2002 (Continued)
way
Additional outcome measures were re-
ported in a subsequent publication (three
year results), recurrence rates, an additional
outcome which was not a pre-specified out-
come measure. However this does not in-
troduce any bias or inaccuracy into the trial
Other bias Low risk States competition of interest: nil

Small study groups
RELACS Study
Methods Study design: Two-centre, prospective, randomised controlled trial

Country: Germany
Setting/Location: A university dermatology department (EVLT treated group) and a
specialized vein clinic (HL/S treated group); Homburg and Bad Bertrich vascular centres
Source of funding: Not indicated
Participants No of patients randomised: Total n = 400 (EVLT n = 200; HL/S n = 200)

No of patients analysed: Total n = 316 (EVLT n = 173; HL/S n = 143)
Exclusions post-randomisation: Total n = 54 (EVLT n = 15; HL/S n = 39) - All declined
to participate
Losses to follow-up: Total n = 30 (EVLT n = 12; HL/S n = 18)
Age - mean years (SD): EVLT 47.9 (10.9); HL/S 48.0 (10.7)
Sex - percent female: EVLT 67%; HL/S 70%
Inclusion criteria: GSV insufficiency with saphenofemoral incompetence and reflux at
least down to the knee level; CVI and/or symptoms caused by GSV incompetence and.
or severe clinical finding at risk of varicose vein bleeding, thrombophlebitis or DVT;
age 18 to 65 years; performance status (according to criteria of the American Society of
Anesthesiologists, of class I - II)
Exclusion criteria: Previous surgical interventions in the groin area with the exception
of inguinal herniotomy; anterior or posterior accessory saphenous vein incompetence;
small saphenous vein insufficient requiring treatment at the same limb; acute DVT or
post-thrombotic syndrome; known thrombophilia associated with a high risk of throm-
boembolism; arterial occlusive disease classified as at least Fontaine stage IIA, and/or
ankle-brachial index below 0.8; active malignant disease (diagnosed during the past five
years); poor compliance or inability to understand the study-related procedures; women
who are pregnant or nursing
Interventions Treatment(s): EVLT (810 nm bare fibres) - laser power delivered in a continuous pull-
back fashion, performed with tumescent local anaesthetic and sedation at surgeons dis-
cretion; model 435 MedArt A/S Hvidovre, Denmark
Control: HLS - transection of all tributaries, flush ligation of SFJ with non absorbable
Ethibond 0-0 suture and neoreflux protection with an invaginating continuous Prolene
RELACS Study (Continued)
4-0 stump suture followed by invagination of GSV to just below the knee. Performed
under tumescent local anaesthetic and sedation at surgeons discretion
Duration: Follow-up at one week (days one to seven), three months, one year and two
year
Outcomes Primary outcomes: Two-year clinical recurrence-free rate according to the classification
of recurrent varices after surgery (REVAS)
Secondary outcomes: Two-year duplex recurrence-free rate at the SFJ, treatment related
adverse effects, HVVSS, QoL (CIVIQ-2), patient satisfaction, cosmetic outcome and
recovery using questionnaires and VAS (range 1 - 5)
Recurrence definition: REVAS criteria, which defined recurrence as the presence of any
new visible or palpable varicosity on the study leg noticed by the examining clinician,
originating form the operated site linked to a saphenofemoral recurrence, to an incom-
petent GSV or perforator at medial thigh level with medical indication for re-operation
Notes Randomisation took place between September 2004 to March 2007

One limb per patient was randomised (for patients with both limbs being eligible, the
one more affected by CVI was chosen for study participation)
Incompetent perforators were ligated and peripheral side branches were removed with
multiple stab avulsions; After three months, those with apparent residual varices and
perforators could be treated with additional phlebectomies or sclerotherapy (exclusively
at this time point)
Risk of bias
Random sequence generation (selection Unclear risk Insufficient description of random se-
bias) quence generation - only described as
’blocks of 10’
Allocation concealment (selection bias) Low risk “Independent randomization was con-
ducted via fax from a remote site”
All outcomes
All outcomes ported
Other bias Unclear risk EVLT and HL/S performed at 2 sepa-

rate clinics; Possibly underpowered: needed
180 patients per group but after dropouts
and losses to follow-up, the EVLT group
had only 173 and HL/S had 143; no agreed
protocol on number of additional phlebec-
RELACS Study (Continued)
tomies - this could affect pain, cosmesis,

QoL etc; all procedures performed under
local tumescent anaesthesia
Subramonia 2010
Methods Study design: Randomised controlled trial

Country: UK
Setting/Location: Hospital
Source of funding: VNUS Medical Technologies funded some of the Closure PLUS
radiofrequency ablation catheters used in the trial. They were not involved in the running
of the trial, data collection, interpretation or analyses
Intention-to-treat analysis: no cross-over occurred

No of patients analysed: Total n = 88 (RFA n = 47; HL/S n = 41)
Exclusions post-randomisation: 2 RFA patients (1 taken off waiting list, 1 did not
receive any treatment) 4 surgery (1 taken off waiting list, 1 developed atrial fibrillation,
1 developed hypertension, 1 operated on a non-trial list)
Losses to follow-up: none at 6 weeks. 53 patients (61 limbs) available at 20 months
Age - median years (IQR): RFA 47 (38 - 58); HL/S 45 (37 - 53)
Sex - M/F: RFA 13/34; HL/S 14/27
No. bilateral limbs randomised: no bilateral limbs were included
Inclusion criteria: Age 18 - 70 years; primary or recurrent GSV reflux on duplex imaging;
duplex confirmed suitable for RFA; pt fit for GA; physical condition allowing ambulation
after surgery; pt give informed consent; pt and surgeon agree intervention is required;
availability for follow-up
Exclusion criteria: Varicose veins without GSV incompetence on duplex; associated
small saphenous or deep venous incompetence; tortuous GSV unsuitable for RFA; GSV
diameter < 3 mm and > 12 mm in supine position; GSV thrombus; patients with per-
manent pacemaker or internal defibrillator; concomitant PVD (ABPI < 0.9); pregnancy;
unable to complete QoL questionnaire due to poor English language skills
Interventions Treatment(s): Radiofrequency ablation; The VNUS® Closure® PLUS intravascular

catheter with bipolar electrodes
Control: HL/S - open surgery
Duration: 1 and 5 week follow-up
Outcomes Primary outcomes: time taken to return to normal household activities

Secondary outcomes: intraoperative complications; duration of the procedure; post
op morbidity (pain, analgesic requirements, sensory abnormalities, wound problems,
phlebitis, skin burns, pigmentation); time to return to driving and patient satisfaction
and QoL
Recurrence definition: not evaluated
Notes
Risk of bias
Subramonia 2010 (Continued)
Random sequence generation (selection Unclear risk Age and sex were ’judged most likely to in-
bias) fluence outcome in the two groups’. Au-
thor contacted for further details:
“A web-based randomisation method was
used (with assistance from the Institute of
Health and Society, Newcastle University,
UK) with stratification to ensure appro-
priate balance between the arms with re-
spect to variables that might influence out-
come in the two groups and to minimise
the risk of confounding. The method used
two stratification variables, age and sex, that
were judged most likely to influence the
outcome in the two groups. Two levels of
each stratification variable were employed:
· Age - ≤ 50 years and > 50 years
· Sex - male or female
Simple randomisation without stratifica-
tion does not guarantee equivalence be-
tween the two groups and several levels
of stratification can make the randomisa-
tion system more complicated and also re-
sult in some small strata. The same proce-
dure was allocated to those with bilateral
varicose veins both of which were suitable
for the trial with a minimum period of 3
months between the procedures. Access to
the website was protected by password and
the file server maintained by the University
of Newcastle had high security protocols.
The researcher alone had knowledge of the
password to access the website. No prob-
lems were encountered either in accessing
the website or in randomising patients dur-
ing the trial.”
Allocation concealment (selection bias) Low risk Used web-based randomisation
Blinding (performance bias and detection High risk Unable to blind surgeon or patient to treat-
bias) ment. No mention that assessors post-op-
All outcomes eratively were blinded
Incomplete outcome data (attrition bias) Low risk Missing outcome data fully reported and
All outcomes balanced in numbers across intervention
groups; all patients were followed up at 5
weeks
Subramonia 2010 (Continued)
way
Other bias Unclear risk Article was written and designed by two
vascular surgeons who perform both pro-
cedures regularly and both authors declare
no personal conflict of interests in either
treatment
Included five patients with recurrent vari-
cose veins. No stratification of these pa-
tients in the results. This could introduce a
potential bias into results such as pain, time
to return to normal activities, QoL etc
The authors standardised their anaesthetic
and inter-operator variability thus reducing
bias
Age and sex variables were controlled in the
randomisation process thus reducing po-
tential confounding
Included 12 patients with bilateral varicose
veins (randomised on one occasion to the
same treatment, but had their limbs treated
with a minimum of 6 weeks in between
treatments, thus treating each limb as a sep-
arate case)
ABPI: ankle brachial pressure index

