Phlebotonics for venous insufficiency

See: Summary of findings 1 Do phlebotonics improve signs and symptoms of venous insufficiency when compared with placebo?

Summary of findings 1. Do phlebotonics improve signs and symptoms of venous insufficiency when compared with placebo?

Outcomes	Anticipated absolute effects *		Relative effect (95% CI)	Number of participants (RCTs)	Certainty of the evidence (GRADE)	Comments
Phlebotonics compared with placebo for venous insufficiency
Patient or population: patients with venous insufficiency Settings: hospital and ambulatory settings Intervention: phlebotonics Comparison: placebo
Outcomes	Risk with placebo	Risk with phlebotonics	Relative effect (95% CI)	Number of participants (RCTs)	Certainty of the evidence (GRADE)	Comments
Oedema in the lower legs (dichotomous variable) Follow‐up: 1‐6 months	575 per 1000	403 per 1000 (362 to 449)	RR 0.70 (0.63 to 0.78)	1245 (13 studies)	⊕⊕⊕⊝ Moderate^a	Phebotonics probably slightly reduce oedema in the lower limb compared to placebo
Oedema in the lower legs (ankle circumference, mm) Follow‐up: 1‐12 months	‐	Mean ankle circumference in the lower legs in the phlebotonic groups was 4.27 mm lower (5.61 to 2.93 lower) than in the placebo groups	‐	2010 (15 studies)	⊕⊕⊕⊝ Moderate^b	Phlebotonics probably slightly reduce ankle perimeter circumference compared to placebo
Quality of life (CIVIQ and other questionnaires) Follow‐up: mean 2‐12 months	‐	The QoL in the phlebotonic groups was 0.06 SMD lower (0.22 lower to 0.1 higher) than in the placebo groups	‐	1639 (5 studies)	⊕⊕⊕⊝ Moderate^c	Phebotonics probably make little or no difference to QoL compared with placebo
Ulcer healing (dichotomous variable) Follow‐up: 1‐12 months	381 per 1000	358 per 1000 (301 to 430)	RR 0.94 (0.79 to 1.13)	461 (6 studies)	⊕⊕⊝⊝ Low^d	Phlebotonics may make little or no difference to ulcer healing compared to placebo
Adverse events Follow‐up: 1‐12 months	158 per 1000	180 per 1000 (161 to 200)	RR 1.14 (1.02 to 1.27)	5789 (37 studies)	⊕⊕⊕⊝ Moderate^e	Phlebotonics probably slightly increase adverse events compared to placebo
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; CIVIQ: Chronic Venous Insufficiency International Questionnaire; QoL: quality of life; RCT: randomised controlled trial; RR: risk ratio
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
^aThe certainty of the evidence was downgraded (1 level) to moderate because of overall risk of bias (10 studies had an unclear risk of bias and two had a high risk of bias) ^bThe certainty of the evidence was downgraded (1 level) to moderate because of overall risk of bias (11 studies had an unclear risk of bias and one had a high risk of bias) ^cThe certainty of the evidence was downgraded (1 level) to moderate because of overall risk of bias (one study had an unclear risk of bias and two had a high risk of bias) ^dThe certainty of the evidence was downgraded (2 levels) to low because of overall risk of bias (four studies had an unclear risk of bias and two had a high risk of bias) and imprecision (low number of events) ^eThe certainty of the evidence was downgraded (1 level) to moderate because of overall risk of bias (28 RCTs had unclear risk of bias and four RCTs had high risk of bias)

Background

Description of the condition

Chronic venous insufficiency (CVI) is a condition in which veins are unable to transport blood unidirectionally toward the heart with flow adapted to tissue drainage needs, temperature regulation and haemodynamic reserve, regardless of their position and activity. CVI first manifests as an increase in venous tension (venous hypertension, or high blood pressure in the veins) with or without reflux (Kurz 1999). Depending on its cause, CVI can be congenital, primary (with undetermined cause) or secondary (post‐thrombotic, post‐traumatic or other). Depending on its pathophysiology, CVI can be related to occlusion (blocked veins), reflux or both. Finally, it might depend on superficial or deep venous systems or on perforator anomalies (Porter 1995).

CVI is an important cause of discomfort and inability to work, and many people find it difficult to live with this condition. Its prevalence has not been clearly determined because available studies regarding this subject are few, and those that are available present limitations. Some studies do not cover the whole pathological spectrum and focus only on varicose veins or ulcers; others do not use standardized definitions of the illness and apply a variety of diagnostic criteria (Nicolaides 2000). As a result, prevalence has been estimated at between 1% and 50% (Evans 1999; Stanhope 1975; Van den Oever 1998). The Framingham Study showed an annual incidence of 2.6% among women and 1.9% among men (Brand 1988). In a recent publication of the Edinburgh Vein Study, annual incidence of CVI was reported as 1% among the general population of the UK (Robertson 2014).

Causes of CVI are unknown, although it has been associated with venous dilation, deformity and valvular venous incompetence. Trophic skin disorders and venous ulcers result from severe varicose illness (Carpentier 2000). Varicose veins have a multi‐factorial origin related to advanced age and certain lifestyles (sedentary life), pregnancy, hereditary factors and obesity. Risk of ulcers may be increased by trauma and previous episodes of deep venous thrombosis (clinical or subclinical) (Scott 1995).

Clinical manifestations of CVI differ according to stage of the illness and can include feelings of heaviness in the extremities, paraesthesia (tingling), cramps, pain, oedema (swellings), varicose veins, skin pigmentation, varicose sores and signs of skin atrophy (wasting). Symptoms are frequently related to extent of disease. Underlying venous disease (superficial, deep or both, with or without obstruction) has a major impact on both manifestations of the disease and response to treatment. Since 1994, criteria develop by the International Consensus Committee on Chronic Venous Disease have been used to define and classify CVI in a standardized fashion (Porter 1995). According to this Consensus, clinical signs (C), aetiology (E), anatomical distribution (A) and physiological conditions (P) ("Clinical‐Etiology‐Anatomy‐Pathophysiology"; CEAP) are used to classify CVI (Porter 1995). A later revision of the CEAP classification established a means of differentiating between chronic venous disorder (referring to all morphology and functional abnormalities of the venous system) and CVI (reserved for more advanced stages of the disease with oedema, skin changes or venous ulcers) (Eklöf 2004). In parallel, a venous clinical severity score (ranging from none (0) to severe (3)) was established to assess pain, varicose veins, venous oedema, skin pigmentation, inflammation, induration, active ulcer (number, duration and size) and use of compression therapy (Vasquez 2010). Recently, a new version of CEAP classification has been published (Lurie 2020), in which Corona phlebectatica was added as a C4c clinical subclass, the modifier “r” introduced for recurrent varicose veins and recurrent venous ulcers and numeric descriptions of the venous segments replaced by their common abbreviations (Lurie 2020).

Description of the intervention

Surgery, sclerotherapy and mechanical compression are generally the preferred treatments for CVI. However, pharmacological treatments or phlebotonics are often used because they are easy to administer, and because compliance with compressive treatments (such as elastic stockings) is often poor.

Phlebotonics represent a heterogeneous group of medications used to treat CVI. Most of these drugs are natural flavonoids extracted from plants. Synthetic products with flavonoid‐like properties are also used to treat venous disorders. In the Anatomical Therapeutic Chemical (ATC) system, phlebotonics are classified as vasoprotective agents (ATC 2015). Within this classification system, active substances are divided into different groups according to the organ or system on which they act and their therapeutic, pharmacological and chemical properties. Phlebotonics are known as venoactive drugs whose mechanism of action is not scientifically well established despite the availability of numerous studies examining their pharmacological and clinical properties. These medications are associated with effects on macrocirculation (e.g. they may improve venous tone) (Tsouderos 1991) and on microcirculatory parameters (e.g. they may decrease capillary hyperpermeability) (Behar 1988).

Why it is important to do this review

Lower limb CVI affects a predominantly adult population and it is a frequent cause for a referral from primary to secondary care (Venous Forum 2011). Although phlebotonics are commercialised in many countries, in others they are not widely available. In some countries, such as Spain, for certain phlebotonics (calcium dobesilate, chromocarbe and naftazone) the CVI indication has been withdrawn, and for several other phlebotonics, such as aminaftone, diosmine, hidrosmine, escin and some rutosides, conditions of use during exacerbations of CVI have been limited to two or three months by the Spanish Ministry of Health (AEM 2002).

Controversy surrounds the clinical relevance of the efficacy and benefit‐risk balance of phlebotonics. Case‐control studies have found that risk of agranulocytosis (reduced numbers of white blood cells, mainly neutrophils) is associated with some phlebotonics (Ibañez 2000; Ibáñez 2005; Kaufman 1991). As efficacy is not well defined and serious harmful effects have been associated with phlebotonics, an evaluation of available evidence is needed.