AVVQ: Aberdeen Varicose Vein Questionnaire (equivalent to AVVSS)
AVVSS: Aberdeen Varicose Vein Severity Score
cc: cubic centimetre
CIVIQ: Chronic Venous Insufficiency Quality of Life Questionnaire
cm: centimetre
CVI: chronic venous insufficiency
DNA: did not attend
DUS: duplex ultrasound
DVT: deep vein thrombosis
EQ-5D: EuroQol 5D
EVLA: endovenous laser ablation (same as EVLT)
EVLT: endovenous laser therapy
GA: general anaesthetic
GSV: great saphenous vein
HHD: hand-held doppler
HL/S: high ligation and stripping
HVVSS: Homburg Varicose Vein Severity Score
Hx: history
IQR: interquartile range
ml: millilitre
mm: millimetre
nm: nanomole
PVD: peripheral vascular disease
QoL: quality of life
REVAS: Recurrent Varices After Surgery
RFA: radiofrequency ablation
s: second
SD: standard deviation
SF-36: Medical Outcomes Short Form-36
SFJ: saphenofemoral junction
SFL/S: saphenofemoral ligation and stripping (equivalent to HL/S)
SSV: small saphenous vein
TCSS: Total Clinical Severity Score
UGFS: ultrasound guided foam sclerotherapy
USS: ultrasound scan
VAS: visual analogue scale
VCSS: Venous Clinical Severity Score
VDS: Venous Disability Score
Characteristics of excluded studies [ordered by study ID]
Study Reason for exclusion
Abela 2008 This paper compared foam sclerotherapy plus SFJ ligation to open surgery. Foam plus SFJ ligation does not
represent ’standard’ foam sclerotherapy
Almeida 2007 Based on a talk given to the 2007 VEITHsymposium, entitled “We’ve Got Plenty of Data to Show that
Endovenous Thermal Ablation is Superior to Open Surgery”. Not a report of any original work but a report
of existing RCTs
Alos 2006 Treatment of reticular or post-operative varices. Foam only
Anon 2008 This study compares two laser techniques
Belcaro 2000 Trial of sclerotherapy versus surgery, surgery = ligation of SFJ only NOT stripping of GSV + multiple stab
avulsions
Blaise 2010 Foam only, does not include comparisons to patients undergoing surgery
BLARA Trial Bilateral veins. Dual therapy. No surgery
Bountouroglou 2006 This paper compared foam sclerotherapy plus SFJ ligation to open surgery. Foam plus SFJ ligation does not
represent ’standard’ foam sclerotherapy
Bush 2008 EVLT only. No comparison made to alternative treatment
Ceulen 2007 Foam only, no comparison made to alternative therapy
(Continued)
Chant 1972 Not foam sclerotherapy
Christenson 2010 Christenson 2010 included the treatment of 200 limbs, randomised to receive open surgery or endovenous
laser ablation. After contacting the author it was confirmed that 40 patients underwent bilateral varicose vein
treatment. It was also confirmed that patients ’limbs’ were randomised, not patients. In fact, eight patients
underwent surgery on one limb and laser on the other. All patients with bilateral varicose veins were treated
on the same day
This clearly biases any results regarding post-operative quality of life scores. The high proportion of bilaterally
treated patients also affects pain scores. Time to return to work is also published, but limbs cannot return to
work independently of one another and subsequently these results are not suitable for our Cochrane review
Trials randomising and analysing results according to number of limbs rather than number of patients as the
unit of analysis means that the standard error of the treatment effect is much smaller than it should be. These
results have a much tighter confidence interval. Subsequently the results of recurrence from this trial cannot
be included in the meta-analysis of our review
Compagna 2010 This paper compared foam sclerotherapy plus ligation to open surgery. Foam plus ligation does not represent
’standard’ foam sclerotherapy
De Medeiros 2006 Compared laser versus open surgery. However the patients undergoing laser also had a high tie of SFJ. This
does not constitute ’standard’ laser treatment and subsequently the trial was excluded
Demagny 2002 Sclerotherapy only: foam versus liquid
Desmyttere 2005 Only evaluating EVLT, no comparison made to alternate therapies
Disselhoff 2008 Cryostripping. Not included
Disselhoff 2011 Compared EVLT with and without ligation
Doran 1975 Fegan’s method; included liquid sclerotherapy
Duffy 2005 Study only included patients without SFJ incompetence
Einarsson 1993 Surgery versus liquid sclerotherapy - not foam
Figueiredo 2009 Figueiredo 2009 randomised 60 patients to receive open surgery or foam sclerotherapy. They included
treatment of both small and great saphenous varicose veins in both the thigh and lower leg. They also treated
accessory varicose veins in the thigh and lower leg, and perforating veins of the thigh and lower leg. In their
original trial they did not include any stratification in the results of who underwent treatment of GSV varicose
veins alone
The author was contacted to enquire about any stratified results
The author provided a table which showed treatment ’success’ versus treated vein segment. Treated segments
included (as above):
great saphenous vein (thigh, great saphenous vein (lower leg), small saphenous vein, accessory vein (thigh),
accessory vein (lower leg), perforating vein (thigh), perforating vein (lower leg)
After further contact the author clarified “success” as: total occlusion and partial recanalisation without reflux
A total of 72 (surgery) and 74 (foam sclerotherapy) treatments were carried out. Yet only 60 patients were
randomised. It is not clear if patients underwent multiple treatments or bilateral treatments
(Continued)
The author presents results differentiating between the treatment of the great saphenous vein in the thigh
and the lower limb. It is unclear how or why this was so
Many aspects of this trials results remain unclear. It was the joint decision of CN, RE and GS that this trial be
excluded on the grounds that its results could not be adequately analysed within the inclusion and exclusion
criteria of this review
Gale 2009 This study compares RFA and EVLA. There is no comparison made to open surgery
Goode 2010 Trial does not compare ablation to open surgery
Hamel-Desnos 2003 No comparison made to alternative therapy
Hamel-Desnos 2005 Foam only and does not include comparisons to patients undergoing surgery
Hamel-Desnos 2008 No comparison made to alternative therapy
Hamel-Desnos 2009 Sclerotherapy only
Hayes 2008 EVLT only versus no treatment
HELP-2 EVLT versus surgery for SSV
Hinchliffe 2006 This trial specifically looked at recurrent varicose veins. Redo groin surgery involved exposing 5 cm of femoral
vein above and below the SFJ. This does not constitute ’conventional’ varicose vein surgery
Hobbs 1968 Used liquid sclerotherapy, not foam
Jia 2010 Compares foam sclerotherapy with ligation to standard stripping
Kabnick 2003 Only evaluating laser. No comparison made to other treatments
Kalodiki 2012 This paper compared foam sclerotherapy plus ligation to open surgery. Foam plus ligation does not represent
’standard’ foam sclerotherapy
Kalteis 2008 This paper compared laser plus SFJ ligation with open surgery. This is not ’standard’ laser therapy, hence the
study was excluded
Kern 2005 Not treatment of GSV. No comparison made to other treatments - foam only
Kuznetsov 2005 Treatment of varicose veins complicated by trophic ulcers
Lattimer 2012 EVLT versus foam sclerotherapy, with no comparison with surgery
Lin 2007 This paper was written in Chinese. Despite complete translation we were not able to extrapolate any mean-
ingful data that would enhance this Cochrane review
Lin 2009 Patients with chronic venous insufficiency, not described as having varicose veins
(Continued)
Liu 2011 Foam sclerotherapy with ligation compared to stripping
Lugli 2009 Sclerotherapy only
Lupton 2002 Not treatment of GSV
Martimbeau 2003 Compares different forms of foam, not two different treatments of varicose veins
Maurins 2009 Only includes patients undergoing endovenous laser ablation
McDaniel 1999 Study does not include patients with varicose veins of the great saphenous vein
Mekako 2006 This study includes varicose veins of other sites as well as the great saphenous veins
Mekako 2007 Only evaluating EVLT. No comparison to alternate treatments
NCT00841178 This study includes varicose veins of other sites as well as the great saphenous veins
Neglen 1993 Compared surgery to liquid sclerotherapy, not foam sclerotherapy
Ogawa 2008 Eight of the 92 patients had treatment of short saphenous vein reflux. There was no stratification in the
results between small and great saphenous vein reflux
Ouvry 2008 Compares two different types of foam, no comparison to different treatment techniques
Rabe 2008 Sclerotherapy treatment only (liquid versus foam)
Rao 2005 Sclerotherapy treatment only (foam versus liquid)
Rasmussen 2010 Compares EVLT, foam sclerotherapy and open surgery for SSV
REACTIV Excluded for two reasons: 1) the intervention surgery versus sclerotherapy included both saphenopopliteal
junction (SPJ) and SFJ incompetence, with no stratification in the analysis of the numbers of each group; 2)
sclerotherapy was not foam sclerotherapy
RECOVERY Trial No comparison made to open surgery. RFA versus laser only
Rutgers 1994 Liquid not foam sclerotherapy
Rybak 2003 Foam treatment only
Sadoun 2003 Foam treatment only
Seddon 1973 Quasi-randomised. Also, used liquid, not foam, sclerotherapy
Selles 2008 Comparing two different foam techniques only
(Continued)
Shepherd 2009 This trial did not include an open surgery arm for comparison
Sica 2006 Foam only, no comparison to alternative treatments
Stotter 2005 Cryostripping. Not included
Theivacumar 2008 Only compares EVLT (two different techniques). No comparison to alternative treatments
Theivacumar 2009 This paper presents the two-year results of both neovascularisation and recurrence for the Darwood 2008
RCT series. However they also include all other patients who were treated (non-randomly) at the same time of
the original RCT. There is no stratification in these results of the randomised and non-randomised patients.
These results could therefore not be utilised and the study was excluded
VEDICO Trial This study compared six treatment options: 1) standard sclerotherapy; 2) high dose sclerotherapy; 3) multiple
vein ligation; 4) stab avulsion; 5) foam-sclerotherapy; 6) surgery (ligation with sclerotherapy). Therefore the
study does not include a true open surgery group, i.e. SFJ ligation + stripping group
NB: a stripping group was included but the authors declared that this was a non-randomised reference group
only
Viarengo 2007 EVLT only, no comparison made to alternative treatments
Vuylsteke 2006 This is a controlled trial, no randomisation was used
Wright 2006 In this study patients were divided into two groups initially by physicians choice based on the extent of their
disease into a surgery or sclerotherapy group. These two groups were then randomised. This introduces a
clear bias from the outset. In the varisolve versus surgery group, ’surgery’ was not standardised. It included
stripping in only 88.3% of cases. Their inclusion criteria for surgery included GSV and SSV and this was
not stratified in the results
Yamaki 2008 Foam only: no other treatment compared
Yang 2013 Compares EVLT with ligation to surgery
Zeh 2003 Foam only, no other treatment compared
EVLA: endovenous laser ablation (equivalent to EVLT)