Objectives

To assess the efficacy and safety of phlebotonics administered orally or topically for treatment of signs and symptoms of lower extremity CVI.

Methods

Criteria for considering studies for this review

Types of studies

We included randomised, double‐blind, controlled trials assessing the efficacy and/or safety of phlebotonics compared with placebo in patients with chronic venous insufficiency (CVI) at any stage of the disease. We did not include studies which were not RCTs or double‐blind. We did not choose specific diagnostic classifications of CVI a priori because most of the studies were carried out before 1994, before the international diagnostic consensus of CVI. Therefore, we included RCTs with different diagnostic criteria. We included studies in which use of compression measures (support tights) was similar across groups.

Types of participants

We included both male and female participants who were 18 years of age and older, suffering from any type of CVI. CVI could be diagnosed according to explicit clinical criteria and/or by objective instruments. Participant background, ethnicity and medical co‐morbidities at the beginning of the study did not influence the decision to include or exclude the study. We excluded studies that included participants with active thrombophlebitis and those including pregnant women.

Types of interventions

We considered the following interventions to treat CVI acceptable for inclusion: treatments including venoactive drugs or phlebotonics, administered orally or topically, at any dosage and independently of the duration of treatment, compared with placebo. We excluded studies that compared phlebotonics among themselves or with any other therapeutic method (i.e. support tights or surgery).

Natural products
- Flavonoids: rutoside, French maritime pine bark extract (also known as pycnogenol), grape seed extract, diosmine and hidrosmine, disodium flavodate
- Saponosides: Centella asiatica
Synthetic products
- Calcium dobesilate, naftazone, aminaftone, chromocarbe

We excluded escin (horse chestnut seed extract), as it is covered in another Cochrane Review (Pittler 2012).

Pentoxifylline is classified as a peripheral vasodilator, not as a vasoprotective agent (ATC 2015); therefore, we excluded it from this review.

Types of outcome measures

We included studies that assessed any of the following outcome measures.

Primary outcomes

Oedema in the lower limb measured by the dichotomous variable 'oedema' and the continuous variables 'ankle perimeter circumference' and 'volume of the leg'
Specific quality of life (QoL) scales (e.g. Chronic Venous Insufficiency International Questionnaire (CIVIQ))

Secondary outcomes

Assessment of CVI: objective signs
- Skin manifestations including venous ulcer healing and trophic alterations (e.g. lipodermatosclerosis (hardening of the skin that may cause red/brown pigmentation and is accompanied by wasting of subcutaneous fat), telangiectasia (tiny blood vessels cause threadlike red lines or patterns on the skin), reticular veins (dilated veins that show as a net‐like pattern on the skin), varicose veins (permanently dilated veins)
Assessment of CVI: subjective symptoms
- Pain in the lower legs
- Cramps in the lower legs
- Restless legs
- Itching in the lower legs
- Feeling of heaviness in the lower legs
- Swelling in the lower legs
- Paraesthesias (abnormal sensations, such as prickling, burning, tingling) in the lower legs
- Participant satisfaction
Adverse events
- Adverse reactions experienced by participants during the trial, as reported by questionnaire or related by participants and specified within the publication

Search methods for identification of studies

Electronic searches

For this update, the Cochrane Vascular Information Specialist conducted systematic searches of the following databases for randomised controlled trials and controlled clinical trials without language, publication year or publication status restrictions:

the Cochrane Vascular Specialised Register via the Cochrane Register of Studies (CRS‐Web searched on 12 November 2019);
the Cochrane Central Register of Controlled Trials (CENTRAL) Cochrane Register of Studies Online (CRSO 2019, issue 10);
MEDLINE (Ovid MEDLINE® Epub Ahead of Print, In‐Process & Other Non‐Indexed Citations, Ovid MEDLINE® Daily and Ovid MEDLINE®) (searched from 1 January 2017 to 12 November 2019);
Embase Ovid (searched from 1 January 2017 to 12 November 2019);
CINAHL Ebsco (searched from 1 January 2017 to 12 November 2019); and
AMED Ovid (searched from 1 January 2017 to 12 November 2019).

The Information Specialist modelled search strategies for other databases on the search strategy designed for CENTRAL. Where appropriate, strategies were combined with adaptations of the highly sensitive search strategy designed by the Cochrane Collaboration for identifying randomised controlled trials and controlled clinical trials (as described in the Cochrane Handbook for Systematic Reviews of Interventions Chapter 6, Lefebvre 2011). Search strategies for major databases are provided in Appendix 1.

The Information Specialist searched the following trials registries on 12 November 2019:

the World Health Organization International Clinical Trials Registry Platform (who.int/trialsearch);
ClinicalTrials.gov (clinicaltrials.gov).

Searching other resources

For this update, we searched the reference lists of articles retrieved by electronic searches for additional citations.

Data collection and analysis

Selection of studies

For this update, two review authors (RV and DS) independently assessed the eligibility of new studies identified by the searches. A third review author (MMZ) helped to resolve disagreements by discussion.

Data extraction and management

For this update, two review authors (RV and MMZ) independently extracted data from new studies and entered them to a previously tested standardized form. A consensus between reviewers were reached if any data extraction discrepancies occurred. We collected information including characteristics of study participants, characteristics of intervention and control groups and outcome characteristics of every group of participants. For cross‐over studies, we extracted and analyzed only data related to the first period of treatment.

Assessment of risk of bias in included studies

For this update, two review authors (RV and MJMZ) independently assessed the risk of bias of the newly included studies. A consensus between review authors was reached by discussion when there was any disagreement. We specifically assessed the randomisation method (sequence generation and allocation concealment); blinding of participants, caregivers/study researchers and outcome assessors to the intervention; whether outcome data were incomplete; and presence of selection bias. Once this information was gathered, review authors classified each study into one of three levels of risk of bias: low, unclear or high, based on the criteria specified in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011).

Measures of treatment effect

We estimated effects of treatment with phlebotonics by using risk ratios (RRs) for dichotomous variables and mean differences (MDs) or standardized mean differences (SMDs) for continuous variables, along with their corresponding 95% confidence intervals (CIs). We calculated SMDs when studies used different instruments to measure the same variable.

Unit of analysis issues

We took the unit of analysis to be the individual participant. For cross‐over studies, we extracted and analyzed only data related to the first period of treatment.

Dealing with missing data

We analyzed dichotomous variables by applying the intention‐to‐treat (ITT) principle to analyze every individual in the randomly assigned treatment group regardless of whether individuals completed treatment or withdrew prematurely from the study. We included in the ITT analysis only studies that provided data from all randomised participants, or that stated the number of participants lost during follow‐up. We numerically imputed missing values due to withdrawal of participants or loss to follow‐up as therapeutic failures in both comparative groups. For continuous variables, we analyzed data as provided by study authors, either per protocol or as ITT values.

Assessment of heterogeneity

We carried out an analysis to detect the presence of heterogeneity by using the I² statistic before obtaining global effect estimators. The I² statistic describes the percentage of total variation across studies that is due to heterogeneity rather than to sampling error (Deeks 2011). When statistical heterogeneity was high (I² > 75%), we did not pool studies. For levels of I² less than 50%, we applied a fixed‐effect model; for levels of I² greater than 50% but less than 75%, we applied a random‐effects model (DerSimonian 1986).

Assessment of reporting biases

We constructed a funnel plot to assess whether the outcome of oedema (dichotomous variable) was subject to publication bias (Figure 1).

Figure 1

Funnel plot of comparison: 1 Phlebotonics vs placebo, outcome: 1.1 Oedema in the lower legs (dichotomous variable).

Data synthesis

We obtained data from the included studies for variables evaluated at the end of treatment. In addition, we obtained data from measures of change when no significant baseline differences were evident between compared groups. We then pooled these together with other similar continuous outcomes.

We split the outcomes of variables measured by ordinal categorical scales into two groups of response. We considered one group as showing success (no signs or symptoms or mild manifestations) and the other as showing failure (moderate, severe or very severe persistence of signs and symptoms).

Subgroup analysis and investigation of heterogeneity

We carried out subgroup analyses in addition to the overall analysis of phlebotonics. These included looking at the effects of the following phlebotonics: rutosides, hidrosmine, diosmine, calcium dobesilate, disodium flavodate, grape seed extract, French maritime pine bark extract, chromocarbe and aminaftone.

Sensitivity analysis

We performed sensitivity analyses to assess the influence on data of assumptions and decisions of review authors during the review process. We re‐analysed data by:

excluding studies that used compression measures;
excluding unpublished studies; and
excluding studies with high or unclear risk of bias in at least one domain.