EVLT: endovenous laser therapy
GSV: great saphenous vein
RCT: randomised controlled trial
RFA: radiofrequency ablation
SFJ: saphenofemoral junction
SPJ: saphenopopliteal junction
SSV: small saphenous vein
Characteristics of ongoing studies [ordered by study ID]
CLASS
Trial name or title CLASS (Comparison of LAser, Surgery and foam Sclerotherapy)
Methods Randomised controlled trial comparing foam sclerotherapy, alone or in combination with endovenous laser
therapy, with open surgery as a treatment for varicose veins
Participants Aim: 1016
Interventions Open surgery for the treatment of varicose veins compared with foam sclerotherapy alone of main great
or small saphenous trunk and non-trunk varicosities and endovenous laser ablation (EVLA) of main trunk
including foam sclerotherapy of non-trunk varicosities performed under local anaesthetic
Outcomes The primary patient outcomes are disease specific (Aberdeen Varicose Vein Questionnaire) and generic quality
of life (EQ-5D, SF-36) at 6 months. The primary economic outcome is the incremental cost per quality
adjusted life years (QALY) at 6 months. The secondary outcomes include (a) costs to the health service and
patients and any subsequent care at 6 months; (b) technical success of venous intervention at 6 weeks and 6
months; (c) clinical success of venous intervention at 6 weeks and 6 months; (d) disease specific and generic
quality of life at 6 weeks; and (e) behavioural recovery at 6 weeks
Starting date 01/06/2008
Contact information Dr Julie Brittenden

c/o Vascular Department
Ward 36
Aberdeen Royal Infirmary
Foresterhill
Aberdeen
AB25 2ZN
UNITED KINGDOM
j.brittenden@abdn.ac.uk
Notes As of May 2014, data from the trial was under peer review, and due to a commitment to another meta-analysis
the authors could not share the unpublished data
Desai 2009
Trial name or title Endovascular venous laser vs Oesch pin stripper in management of primary varicosities of the great saphenous
vein: a randomized control trial
Methods Prospective, randomised controlled trial comparing endovenous laser therapy to high ligation and stripping
using the Oesch pin stripper
Participants n = 60 participants included
Interventions Endovenous laser therapy compared with high ligation and stripping using Oesch pin stripper
Outcomes Complications, duration of hospital stay, pain
Desai 2009 (Continued)
Starting date
Contact information Devang Desai email: docdevang@gmail.com
Notes Author contacted for more data as the current abstract publication is not enough to determine inclusion.
Author responded saying they would get us the data, but was not able to supply the data in time for the
current update
RAFPELS Trial
Trial name or title Prospective randomised trial comparing the new endovenous procedures versus open surgery for varicose
veins due to great saphenous vein incompetence (RAFPELS)
Methods RCT
Participants Estimated enrolment: 1000
Interventions Procedure: High ligation of the great saphenous vein + stripping

Procedure: Endovenous laser ablation
Procedure: Radiofrequency ablation
Procedure: Foam sclerotherapy
Outcomes Primary outcome measures:

Recurrence rate and complications after venous surgery. Duplex ultrasound and clinical evaluation (time
frame: 3 years)
Secondary outcome measures:

Quality of life with SF36, Aberdeen Vein Score, VCSS and Visual Analogue Scale Scoring (time frame: 3
years)
Starting date Study start date: January 2008
Contact information Anders Hellberg, MD, PhD Tel +4621173000, email anders.hellberg@ltv.se
Achilleas Karkamanis, MD Tel +4621173000, email achilleas.karkamanis@ltv.se
Notes Estimated study completion date: August 2011

Estimated primary completion date: January 2011 (final data collection date for primary outcome measure)
Website reports data was collected in January 2014, no indication of a publication at this time
Venermo 2013
Trial name or title Comparison of treatments in venous insufficiency
Methods Randomised controlled trial, using numbered containers to implement randomisation sequence
Participants Enrolled: n = 212; patients 20 - 70 years with venous insufficiency
Venermo 2013 (Continued)
Interventions 1) Laser ablation - a thin laser fibre is inserted through a tiny distal entry point. The probe is guided into
place using ultrasound and the procedure is performed under local tumescence anaesthesia; Laser energy is
delivered to seal the faulty vein
2) Foam sclerotherapy - involves an injection of foam (sodium tetradecyl sulfate mixed with air according to
the Thessari method) directly into the venous trunk under ultrasound control
3) Operative treatment - great saphenous vein will be removed after flush ligation by femoral vein and stripping
of the trunk
Outcomes Recanalisation or reflux of the treated venous trunk, symptom relief, symptoms evaluated by CEAP-classifi-
cation and degree of disability, complications, quality of life
Follow-up at one month and 12 months
Starting date October 2007
Contact information Maarit Venermo, MD, PhD, Helsinki University Central Hospital
Notes Final data had been collected, but results were not published at the time of the update of the review, and
authors did not supply data for inclusion
Wilhelmi 2009
Trial name or title Effectiveness and clinical outcome following endovenous therapy of primary varicose veins: first results of a
randomised, prospective study comparing the VNUS ClosureFast system, 980 nm and 1470 nm lasers and
open surgery
Methods Prospective, randomised controlled trial
Participants 201 participants (243 legs), with primary varicose veins
Interventions 1) VNUS ClosureFast

2) 980 nm laser
3) 1470 nm laser
4) open surgery
Outcomes Time to normal physical activity, pain score, use of analgesics, CIVIQ, duplex and complication rates
Starting date Unclear
Contact information M Wilhelmi

Div of Cardiac-, Throacic-, Transplantations- and Vascular Surgery, Hannover Medical School, Hanover,
Germany
Notes Authors contacted but could not present full data at the time of the review update
DATA AND ANALYSES
Comparison 1. Foam sclerotherapy versus surgery
No. of No. of
Outcome or subgroup title studies participants Statistical method Effect size
1 Recurrence 3 Odds Ratio (M-H, Random, 95% CI) Subtotals only

1.1 Recurrence noted by 3 766 Odds Ratio (M-H, Random, 95% CI) 1.74 [0.97, 3.12]
clinicians
patients
2 Recanalisation 1 Odds Ratio (M-H, Fixed, 95% CI) Totals not selected
2.1 Early (< 4 months) 1 Odds Ratio (M-H, Fixed, 95% CI) 0.0 [0.0, 0.0]
2.2 Late ( > 4months) 1 Odds Ratio (M-H, Fixed, 95% CI) 0.0 [0.0, 0.0]
3 Neovascularisation 1 Odds Ratio (M-H, Fixed, 95% CI) Totals not selected
4 Technical failure 2 429 Odds Ratio (M-H, Fixed, 95% CI) 0.44 [0.12, 1.57]
Comparison 2. Laser ablation versus surgery
No. of No. of
1 Recurrence 7 Odds Ratio (M-H, Random, 95% CI) Subtotals only

clinicians
patients
2 Recanalisation 4 Odds Ratio (M-H, Random, 95% CI) Subtotals only
2.1 Early (< 4 months) 3 519 Odds Ratio (M-H, Random, 95% CI) 1.05 [0.09, 12.77]
2.2 Late (> 4 months) 4 743 Odds Ratio (M-H, Random, 95% CI) 4.14 [0.76, 22.65]
3 Neovascularisation 4 760 Odds Ratio (M-H, Fixed, 95% CI) 0.05 [0.01, 0.22]
4 Technical failure 6 1255 Odds Ratio (M-H, Fixed, 95% CI) 0.29 [0.14, 0.60]
5 Long term recurrence (≥ 5 years) 1 Odds Ratio (M-H, Fixed, 95% CI) Totals not selected
6 Long term recanalisation (≥ 5 1 Odds Ratio (M-H, Fixed, 95% CI) Totals not selected
years)
7 Long term technical failure (≥ 5 1 Odds Ratio (M-H, Fixed, 95% CI) Totals not selected
years)
Comparison 3. Radiofrequency ablation versus surgery
No. of No. of
1 Recurrence 4 Odds Ratio (M-H, Fixed, 95% CI) Subtotals only

1.1 Recurrence noted by 4 487 Odds Ratio (M-H, Fixed, 95% CI) 0.82 [0.49, 1.39]
clinicians
1.2 Recurrence noted by 1 28 Odds Ratio (M-H, Fixed, 95% CI) 2.0 [0.30, 13.26]
patients
2 Recanalisation 4 Odds Ratio (M-H, Random, 95% CI) Subtotals only
2.1 Early (< 4 months) 4 469 Odds Ratio (M-H, Random, 95% CI) 0.68 [0.01, 81.18]
2.2 Late ( > 4months) 3 325 Odds Ratio (M-H, Random, 95% CI) 1.09 [0.39, 3.04]
3 Neovascularisation 2 93 Odds Ratio (M-H, Fixed, 95% CI) 0.31 [0.06, 1.65]
4 Technical failure 5 645 Odds Ratio (M-H, Random, 95% CI) 0.82 [0.07, 10.10]
Analysis 1.1. Comparison 1 Foam sclerotherapy versus surgery, Outcome 1 Recurrence.
Review: Endovenous ablation (radiofrequency and laser) and foam sclerotherapy versus open surgery for great saphenous vein varices
Comparison: 1 Foam sclerotherapy versus surgery
Outcome: 1 Recurrence
Study or subgroup Foam Surgery Odds Ratio Weight Odds Ratio