Summary of findings and assessment of certainty of the evidence

We created one 'Summary of findings' table to present the main findings for 'Phlebotonics compared with placebo for venous insufficiency' using GRADE profiler software (GRADEpro 2008). See summary of findings Table 1. We used the principles of the GRADE system to assess the certainty of the body of evidence associated with the main outcomes listed below. The GRADE approach appraises the certainty of a body of evidence according to the extent to which one can be confident that an estimate of effect or association reflects the item being assessed. Evaluation of the certainty of a body of evidence considers within‐study risk of bias, indirectness of the evidence, inconsistency (heterogeneity in the data), imprecision (precision of effect estimates) and publication bias (Schünemann 2011).

Two review authors (DS and RV) independently assessed the certainty of the body of evidence for the following outcomes.

Oedema in the lower legs (dichotomous variable)
Oedema in the lower legs (circumference mm)
QoL
Ulcer healing
Adverse events

Results

Description of studies

Results of the search

For this update we identified three new included studies (NCT01848210; Rabe 2015; Rabe 2016); two new ongoing studies (Barattini 2019; NCT03833024); and two new studies were excluded (EudraCT2009‐014681‐25; ISRCTN54360155). See Figure 2. Details of all studies are provided in the Characteristics of included studies, Characteristics of ongoing studies and Characteristics of excluded studies tables.

Figure 2

Study flow diagram.

Included studies

For this update, we identified three new included studies (NCT01848210; Rabe 2015; Rabe 2016). In total with those identified in earlier versions, we included 69 studies. See Characteristics of included studies tables.

Most studies were published in English, but four were published in German (Biland 1982; Kiesewetter 1997; Koscielnny 1996; Pedersen 1992), seven in French (Cauwenberge 1978; Chassignolle 1994; Planchon 1990; Thebaut 1985; Vin 1994; Welch 1985; Zucarelli 1987), four in Spanish (Flota‐Cervera 2008; Klüken 1971; Marinello 2002; Serralde 1990), three in Italian (Allegra 1981; Lazzarini 1982; Pecchi 1990), and one in Spanish, French and Dutch (Padrós 1972).

Of the 69 included double‐blind, placebo‐controlled clinical trials, we did not include 13 studies in the efficacy analysis. Of these, 10 studies corresponded to the rutoside group (Bergqvist 1981; Cloarec 1994; Jongste 1986; Mann 1981; Nocker 1990; Prerovsky 1972; Renton 1994; Rose 1970; Rudofsky 1989; Sentou 1984), two corresponded to calcium dobesilate (Padrós 1972; Pecchi 1990) and another corresponded to French bark pine extract (Petrassi 2000).

We excluded these studies from the efficacy analysis for the following reasons.

Only mean data were provided without standard deviations (SDs) or standard errors (SEs) (Sentou 1984).
Medians were provided instead of means (Renton 1994).
Outcomes were reported by graph only (Nocker 1990; Rose 1970; Rudofsky 1989).
First period data were not provided in studies of cross‐over design (Padrós 1972; Prerovsky 1972).
No data were provided for any variable (Bergqvist 1981; Cloarec 1994; Jongste 1986).
Measured changes were reported when significant differences in baseline were noted between compared groups (Mann 1981; Petrassi 2000).
A quasi‐randomisation method was used in which treatments were alternatively allocated depending on participants' order of arrival (Pecchi 1990).
At baseline, a significant imbalance in the ulcer area was evident between groups (1130 mm² in the rutoside group vs 430 mm² in the placebo group; P = 0.039) (Mann 1981).

Of the 56 studies with oral phlebotonics included in the efficacy analysis, studied phlebotonics corresponded to 28 studies of rutosides (Balmer 1980; Burnand 1989; Cloarec 1996; Cauwenberge 1972; Cauwenberge 1978; Cesarone 2002; Cornu‐Thenard 1985; Diebschlag 1994; Ihme 1996; Jongste 1986; Jongste 1989; Kiesewetter 1997; Koscielnny 1996; Klüken 1971; Kriner 1985; Languillat 1988; Laurent 1988; MacLennan 1994; NCT01848210; Parrado 1999; Pedersen 1992; Pulvertaft 1983; Schultz‐Ehrenburg 1993; Serralde 1990; Unkauf 1996; Vanscheidt 2002a; Vanscheidt 2002b; Vin 1994), 11 of hidrosmine and diosmine (Chassignolle 1994; Danielsson 2002; Dominguez 1992; Fermoso 1992; Gilly 1994; Guilhou 1997; Planchon 1990; Rabe 2015; Thebaut 1985; Welch 1985; Zucarelli 1987), 10 of calcium dobesilate (Casley‐Smith 1988; DOBESILATO500/2; Flota‐Cervera 2008; Hachen 1982; Labs 2004; Marinello 2002; Martinez‐Zapata 2008; Rabe 2011; Rabe 2016; Widmer 1990), two of Centella asiatica (Allegra 1981; Pointel 1986), two of aminaftone (Belczak 2014; Lazzarini 1982), two of French maritime pine bark extract (Arcangeli 2000; Petrassi 2000) and one of grape seed extract (Thebaut 1985). No studies using topical phlebotonics or chromocarbe or naftazone or disodium flavodate fulfilled the inclusion criteria. Length of treatment and participant follow‐up ranged from 28 days to four months, except for three studies, in which follow‐up lasted six months or more (DOBESILATO500/2; MacLennan 1994; Martinez‐Zapata 2008).