M- M-
H,Random,95% H,Random,95%
n/N n/N CI CI
1 Recurrence noted by clinicians

FOAM-Study 75/213 37/177 44.1 % 2.06 [ 1.30, 3.25 ]
Magna 2007 (1) 21/77 8/68 25.1 % 2.81 [ 1.15, 6.86 ]
Rasmussen 2011 (2) 17/123 16/108 30.8 % 0.92 [ 0.44, 1.93 ]
Subtotal (95% CI) 413 353 100.0 % 1.74 [ 0.97, 3.12 ]

Total events: 113 (Foam), 61 (Surgery)
Heterogeneity: Tau2 = 0.15; Chi2 = 4.44, df = 2 (P = 0.11); I2 =55%
Test for overall effect: Z = 1.85 (P = 0.064)
2 Recurrence noted by patients
FOAM-Study 24/213 16/177 100.0 % 1.28 [ 0.66, 2.49 ]
Subtotal (95% CI) 213 177 100.0 % 1.28 [ 0.66, 2.49 ]

Heterogeneity: not applicable
0.01 0.1 1 10 100

Favours Foam Favours Surgery
(1) Legs
(2) Legs
Analysis 1.2. Comparison 1 Foam sclerotherapy versus surgery, Outcome 2 Recanalisation.
Outcome: 2 Recanalisation
Study or subgroup Foam Surgery Odds Ratio Odds Ratio

n/N n/N M-H,Fixed,95% CI M-H,Fixed,95% CI
1 Early (< 4 months)

Rasmussen 2011 (1) 5/144 7/135 0.66 [ 0.20, 2.12 ]
2 Late ( > 4months)

Rasmussen 2011 (2) 20/123 4/108 5.05 [ 1.67, 15.28 ]
0.01 0.1 1 10 100

(1) Legs
(2) Legs
Analysis 1.3. Comparison 1 Foam sclerotherapy versus surgery, Outcome 3 Neovascularisation.
Outcome: 3 Neovascularisation
Study or subgroup Foam Surgery Odds Ratio Odds Ratio

Magna 2007 (1) 0/77 7/68 0.05 [ 0.00, 0.94 ]
0.001 0.01 0.1 1 10 100 1000

(1) Legs
Analysis 1.4. Comparison 1 Foam sclerotherapy versus surgery, Outcome 4 Technical failure.
Outcome: 4 Technical failure
Study or subgroup Foam Surgery Odds Ratio Weight Odds Ratio

Magna 2007 (1) 0/77 3/68 48.3 % 0.12 [ 0.01, 2.38 ]
Rasmussen 2011 (2) 3/143 4/141 51.7 % 0.73 [ 0.16, 3.34 ]
Total (95% CI) 220 209 100.0 % 0.44 [ 0.12, 1.57 ]

Heterogeneity: Chi2 = 1.16, df = 1 (P = 0.28); I2 =14%
Test for subgroup differences: Not applicable
0.001 0.01 0.1 1 10 100 1000

(1) Legs
(2) Legs
Analysis 2.1. Comparison 2 Laser ablation versus surgery, Outcome 1 Recurrence.
Comparison: 2 Laser ablation versus surgery
Study or subgroup Laser Surgery Odds Ratio Weight Odds Ratio

M- M-
n/N n/N CI CI

Flessenkaemper 2013 11/127 4/128 11.2 % 2.94 [ 0.91, 9.49 ]
HELP-1 5/124 23/113 13.1 % 0.16 [ 0.06, 0.45 ]
Magna 2007 (1) 9/78 8/68 13.0 % 0.98 [ 0.36, 2.69 ]
Pronk 2010 (2) 5/62 5/68 10.0 % 1.11 [ 0.30, 4.02 ]
Rasmussen 2007 (3) 18/69 25/68 16.9 % 0.61 [ 0.29, 1.26 ]
Rasmussen 2011 (4) 14/121 16/108 16.3 % 0.75 [ 0.35, 1.62 ]
RELACS Study 28/173 33/143 19.5 % 0.64 [ 0.37, 1.13 ]
Subtotal (95% CI) 754 696 100.0 % 0.72 [ 0.43, 1.22 ]

Total events: 90 (Laser), 114 (Surgery)
Flessenkaemper 2013 10/127 12/128 49.2 % 0.83 [ 0.34, 1.99 ]
Pronk 2010 (5) 3/62 3/68 14.1 % 1.10 [ 0.21, 5.67 ]
Rasmussen 2007 (6) 8/69 9/68 36.6 % 0.86 [ 0.31, 2.38 ]
Subtotal (95% CI) 258 264 100.0 % 0.87 [ 0.47, 1.62 ]

Heterogeneity: Tau2 = 0.0; Chi2 = 0.09, df = 2 (P = 0.95); I2 =0.0%
0.01 0.1 1 10 100

Favours Laser Favours Surgery
(1) Legs
(2) Legs
(3) Legs
(4) Legs
(5) Legs
(6) Legs
Analysis 2.2. Comparison 2 Laser ablation versus surgery, Outcome 2 Recanalisation.

M- M-
n/N n/N CI CI

Rasmussen 2007 (1) 2/69 0/68 30.1 % 5.07 [ 0.24, 107.67 ]
Darwood 2008 (2) 3/71 0/32 30.6 % 3.32 [ 0.17, 66.21 ]
Rasmussen 2011 (3) 1/144 7/135 39.3 % 0.13 [ 0.02, 1.05 ]
Subtotal (95% CI) 284 235 100.0 % 1.05 [ 0.09, 12.77 ]

2 Late (> 4 months)
Rasmussen 2007 (4) 3/69 0/68 18.8 % 7.21 [ 0.37, 142.29 ]
RELACS Study 24/173 0/143 20.1 % 47.03 [ 2.83, 780.68 ]
Darwood 2008 (5) 6/49 1/12 25.3 % 1.53 [ 0.17, 14.11 ]
Rasmussen 2011 (6) 7/121 4/108 35.8 % 1.60 [ 0.45, 5.61 ]
Subtotal (95% CI) 412 331 100.0 % 4.14 [ 0.76, 22.65 ]

Test for subgroup differences: Chi2 = 0.79, df = 1 (P = 0.37), I2 =0.0%
0.001 0.01 0.1 1 10 100 1000

(1) Legs
(2) Legs
(3) Legs
(4) Legs
(5) Legs
(6) Legs
Analysis 2.3. Comparison 2 Laser ablation versus surgery, Outcome 3 Neovascularisation.

Darwood 2008 (1) 0/49 1/12 7.8 % 0.08 [ 0.00, 2.03 ]
HELP-1 0/124 17/113 60.4 % 0.02 [ 0.00, 0.37 ]
Magna 2007 (2) 0/78 7/68 26.4 % 0.05 [ 0.00, 0.93 ]
RELACS Study 0/173 1/143 5.4 % 0.27 [ 0.01, 6.77 ]
Total (95% CI) 424 336 100.0 % 0.05 [ 0.01, 0.22 ]

Heterogeneity: Chi2 = 1.50, df = 3 (P = 0.68); I2 =0.0%
0.001 0.01 0.1 1 10 100 1000

(1) Legs
(2) Legs
Analysis 2.4. Comparison 2 Laser ablation versus surgery, Outcome 4 Technical failure.

Darwood 2008 (1) 1/71 4/32 17.8 % 0.10 [ 0.01, 0.93 ]
HELP-1 1/137 10/132 33.2 % 0.09 [ 0.01, 0.71 ]
Magna 2007 (2) 2/78 3/68 10.3 % 0.57 [ 0.09, 3.52 ]
Rasmussen 2007 (3) 3/69 2/68 6.3 % 1.50 [ 0.24, 9.27 ]
Rasmussen 2011 (4) 0/143 4/141 14.8 % 0.11 [ 0.01, 2.00 ]
RELACS Study 2/143 6/173 17.6 % 0.39 [ 0.08, 1.99 ]
Total (95% CI) 641 614 100.0 % 0.29 [ 0.14, 0.60 ]

0.001 0.01 0.1 1 10 100 1000

(1) Legs
(2) Legs
(3) Legs
(4) Legs
Analysis 2.5. Comparison 2 Laser ablation versus surgery, Outcome 5 Long term recurrence (≥ 5 years).
Outcome: 5 Long term recurrence (≥ 5 years)
Study or subgroup Laser Surgery Odds Ratio Odds Ratio

Rasmussen 2007 (1) 24/67 25/67 0.94 [ 0.46, 1.89 ]
0.01 0.1 1 10 100

(1) Legs
Analysis 2.6. Comparison 2 Laser ablation versus surgery, Outcome 6 Long term recanalisation (≥ 5 years).
Outcome: 6 Long term recanalisation (≥ 5 years)

Rasmussen 2007 (1) 0/67 9/67 0.05 [ 0.00, 0.80 ]
0.001 0.01 0.1 1 10 100 1000

(1) Legs
Analysis 2.7. Comparison 2 Laser ablation versus surgery, Outcome 7 Long term technical failure (≥ 5
years).
Outcome: 7 Long term technical failure (≥ 5 years)

Rasmussen 2007 (1) 2/67 3/67 0.66 [ 0.11, 4.06 ]
0.001 0.01 0.1 1 10 100 1000

(1) Legs
Analysis 3.1. Comparison 3 Radiofrequency ablation versus surgery, Outcome 1 Recurrence.
Comparison: 3 Radiofrequency ablation versus surgery
Study or subgroup RFA Surgery Odds Ratio Weight Odds Ratio


EVOLVeS Study (1) 5/36 6/29 18.7 % 0.62 [ 0.17, 2.28 ]
Helmy ElKaffas 2011 12/81 9/81 25.0 % 1.39 [ 0.55, 3.51 ]
Rasmussen 2011 (2) 9/124 16/108 51.7 % 0.45 [ 0.19, 1.06 ]
Rautio 2002 5/15 2/13 4.7 % 2.75 [ 0.43, 17.49 ]
Subtotal (95% CI) 256 231 100.0 % 0.82 [ 0.49, 1.39 ]

Total events: 31 (RFA), 33 (Surgery)
Rautio 2002 4/15 2/13 100.0 % 2.00 [ 0.30, 13.26 ]
Subtotal (95% CI) 15 13 100.0 % 2.00 [ 0.30, 13.26 ]
0.01 0.1 1 10 100

Favours RFA Favours Surgery
(Continued . . . )
(. . .
Continued)
Heterogeneity: not applicable
0.01 0.1 1 10 100

(1) Legs
(2) Legs
Analysis 3.2. Comparison 3 Radiofrequency ablation versus surgery, Outcome 2 Recanalisation.