Overall, we included 7690 participants in the meta‐analysis; 83% were female and 17% were male; mean age was 50 years (range 32 to 62 years). The mean number of participants included per clinical trial was 150 (range 20 to 1137). All participants met the respective CVI criteria of every study, although we noted variation between studies in degree of progression to CVI, as well as in diagnostic classification criteria applied. Only 22% of studies reported the diagnostic classification used. Among studies that did report on the diagnostic classification of CVI, the CEAP classification was used most often (Belczak 2014; Danielsson 2002; DOBESILATO500/2; Labs 2004; Marinello 2002; Martinez‐Zapata 2008; Rabe 2011; Rabe 2015; Rabe 2016; Vanscheidt 2002a; Vanscheidt 2002b), followed by Widmer's classification (Casley‐Smith 1988; Cloarec 1996; Koscielnny 1996; Parrado 1999; Unkauf 1996). Wert's was the only other classification used (Kiesewetter 1997).

Differences in severity of disease were observed: some studies were performed with participants at early and symptomatic CVI stages (Cornu‐Thenard 1985; Danielsson 2002; Gilly 1994; Hachen 1982; Thebaut 1985), and others included participants at advanced stages because of long progression of the disease or the presence of venous ulcers (Casley‐Smith 1988; DOBESILATO500/2; Guilhou 1997; Lazzarini 1982; Marinello 2002; Planchon 1990; Schultz‐Ehrenburg 1993; Vanscheidt 2002a). However, most studies included participants at moderate CVI stages with oedema, skin pigmentation, varicose veins and post‐thrombotic syndromes.

Ten studies specified that investigators used additional compression therapy (DOBESILATO500/2; Guilhou 1997; Laurent 1988; Lazzarini 1982; Marinello 2002; Martinez‐Zapata 2008; Planchon 1990; Rabe 2011; Schultz‐Ehrenburg 1993; Zucarelli 1987).

Eleven studies used a visual analogue scale (VAS) to measure subjective variables (Alterkamper 1987; Cesarone 2002; DOBESILATO500/2; Labs 2004; Martinez‐Zapata 2008; Rabe 2011; Rabe 2015; Unkauf 1996; Vanscheidt 2002b; Widmer 1990; Zucarelli 1987). Other studies used ordinal categorical scales with a scoring system from ‐3 to +1 (Hachen 1982), ‐1 to + 1 (Casley‐Smith 1988), 0 to 1 (Ihme 1996), 0 to 2 (Biland 1982; Ihme 1996; Kiesewetter 1997), 0 to 3 (Allegra 1981; Arcangeli 2000; Cloarec 1996; Cornu‐Thenard 1985; Danielsson 2002; Diebschlag 1994; Dominguez 1992; Gilly 1994; Jongste 1989; Languillat 1988; Laurent 1988; Lazzarini 1982; Parrado 1999; Planchon 1990; Pointel 1986; Pulvertaft 1983; Serralde 1990; Thebaut 1985; Tsouderos 1989; Welch 1985), 0 to 4 (Balmer 1980; Chassignolle 1994; Fermoso 1992; Flota‐Cervera 2008), 0 to 5 (NCT01848210; Rabe 2011), 0 to 7 (Labs 2004) or 0 to 9 (Dominguez 1992). Likewise, some of these scales were used to evaluate signs or objective variables such as oedema or trophic disorders. Methods used to measure oedema included metric tape to measure ankle or calf circumference and plethysmographic values (used in most studies) to determine leg volume.

Excluded studies

For this update, we identified two new studies that were excluded (EudraCT2009‐014681‐25; ISRCTN54360155). Four previously excluded studies were also identified by the search (Belczak 2014; Kiesewetter 1997; Prerovsky 1972; NCT01532882), making a total of 104 studies excluded for a variety of reasons (see Characteristics of excluded studies for details). We summarise the exclusion details below.