M- M-
n/N n/N CI CI
EVOLVeS Study (1) 4/43 0/34 49.7 % 7.86 [ 0.41, 151.28 ]
Rasmussen 2011 (2) 0/141 7/135 50.3 % 0.06 [ 0.00, 1.07 ]
Rautio 2002 0/15 0/13 Not estimable
Subramonia 2010 0/47 0/41 Not estimable
Subtotal (95% CI) 246 223 100.0 % 0.68 [ 0.01, 81.18 ]

2 Late ( > 4months)
EVOLVeS Study (3) 3/36 3/29 37.2 % 0.79 [ 0.15, 4.23 ]
Rasmussen 2011 (4) 6/124 4/108 62.8 % 1.32 [ 0.36, 4.81 ]
Subtotal (95% CI) 175 150 100.0 % 1.09 [ 0.39, 3.04 ]
0.001 0.01 0.1 1 10 100 1000

(Continued . . . )
(. . .Continued)
M- M-
n/N n/N CI CI
Heterogeneity: Tau2 = 0.0; Chi2 = 0.23, df = 1 (P = 0.63); I2 =0.0%
Test for subgroup differences: Chi2 = 0.04, df = 1 (P = 0.85), I2 =0.0%
0.001 0.01 0.1 1 10 100 1000

(1) Legs
(2) Legs
(3) Legs
(4) Legs
Analysis 3.3. Comparison 3 Radiofrequency ablation versus surgery, Outcome 3 Neovascularisation.

EVOLVeS Study (1) 1/36 4/29 81.2 % 0.18 [ 0.02, 1.70 ]
Rautio 2002 1/15 1/13 18.8 % 0.86 [ 0.05, 15.22 ]
Total (95% CI) 51 42 100.0 % 0.31 [ 0.06, 1.65 ]

Heterogeneity: Chi2 = 0.71, df = 1 (P = 0.40); I2 =0.0%
0.01 0.1 1 10 100

(1) Legs
Analysis 3.4. Comparison 3 Radiofrequency ablation versus surgery, Outcome 4 Technical failure.

M- M-
n/N n/N CI CI
EVOLVeS Study (1) 2/44 0/36 23.3 % 4.29 [ 0.20, 92.36 ]
Helmy ElKaffas 2011 6/81 0/81 24.3 % 14.03 [ 0.78, 253.37 ]
Rasmussen 2011 (2) 0/146 4/141 24.1 % 0.10 [ 0.01, 1.95 ]
Subramonia 2010 1/47 7/41 28.4 % 0.11 [ 0.01, 0.90 ]
Total (95% CI) 333 312 100.0 % 0.82 [ 0.07, 10.10 ]

0.001 0.01 0.1 1 10 100 1000

(1) Legs
(2) Legs
ADDITIONAL TABLES
Table 1. Study sample sizes
Technique Study Patients Patients analysed

randomised
Overall Overall Surgery Alt Rx
Foam versus FOAM-Study 460 390 177 213

surgery
Magna 2007*ˆ 160 145 68 77
Rasmussen 2011* 250 204 97 107
Total 870 739 342 397
Table 1. Study sample sizes (Continued)
Laser versus Darwood 2008 118 95 30 65

surgery
Flessenkaemper 301 225 128 127
2013**
HELP-1 280 237 113 124
Magna 2007*ˆ 160 146 68 78
Pronk 2010ˆ 130 130 68 62
Rasmussen 2007 121 88 41 47
Rasmussen 2011* 250 204 97 107
RELACS Study 400 316 143 173
Total 1760 1441 688 783
RFA versus surgery EVOLVeS Studyˆ 86 65 29 36
Helmy ElKaffas 180 162 81 81

2011
Rasmussen 2011* 250 203 97 106
Rautio 2002 33 28 13 15
Subramonia 2010 93 88 41 47
Total 642 546 261 285

Alt Rx: alternative treatment
*Includes multiple comparisons in different treatment categories
**Included a third treatment group not evaluated in this review
ˆnumber of limbs rather than number of patients randomised or patients analysed
Table 2. Age and sex of participants
Technique Study Age (years) Sex (F:M)
Surgery Alt Rx Surgery Alt Rx
Foam versus FOAM-Study 54.6 (13.4) 55.8 (13.4) 162:65 175:58

surgery mean (SD) mean (SD)
Magna 2007 52 (15.59) 56 (13.30) 46:22 52:25

mean (SD) mean (SD)
Table 2. Age and sex of participants (Continued)
Rasmussen 2011 50 (19 - 72) 51 (18 - 75) 95:29 94:30

mean (range) mean (range)
Laser versus Darwood 2008 49 (38.5 - 57.5) EVLT1 42 (30.5 - 16:14 EVLT1 22:16
surgery median (IQR) 54.5) EVLT2 16:11
EVLT2 52 (35 - 59)
median (IQR)
Flessenkaemper 47.7 (11.5) 47.4 (12.9) 112:47 97:45

2013 mean (SD) mean (SD)
HELP-1 49 (13) 49 (14) 90:47 85:54

mean (SD) mean (SD)
Magna 2007 52 (15.59) 49 (15.03) 46:22 54:24

mean (SD) mean (SD)
Pronk 2010 50 (10.5) 49 (11.0) 53:15 46:16

mean (SD) mean (SD)
Rasmussen 2007 54 (22 - 78) 53 (26 - 79) 43:16 41:21

Rasmussen 2011 50 (19 - 72) 51 (23 - 75) 95:29 90:35

RELACS Study 48.0 (10.7) 47.9 (10.9) 124:61 113:48

mean (SD) mean (SD)
RFA versus surgery EVOLVeS Study 47 (4) 49 (4) 26:10 32:12

mean (SD) mean (SD)
Helmy ElKaffas 34.9 (3.7) 33.1 (2.6) 45:45 48:42

2011 mean (SD) mean (SD)
Rasmussen 2011 50 (19 - 72) 50 (19 - 72) 95:29 88:37

Rautio 2002 38 (6.8) 33 (6.7) 12:1 14:1

mean (SD) mean (SD)
Subramonia 2010 45 (37 - 53) 47 (38 - 58) 27:14 34:13

median (IQR) median (IQR)
IQR: interquartile range
SD: standard deviation
Table 3. Laser technique used
Study Laser Pulsed/Continuous Energy Technique
Darwood 2008 810 nm diode 1) Pulsed 12W 1 sec pulses, 1 sec intervals
2) Continuous 14W withdrawn 2 - 3 mm/sec
Flessenkaemper 2013 980 nm diode Continuous 30W not indicated
HELP-1 810 nm diode Continuous 14W not indicated
Magna 2007 940 nm diode Continuous not indicated not indicated
Pronk 2010 980 nm diode Continuous 12W not indicated
Rasmussen 2007 980 nm diode Pulsed 12W 1.5 sec pulses, 1.5 sec intervals
Rasmussen 2011 980 nm diode 1) Pulsed not indicated not indicated
2) Continuous
1470 nm diode 1) Pulsed
2) Continuous
RELACS Study 810 nm diode Continuous 20W not indicated

sec: seconds
Table 4. Outcome measures
Tech- Study Outcome measure

nique
Techni- Time to Pain QoL/ Compli- Cost Neo Recur- Satisfac- Duration
cal fail- re- Severity cations vasculari- rence tion/ of
ure turn to score sation cosmesis proce-
work/ dure
N activ-
ities/
driving
√ √ √ √ √
Foam FOAM-
versus Study
surgery
√ √ √ √
Magna
2007
Table 4. Outcome measures (Continued)
√ √ √ √ √ √ √
Ras-
mussen
2011
√ √ √ √ √ √
Laser Dar-
versus wood
surgery 2008
√ √ √ √ √
Flessenkaem-
per
2013
√ √ √ √ √ √ √ √ √
HELP-1
√ √ √ √
Magna
2007
√ √ √ √ √ √
Pronk
2010
√ √ √ √ √ √ √
Ras-
mussen
2007
√ √ √ √ √ √ √
Ras-
mussen
2011
√ √ √ √ √ √ √ √
RELACS
Study
√ √ √ √ √ √ √
RFA
EVOLVeS
versus
Study
surgery
√ √ √ √
Helmy
ElKaffas
2011
√ √ √ √ √ √ √
Ras-
mussen
2011
√ √ √ √ √ √ √ √ √
Rautio
2002
√ √ √ √ √ √
Subra-
monia
2010
N: number
QoL: quality of life
Table 5. Additional phlebectomies
Technique Study Additional phlebectomies
Surgery Alt Rx
Foam versus surgery FOAM-Study yes yes
Magna 2007 yes yes
Rasmussen 2011 yes yes
Laser versus surgery Darwood 2008 yes no
Flessenkaemper 2013 yes yes
HELP-1 yes yes
Magna 2007 yes yes
Pronk 2010 yes yes
RELACS Study yes yes
RFA versus surgery EVOLVeS Study yes yes
Helmy ElKaffas 2011 yes yes
Rautio 2002 yes yes
Subramonia 2010 yes yes

Table 6. Recurrence and neovascularisation
Technique Study - final Neovascularisation (%) Recurrence noted Recurrence noted

time point by clinician at final time point by patients at final time point
(%) (%)
Surgery Alt Rx Surgery Alt Rx Surgery Alt Rx
Foam versus FOAM-Study not reported 37/177 (21) 75/213 (35) 16/177 (9.0) 24/213 (11.3)
surgery - 2 yrs
Magna 2007 - 7/68 (10.3) 0/77 (0) 8/68 (11.8) 21/77 (27.3) Authors make a brief comment on
1 yr symptomatic recurrence, but not
enough detail is given to extrapo-
late data
Rasmussen not reported 16/108 (14.8) 17/123 (13.8) not reported

2011ˆ - 1 yr
Laser versus Darwood 1/12 (8.3)* 0/49 (0) no results not reported
surgery 2008 - 1 yr
Flessenkaem- not reported 4/128 (3.1) 11/127 (8.7) 12/128 (9.4) 10/127 (7.9)
per
2013 - 6
months
HELP-1 - 1 yr 17/113 (15.0) 0/124 (0) 23/113 (20.3) 5/124 (4.0) not reported
Magna 2007 - 7/68 (10.3) 0/78 (0) 8/68 (11.8) 9/78 (11.5) not reported
1 yr
Pronk 2010- 1 not reported 5/68 (7.4) 5/62 (8.1) 3/68 (4.4) 3/62 (4.8)
yr
Rasmussen not reported 25/68 (37) 18/69 (26) 9/68 (13) 8/69 (12)
2007 - 2 yrs

2011ˆ - 1 yr
RELACS 1/143 (0.7) 0/173 (0) 33/143 (23.1) 28/173 (16.2) Authors make a brief comment on
Study - 2 yrs symptomatic recurrence, but not
enough detail is given to extrapo-
late data
RFA versus EVOLVeS 4/29 (13.8) 1/36 (2.8) 6/29 (20.9) 5/36 (14.3) not reported
surgery Study - 2 yrs
Table 6. Recurrence and neovascularisation (Continued)
Helmy not reported 9/81 (11.1) 12/81 (14.8) not reported

ElKaffas 2011
- 2 yrs

2011ˆ - 1 yr
Rautio 2002 - 1/13 (8) 1/15 (7) 2/13 (15.4) 5/15 (33.3) 2/13 (15.4%) 4/15 (26.7)
2 yrs
* Only 12 limbs available for follow-up at 1 year
ˆReported as limbs and not patients
yr: year
Table 7. Re-intervention rates (technical failure)
Technique Study Re-intervention (%)
Surgery Alt Rx
Foam versus surgery FOAM-Study 10/177 (5.6) 40/213 (18.8)
Rasmussen 2011 No data 5/144 (3.5)
Laser versus surgery Rasmussen 2007 6/68 (8.8) 9/69 (13)
RELACS Study 2/143 (1.4) 6/173 (3.5)
RFA versus surgery Helmy ElKaffas 2011 6/81 (7.4) 0/81 (0)
Rautio 2002 2/13 (15.4) 2/15 (13.3)

Alt Rx = alternative treatment
Table 8. Quality of life and venous severity scores
Tech- Quality of life score Severity score

nique Study
V-Q/ SF- EQ- SF- VCSS TCSS VDS Hach

SymQ AVVSS CIVIQ236 RAND- 5D 6D CEAP VSDS HVVSS
36
√ √
Foam
FOAM-
ver-
Table 8. Quality of life and venous severity scores (Continued)
sus Study
surgery
√ √ √
Magna
2007
√ √ √
Ras-
mussen
2011
√ √
Laser Dar-
ver- wood
sus 2008
surgery √ √ √ √
Flessenkaem-
per
2013
√ √ √ √ √ √
HELP-
1
√ √ √
Magna
2007
√ √
Pronk
2010
√ √ √
Ras-
mussen
2007
√ √ √
Ras-
mussen
2011
√ √
RELACS
Study
√ √ √
RFA
EVOLVeS
ver-
Study
sus
surgery √ √ √
Ras-
mussen
Table 8. Quality of life and venous severity scores (Continued)
2011
√ √ √ √
Rautio
2002
√ √ √ √
Sub-
ramo-
nia
2010
AVVSS: Aberdeen Varicose Vein Symptom Severity Score
CEAP: Clinical severity, Etiology, Anatomy, Pathophysiology
CIVIQ2: Chronic Venous Insufficiency Quality of Life Questionnaire
EQ-5D: EuroQol-5D
Hach: Hach classification
HVVSS: Homburg Varicose Vein Severity Score
RAND-36: Short term RAND-36 (validated for Finland)
SF-36: Medical Outcomes Study Short Form 36
SF-6D: a variation of the Medical Outcomes Study Short Form 36
TCSS: Total Clinical Severity Score
VCSS: Venous Clinical Severity Score
VDS: Venous Disability Score
V-Q/SymQ: VEINES-QoL/Sym questionnaire
VSDS: Venous Segmental Disability Score
Table 9. Post-operative complications
Early (within three months)
Minor (not requiring intervention) (%) Major (requiring interven-

Study tion) (%)
Ad- Haematoma Saphenous Thermal in- Wound Bruising Phlebitis Wound Other
verse (wound or nerve injury jury/inflam- problems and pigmen- Problems
event thigh) mation (groin/stab) tation
Surg Alt Surg Alt Surg Alt Surg Alt Surg Alt Surg Alt Surg Alt Surg Alt
Tech- Rx Rx Rx Rx Rx Rx Rx Rx
nique
Foam versus surgery
3/ 0/ 6/ 0/ 0/ 17/ 4/ 0/ 0/ 2/
FOAM-
176 217 176 217 176 217 176 217 176 217
(1.7) (0) (3.4) (0) (0) (7.8) (2.3) (0) (0) (0.9)
Study
Table 9. Post-operative complications (Continued)
4/68 1/77 0/68 1/77 3/68 0/77 0/68 0/77

Magna
(5.9) (1.3) (0) (1.3) (4.4) (0) (0) (0)
2007
Ras- 1/ 1/ 5/ 2/ 6/ 8/ 5/ 17/ 1/ 4/ 1/ 1/
mussen 119 124 119 124 119 124 119 124 119 124 119 124
(0.8) (0.8) (4.2) (1.6) (5.0) (6.5) (4.2) (13. (0.8) (3.2) (0.8) (0.8)
2011**** 7)
4/ 1/ 15/ 3/ 6/ 9/ 5/ 34/ 8/ 4/ 1/ 3/
Over- 295 341 363 418 187 201 295 341 363 418 363 418
all (1.4) (0.3) (4.1) (0.7) (3.2) (4.5) (1.7) (10. (2.2) (1.0) (0.3) (0.7)
(foam 0)
ver-
sus
surgery)
Laser versus surgery
Dar- 0/32 0/80 4/32 1/80 0/32 0/80 1/32 1/80 2/32 1/80 0/32 9/80 2/32 0/80 1/32 0/80
wood (0) (0) (13) (1) (0) (0) (3) (1) (6) (1) (0) (11) (6) (0) (3)** (0)
2008*****
2/ 3/ 108/ 68/ 1/ 1/
Flessenkaem-
159 142 159 142 159 142
per
(1.3) (2.1) (68. (47. (0.6) (0.7)
2013
0) 9)
11/ 1/ 3/ 5/ 6/ 4/ 8/ 2/
HELP-
133 137 133 137 133 137 133 137
(8.3) (0.7) (2.2) (3.6) (4.5) (2.9) (6.0) (1.5)
1
4/68 2/78 0/68 2/78 3/68 0/78 0/68 0/78

Magna
(5.9) (2.6) (0) (2.6) (4.4) (0) (0) (0)
2007
Ras- 5/86 3/69 1/68 1/69 0/68 0/69 1/68 0/69 15/ 7/69 2/68 2/69 1/68 0/69 0/68 0/69
mussen (8) (5)* (2) (2) (0) (0) (2) (0) 68 (11) (3) (3) (2) (0) (0) (0)
(25) ****
2007*****
Ras- 1/ 1/ 5/ 3/ 6/ 3/ 5/ 4/ 1/ 0/ 1/ 0/
mussen 119 125 119 125 119 125 119 125 119 125 119 125
(0.8) (0.8) (4.2) (2.4) (5.0) (2.4) (4.2) (3.2) (0.8) (0) (0.8) (0)
2011*****
145/ 169/ 4/ 20/ 0/ 1/ 1/ 3/

RELACS
161 185 161 185 161 185
Study
(90. (91. (2.5) (10. (0) (0.5) 161 185

1) 4) 8) (0.6) (1.6)
17/ 5/ 10/ 10/ 0/ 0/ 2/ 1/ 277/ 255/ 17/ 39/ 15/ 3/ 4/ 4/

Over- 102 411 414 414 100 149 100 149 708 816 513 596 581 674 607 679
all (4.6) (1.2) (2.4) (2.4) (0) (0) (0.2) (0.7) (39. (31. (3.3) (6.5) (2.6) (0.4) (0.7) (0.6)
(laser 1) 3)
ver-
sus
surgery)
RFA versus surgery
30/ 1/81 3/81 9/81 0/81 0/81 0/81 6/81 3/81 0/81 1/81 0/81
Helmy
81 (1.2) (3.7) (11. (0) (0) (0) (7.4) (3.7) (0) (1.2) (0)
ElKaf- (37. 1)
fas 0)
2011
20/ 9/44 5/36 10/ 3/36 6/44 16/ 8/44 23/ 15/ 3/36 6/44 2/36 0/44 0/36 0/44
EVOLVeS
36 (20) (14) 44 (8) (14) 36 (18) 36 44 (8) (14) (6) (0) (0) (0)
Study (56) (23) (44) (64) (34) ***
Ras- 1/ 0/ 5/ 6/ 6/ 8/ 5/ 12/ 1/ 1/ 1/ 0/
mussen 119 121 119 121 119 121 119 121 119 121 119 121
(0.8) (0) (4.2) (5.0) (5.0) (6.6) (4.2) (9.9) (0.8) (0.8) (0.8) (0)
2011*****
4/13 1/15 3/13 2/15 0/13 2/15 0/13 0/15 0/13 0/15 0/13 3/15 0/13 0/15 0/13 0/15
Rautio
(31) (7) (23) (13) (0) (13) (0) (0) (0) (0) (0) (20) (0) (0) (0) (0)
2002
Sub- 1/41 0/47 20/ 9/47 0/41 0/47 6/41 0/47 0/41 5/47 0/41 0/47 0/41 0/47 0/41 0/47
ra- (2) (0) 41 (19) (0) (0) (15) (0) (0) (11) (0) (0) (0) (0) (0) (0)
mo- (49)
nia
2010
56/ 11/ 36/ 36/ 3/ 8/ 22/ 8/ 29/ 28/ 8/ 27/ 6/ 1/ 2/ 0/