We excluded 58 studies because the intervention used by researchers was not included in this Cochrane Review (Akbulut 2010; Bacci 2003; Bastide 1976; Batchvarova 1989a; Behar 1993; Bello 1990; Bento 2006; Berson 1978; Bohm 1989; Bolliger 1972; Bosse 1985; Brami 1983; Carstens 1985; Cataldi 2001; Cesarone 2001b; Chiummariello 2009; Cospite 1996; de Parades 1990; Delacroix 1981; Delecluse 1991; Dustmann 1984; Erdlen 1989; Erler 1991; EudraCT2009‐014681‐25; Henriet 1995; Horvath 1985; Janssens 1999a; Kiesewetter 2000; Koltringer 1993; Krähenbühl 1975; Krcílek 1973; Languillat 1988b; Marastoni 1982; Monteil‐Seurin 1993; Morales 1993; NCT02191163; NCT02191254; NCT02191280; Neumann‐Mangoldt 1979; Nill 1970; Ottillinger 2001; Paciaroni 1982; Partsch 1981; Paul 1983; Pauschinger 1987; Pointel 1987b; Pokrovskii 2005; Rabe 2011b; Riccioni 2004; Sanctis 2001; Steiner 1990; Steiner 1992; Topalov 1990; Turio 2000; ISRCTN54360155; Weindorf 1987; Widmer 1972; Zuccarelli 1996).
We excluded 30 studies because researchers assessed no clinical endpoints or reported only outcomes not included in this Cochrane Review (Androulakis 1989; Auteri 1990; Belcaro 1995; Belcaro 2008; Boisseau 1995; Bort 1995; Cesarone 1992; Cesarone 1994; Cesarone 2001; Cesarone 2001c; Cesarone 2002b; Chant 1973; Clemens 1986; Duchene 1988; Forconi 1977; Gonzalez‐Fajardo 1990; Incandela 1995; Incandela 1996; Janssens 1999; Kalus 2004; Kostering 1985; Languillat 1989; Le Dévéhat 1989; Le Dévéhat 1997; Naser‐Hijazi 2004; Neumann 1988; Neumann 1990; Questel 1983; Roztocil 1977; Seydewitz 1992).
We excluded 16 studies because they were not double‐blinded (Belcaro 1989; Blume 1996; Cesarone 2001a; Cesarone 2010; De Anna 1989; De Sanctis 2001; Frausini 1985; Glinski 1999; Granger 1995; Incandela 2001; Incandela 2002; Menyhei 1994; NCT01654016; Petruzzellis 2002; Roztocil 2003; Steru 1988).

Ongoing studies

For this update, we identified two new ongoing studies (Barattini 2019; NCT03833024). This brings the total number of ongoing studies included to four (Barattini 2019; ISRCTN18841175; NCT01532882; NCT03833024). Details of these can be found in the Characteristics of ongoing studies table.

Risk of bias in included studies

Overall, only four studies (Labs 2004; Martinez‐Zapata 2008; Rabe 2011; Vanscheidt 2002a) were at low risk of bias (see Characteristics of included studies, Figure 3 and Figure 4).

Figure 3

Methodological quality graph: review authors' judgements about each methodological quality item presented as percentages across all included studies.

Figure 4

Methodological quality summary: review authors' judgements about each methodological quality item for each included study.

Allocation

Of the 69 studies included, 19 (28%) submitted details on the randomisation process and were assessed as being at low risk (see Figure 4 and Characteristics of included studies). The remaining studies were all judged to be at an unclear risk of bias.

Only 12 (17%) studies provided an accurate explanation of the allocation concealment process. Two used the sealed envelope method (Danielsson 2002; Pedersen 1992), four used indistinguishable number packaging (Biland 1982; Padrós 1972; Rabe 2011; Rose 1970), one used randomised numbered bottles provided by an external investigator (Belczak 2014), two used allocation concealment by direct phone calls (DOBESILATO500/2; Martinez‐Zapata 2008), and the remaining three studies used computerised random assignment (Jongste 1989; Labs 2004; Vanscheidt 2002a).

Blinding

Of the 69 studies included, 41 (59%) reported that the placebo used was identical to the active treatment; thus participants, study researchers and outcome assessors were blinded to the intervention and these were judged to have a low risk of bias. The other 28 studies did not mention whether placebo had identical characteristics to those of the active drug and so were at an unclear risk of bias (see Figure 4 and Characteristics of included studies).