Over- 290 308 290 308 171 187 90 106 209 227 290 308 290 308 290 308
all (19. (3.6) (12. (11. (1.8) (4.3) (24. (7.5) (13. (12. (2.8) (8.8) (2.1) (0.3) (0.7) (0)
(RFA 3) 4) 7) 4) 9) 3)
ver-
sus
surgery)
Late (beyond three months)
Ad- Haematoma Saphenous Thermal in- Wound Bruising Phlebitis Wound Other
verse (wound or nerve injury jury/ problems and pigmen- Problems
event thigh) inflamma- (groin/stab) tation
tion
Surg Alt Surg Alt Surg Alt Surg Alt Surg Alt Surg Alt Surg Alt Surg Alt
Tech- Rx Rx Rx Rx Rx Rx Rx Rx
nique
Foam versus surgery
2/ 12/
FOAM-
177 213
(1.1) (5.6)
Study
1/68 1/77 0/68 1/77

Magna
(1.5) (1.3) (0) (1.3)
2007
1/68 1/77 2/ 13/

Over- (1.5) (1.3) 245 290
all (0.8) (4.5)
(foam
ver-
sus
surgery)
Laser versus surgery
Dar- 0/34 0/80 1/34 0/80 0/34 0/80 0/34 0/80 0/34 0/80 0/34 0/80 0/34 0/80 0/34 0/80
wood (0) (0) (3) (0) (0) (0) (0) (0) (0) (0) (0) (0) (0) (0) (0) (0)
2008*****
1/68 0/78 0/68 1/78

Magna
(1.5) (0) (0) (1.3)
2007
Ras- 0/68 0/96 1/68 0/96 0/68 0/96 0/68 0/96 0/68 0/96 0/68 0/96 0/68 0/96 0/68 0/96
mussen (0) (0) (2) (0) (0) (0) (0) (0) (0) (0) (0) (0) (0) (0) (0) (0)
2007
*****
0/ 0/ 3/ 0/ 0/ 0/ 0/ 0/ 0/ 1/ 0/ 0/ 0/ 0/ 0/ 0/
Over- 102 176 170 254 102 176 102 176 170 254 102 176 102 176 102 176
all (0) (0) (1.8) (0) (0) (0) (0) (0) (0) (0.4) (0) (0) (0) (0) (0) (0)
(laser
ver-
sus
surgery)
RFA versus surgery
3/34 0/43 0/34 0/43 0/34 0/43 0/34 0/43 1/34 0/43 2/34 0/43 0/34 0/43 0/34 0/43
EVOLVeS
(9) (0) (0) (0) (0) (0) (0) (0) (3) (0) (6) (0) (0) (0) (0) (0)
Study
0/13 0/15 5/13 1/15 0/13 0/15 0/13 0/15 0/13 0/15 0/13 0/15 0/13 0/15 0/13 0/15
Rautio
(0) (0) (38) (3) (0) (0) (0) (0) (0) (0) (0) (0) (0) (0) (0) (0)
2002
Sub- No follow-up beyond 5 weeks

ra-
mo-
nia
2010
3/47 0/58 5/47 1/58 0/47 0/58 0/47 0/58 1/47 0/58 2/47 0/58 0/47 0/58 0/47 0/58
Over- (6.4) (0) (10. (1.7) (0) (0) (0) (0) (2.1) (0) (4.3) (0) (0) (0) (0) (0)
all 6)
(RFA
ver-
sus
surgery)
* in one patient the saphenous thrombus extended into the femoral vein, it resolved without intervention
** post-operative acute respiratory distress syndrome (requiring seven days intensive therapy unit (ITU) support) following aspiration
post-operatively
*** includes one patient that required debridement and IV antibiotics for a ’thigh and calf infection’
**** groin infection requiring antibiotics
***** results only available per limb, not per patient
Surg: surgery
Table 10. Length of procedure or operative time
Technique Study Time (min)
Surgery Alt Rx
Foam versus surgery Rasmussen 2011- mean (range)* 32 (15 - 80) 19 (5 - 145)
Laser versus surgery HELP-1- mean (SD) 61 (14) 67 (16)
Rasmussen 2011- mean (range)* 32 (15 - 80) 26 (12 - 80)
RFA versus surgery EVOLVeS Study- mean (SD) 89 (12) 74 (10)
Table 10. Length of procedure or operative time (Continued)
Helmy ElKaffas 2011- mean (SD) 45 (13) 40 (12)
Rasmussen 2011- mean (range)* 32 (15 - 80) 27 (12 - 80)
Rautio 2002- mean (SD) Operating time: 57 (11) Operating time: 75 (16.6)
Operating room time: 99 (12.9) Operating room time: 115 (18.3)
Recovery room time: 198 (40.7) Recovery room time: 227 (57.6)
Subramonia 2010- median (range) Theatre time: 55 (48 - 63) Theatre time: 82 (73 - 91)
Procedure time: 48 (39 - 54) Procedure time: 76 (67 - 84)
*Surgeon’s time
min: minutes
Table 11. Duration of hospital stay
Technique Study Length of hospital stay

% day case
Surgery Alt Rx
Foam versus surgery FOAM-Study 100 not indicated
Laser versus surgery Darwood 2008 100 100
Flessenkaemper 2013 ~100 ~100
HELP-1 78.8 100
Pronk 2010 100 100
Rasmussen 2007 100 100
RFA versus surgery EVOLVeS Study 86* 95**
Helmy ElKaffas 2011 0 100

* 5 patients kept overnight
** 2 patients kept overnight
Table 12. Procedural costs
Technique Study Cost (£) Apparent saving of Alt Rx
Surgery Alt Rx
Foam versus surgery FOAM-Study* 1824 774 1050
Rasmussen 2011** 2199 1554 645
Laser versus surgery Rasmussen 2007*** 2,066.14 2,275.06 -208.92
Rasmussen 2011** 2199 2200 -1
RFA versus surgery Helmy ElKaffas 2011*** 253.74 417.71 -163.97
Rasmussen 2011** 2199 1996 203
Rautio 2002*** 1,339.73 974.54 365.19
Subramonia 2010* **** 559.13 1275.90 -716.77

*Hospital costs
**Inclusive of indirect costs
***Converted from other currency
****The catheter cost in the RFA group increase the cost significantly, as did the increased time in the operating theatre
Table 13. Time to return to work and normal activities
Technique Study Time to return to work (days) Time to return normal activities
Surgery Alt Rx Surgery Alt Rx
Foam versus Rasmussen 2011 - 4.3 (0 - 42) 2.9 (0 -33) 4 (0 - 30) 1 (0 - 30)
surgery median (range)
Laser versus Darwood 2008* - 17 (7.25 - 33.25) laser 1: 4 (2.5 - 7) 7 (2 - 26) laser 1: 2 (0 - 7)
surgery median (IQR) laser 2: 4 (1 - 12) laser 2: 2 (0 - 7)
HELP-1 - median 14 (13 - 28) 4 (2 - 14) 14 (7 - 25) 3 (1 - 10)

(range)
Pronk 2010 - mean 4.15 (3.72) 4.38 (5.43) 3.20 (4.01) 3.16 (4.34)
(SD)
Rasmussen 2007 - 7.6 (4.9) 7 (6) 7.7 (6.1) 6.9 (7)

mean (SD)
Table 13. Time to return to work and normal activities (Continued)
Rasmussen 2011 - 4.3 (0 - 42) 3.6 (0 - 46) 4 (0 - 30) 2 (0 - 25)

median (range)
RELACS Study - 11.8 10.4

mean
RFA versus surgery EVOLVeS Study*** 12.4 4.7 3.89 1.15

- mean
Helmy ElKaffas 7 (2.6) 3 (3)

2011 - mean (SD)
Rasmussen 2011 - 4.3 (0 - 42) 2.9 (0 - 14) 4 (0 - 30) 1 (0 - 30)

median (range)
Rautio 2002** - actual: 15.6 (6) actual: 6.5 (3.3) no data no data
mean (SD) perceived: 19.2 (10) perceived: 6.1 (4.4)
Subramonia 2010 - 18.5 (11 - 28) 10 (4 - 13) 12.5 (4 - 21) 3 (0 - 7)

median (IQR)
*presented both laser techniques separately
**sick leave days taken and patient’s own perception of required sick leave
***adjusted according to the number of phlebectomies performed, and the type of anaesthetic used
Table 14. Type of anaesthetic used
Technique Study General anaesthesia Tumescent anaesthesia Local anaesthesia
Surgery Alt Rx Surgery Alt Rx Surgery Alt Rx
Foam versus FOAM- All (general or

surgery Study* spinal)
Magna 2007* All (general or

spinal)
Rasmussen All** All** All** All**

2011
Laser versus Darwood All All All All

surgery 2008
Flessenkaem- All All

per
2013
Table 14. Type of anaesthetic used (Continued)
HELP-1 All All All
Magna 2007 All (general or All All

spinal)
Pronk 2010 All All All All
Rasmussen All All All All

2007

2011
RELACS All** All** All** All**

Study
RFA versus EVOLVeS 19/36 12/44 17/36 32/44 17/36 32/44

surgery Study
Helmy All All All

ElKaffas 2011

2011
Rautio 2002 All All
Subramonia All All All All

2010
*Anaesthesia not reported for one or more treatment groups
**Sedation at surgeon’s discretion
APPENDICES
Appendix 1. CENTRAL search strategy
#1 MeSH descriptor: [Sclerotherapy] explode all trees 453
#2 MeSH descriptor: [Sclerosing Solutions] explode all trees 299
#3 sclero*:ti,ab,kw 5977
#4 tetradecyl near/2 (sulfate or sulphate):ti,ab,kw 56
#5 MeSH descriptor: [Sodium Tetradecyl Sulfate] explode all trees 37
#6 MeSH descriptor: [Saline Solution, Hypertonic] explode all 373

trees
#7 MeSH descriptor: [Ethanolamines] this term only 1389
#8 polydocanol or polidocanol:ti,ab,kw 156
#9 saline:ti,ab,kw 13867
#10 ethanolamine near/2 oleate:ti,ab,kw 58
#11 sodium near/2 morrhuate:ti,ab,kw 18
#12 sotradecol:ti,ab,kw 6
#13 aetoxisclerol or aethoxysclerol:ti,ab,kw 14
#14 aetoxiskerol or aethoxyskerol:ti,ab,kw 1
#15 Turbofoam:ti,ab,kw 2
#16 foam* or microfoam*:ti,ab,kw 993
#17 varisolve:ti,ab,kw 2
#18 MeSH descriptor: [Laser Therapy] explode all trees 2963
#19 (endovenous or EVLA or EVLT or radiofrequency or laser* or 10714

ablation* or obliteration* or RFA):ti,ab,kw
#20 MeSH descriptor: [Catheter Ablation] explode all trees 1041
#21 #1 or #2 or #3 or #4 or #5 or #6 or #7 or #8 or #9 or #10 or 32434

#11 or #12 or #13 or #14 or #15 or #16 or #17 or #18 or #
19 or #20
(Continued)
#22 MeSH descriptor: [Vascular Surgical Procedures] explode all 11931

trees
#23 MeSH descriptor: [Ligation] explode all trees 533
#24 surg* or ligat* or strip* or phlebectomy:ti,ab,kw 81940
#25 #22 or #23 or #24 88224
#26 #21 and #25 7845
#27 MeSH descriptor: [Varicose Veins] explode all trees 777
#28 MeSH descriptor: [Venous Insufficiency] explode all trees 362
#29 MeSH descriptor: [Saphenous Vein] explode all trees and with 173
qualifier(s): [Surgery - SU]
#30 (varic* or incomp* or insuffici* or tortuous or sapheno* or 16631

GSV or CVI):ti,ab,kw
#31 #27 or #28 or #29 or #30 16658
#32 #26 and #31 in Trials 545
WHAT’S NEW
Date Event Description
3 June 2014 New search has been performed Searches re-run. Eight additional included studies and 12
additional excluded studies identified
3 June 2014 New citation required but conclusions have not changed Searches re-run. Eight additional included studies and 12
additional excluded studies identified. Review text up-
dated accordingly. New author joined review team
CONTRIBUTIONS OF AUTHORS
For the current update of the review, study selection, data extraction, quality assessment and manuscript development were performed
by CN and RB. GS resolved any disagreements regarding study inclusion. RE, PC, HB, VB and GS provided support by appraising
the manuscript and adding comments and suggestions.
For the previous version of this review CN, RE, VB, PC, HB and GS independently reviewed studies for inclusion. Data extraction was
performed by CN and RE, and was cross-checked by VB. CN, RE and VB assessed the methodological quality of trials and extracted
data. GS resolved any disagreements regarding methodological quality of trials. The final contents of the review were agreed by GS and
VB.
DECLARATIONS OF INTEREST
VB: I am a co-editor, along with Professor Stansby, of a textbook entitled ’Postgraduate Vascular Surgery - A Candidate’s guide to the
FRCS’ published by Cambridge University Press and thus entitled to royalties.
SOURCES OF SUPPORT
Internal sources
• No sources of support supplied
External sources
• Chief Scientist Office, Scottish Government Health Directorates, The Scottish Government, UK.
The PVD Group editorial base is supported by the Chief Scientist Office.
DIFFERENCES BETWEEN PROTOCOL AND REVIEW

The title was altered for clarity: the term ’conventional surgery’ can be misleading, ’open surgery’ was used in its place. This term has
been changed throughout the text, as well.
Additional outcome data were collected on post-operative pain and time to return to work and normal activities.
For the primary outcome ’recurrence or recanalisation’, the description ’determined by ultrasound imaging, at set time intervals post-
procedure’ was removed. The included studies had many different definitions and ways of evaluating this outcome, which are presented
in the Characteristics of included studies, and therefore using only ’determined by ultrasound imaging’ was limiting the scope of the
outcome.
The third primary outcome ’need for re-intervention due to failure of a procedure (technical failure)’ was amended to ’technical failure’
as re-intervention and technical failure are not necessarily the same thing. Table 6 shows the available data for re-intervention.
The term ’operative costs’ was changed to ’procedural costs’. This is more inclusive as UGFS is not technically considered an operation.
’Type of anaesthetic’ was removed from the secondary outcomes list as it is a characteristic rather than an outcome. It is reported under
’Additional outcomes’ and information is provided in Table 14.
INDEX TERMS
Medical Subject Headings (MeSH)

∗ Saphenous Vein; Catheter Ablation [∗ methods]; Endovascular Procedures [∗ methods]; Laser Therapy [∗ methods]; Randomized Con-
trolled Trials as Topic; Recurrence; Sclerotherapy [∗ methods]; Varicose Veins [∗ therapy]
MeSH check words

Humans

Database - of - Systematic - Reviews in Varicose Vein

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Database - of - Systematic - Reviews in Varicose Vein

Uploaded by

Copyright:

Available Formats

Cochrane Database of Systematic Reviews

Endovenous ablation (radiofrequency and laser) and foam

Nesbitt C, Bedenis R, Bhattacharya V, Stansby G

Nesbitt C, Bedenis R, Bhattacharya V, Stansby G.

Endovenous ablation (radiofrequency and laser) and foam

Craig Nesbitt1 , Rachel Bedenis2 , Vish Bhattacharya3 , Gerard Stansby4

Editorial group: Cochrane Vascular Group.

PLAIN LANGUAGE SUMMARY

Description of the intervention

Assessment of risk of bias in included studies

Dealing with missing data

Subgroup analysis and investigation of heterogeneity

(i) Foam versus surgery

(ii) Laser versus surgery

Characteristics of included studies [ordered by study ID]

Methods Study design: Prospective, randomised controlled trial

Interventions Treatment(s): 2 EVLT techniques:

Bias Authors’ judgement Support for judgement

Allocation concealment (selection bias) Low risk Used sealed envelopes

Other bias Unclear risk Authors reported difficulty recruiting pa-

concomitant mini-phlebectomies. This

Methods Study design: Multicentre, prospective, randomised controlled trial

presented rates of neovascularisation in the groin and recurrence at 2 years

Bias Authors’ judgement Support for judgement

Methods Study design: Multicentre, prospective, randomised controlled trial

Outcomes Primary outcomes: Inguinal venous reflux after 2 years

Notes May 2005 to July 2009

Bias Authors’ judgement Support for judgement

Other bias Unclear risk Possibly underpowered; power calculation

Methods Study design: Multicentre, prospective randomised controlled trial

Participants No of patients randomised: Total n = 460 (UGFS n = 233; HL/S n = 227)

Outcomes Primary outcomes: Recurrence

Notes October 2005 to December 2007

Bias Authors’ judgement Support for judgement

Allocation concealment (selection bias) Low risk Used computer-generated randomisation

Selective reporting (reporting bias) Low risk All outcomes reported on

Other bias Unclear risk Mini-phlebectomies were performed at the

As seen in the commentary letter from MJ

Helmy ElKaffas 2011

Methods Study design: Prospective, randomised controlled trial

Participants No of patients randomised: Total n = 180 (RFA n = 90; HL/S n = 90)

Interventions Treatment(s): Ultrasound guided radiofrequency ablation- RFA Closuresystem, using

Notes Conducted between May 2006 and January 2009

Bias Authors’ judgement Support for judgement

Random sequence generation (selection Unclear risk No description of sequence generation

Methods Study design: Single-centre, prospective, randomised controlled trial

Participants No of patients randomised: Total n = 280 (EVLT n = 140; HL/S n = 140)

Outcomes Primary outcomes: QoL (UK SF-36 V1); recurrence

Notes September 2004 to March 2009

Bias Authors’ judgement Support for judgement

Selective reporting (reporting bias) Low risk All outcomes reported on

Other bias Unclear risk Possibly underpowered; power calculation

Methods Study design: SIngle-centre*, prospective, randomised controlled trial

Outcomes Primary outcomes: Anatomic success according to DUS, neovascularisation

Notes January 2007 to May 2010

Bias Authors’ judgement Support for judgement

Other bias Unclear risk Possibly underpowered; power calculation

Methods Study design: Single-centre, prospective, non-blinded, randomised controlled trial

Age - mean years (SD): EVLT 49 (11.0); HL/S 50 (10.5)

Notes June 2007 to December 2008

Bias Authors’ judgement Support for judgement

Allocation concealment (selection bias) Low risk Used computer randomisation

Selective reporting (reporting bias) Low risk All outcomes reported on