Incomplete outcome data

Of the 69 studies included, 52 (75%) reported participant withdrawals, and thus were at low risk of bias. The percentage of withdrawn participants ranged from 0% to 42.5% (see Characteristics of included studies). Only eight (12%) studies included in the efficacy analysis stated that investigators carried out an ITT analysis (Dominguez 1992; Guilhou 1997; Ihme 1996; Martinez‐Zapata 2008; Rabe 2011; Rabe 2016; Unkauf 1996; Vanscheidt 2002a). Seven studies had high risk of bias in this domain (Cauwenberge 1978; DOBESILATO500/2; Mann 1981; Rabe 2016; Rose 1970; Sentou 1984; Vanscheidt 2002b): four described an important percentage of losses (42.5% Cauwenberge 1978; 18% Mann 1981; 39% Rose 1970; 34% Vanscheidt 2002b), one interrupted recruitment because financial support was interrupted (DOBESILATO500/2) and one did not specify the number of participants included (Sentou 1984). In the Rabe 2016 study, 14.8% of the randomised participants were lost during follow‐up and major protocol violations were reported for 42.4% of the randomised participants. Ten studies were judged to be at unclear risk of bias because the reasons for dropouts (Cauwenberge 1972), or the number of dropouts, were not provided (Cornu‐Thenard 1985; Kiesewetter 1997; Klüken 1971; Kriner 1985; Lazzarini 1982; Nocker 1990; Padrós 1972; Pedersen 1992), or the standard deviation was lacking in the results (Thebaut 1985).

Selective reporting

Of the 69 studies included, 57 (84%) reported all outcomes specified in the methods section and were judged as being at low risk of reporting bias. We evaluated seven studies as having high risk of selective reporting bias because we noted differences between outcomes reported in the methods and results sections (Cloarec 1994; Jongste 1986; Jongste 1989; Mann 1981; Rabe 2015), and because data before the cross‐over were not reported (Padrós 1972; Rose 1970). One study was interrupted, and results of this study were not published (DOBESILATO500/2). Lazzarini 1982 provided no information about adverse events. Two studies were judged to be at unclear risk of reporting bias because characteristics of participants were not provided ((Allegra 1981) and outcomes were not reported in methods and neither a protocol was published (Klüken 1971).

Figure 1 shows that all studies, except one (Casley‐Smith 1988), are located symmetrically around the effect measure at the top of the pyramid, indicating highly precise results (Cauwenberge 1972; Cornu‐Thenard 1985; Kiesewetter 1997; Klüken 1971; Kriner 1985; Lazzarini 1982; Nocker 1990; Padrós 1972; Pedersen 1992; Thebaut 1985). Apart from one imprecise study favouring phlebotonics, no small or heterogeneous studies provided results favouring placebo or phlebotonics (Casley‐Smith 1988).

Other potential sources of bias

No other potential sources of bias were detected.

Effects of interventions

See summary of findings Table 1 for the main comparison. Results of all analyzed outcomes are specified in an additional Table 1. Results of outcomes analyzed by active agent (aminaftone, calcium dobesilate, Centella asiatica, diosmine and hidrosmine, French maritime pine bark extract, grape seed extract and rutosides) are specified in Table 2; Table 3; Table 4; Table 5; Table 6; Table 7; and Table 8, respectively.

Table 1. Results of all outcomes analysed (all phlebotonics)

Variables	Dichotomous	Continuous
Oedema	RR 0.70 (0.63 to 0.78)	‐
Oedema (mm)	‐	MD ‐4.27 (‐5.61 to ‐2.93)
Oedema (volume)	‐	SMD ‐0.24 (‐0.33 to ‐0.15)
Ulcer cured	NS	‐
Trophic disorders	RR 0.87 (0.81 to 0.95)	‐
Pain	‐	SMD ‐0.35 (‐0.54, ‐0.17)
Cramps	RR 0.72 (0.58 to 0.89)	‐
Restless legs	RR 0.81 (0.72 to 0.91)	‐
Itching	‐	‐
Heaviness	‐	‐
Swelling	RR 0.63 (0.50 to 0.80)	‐
Paraesthesia	RR 0.67 (0.50 to 0.88)	NS
Quality of life	‐	NS
Global assessment by the participant	‐	‐
Adverse events	RR 1.14 (1.02 to 1.27)	‐
Note: No measures of effect are specified when I² was > 75% for the subgroup

RR: risk ratio
MD: mean difference
NS: non‐significant
RR: risk ratio
SMD: standardized mean difference

Table 2. Results by pharmacological group: aminaftone

Variables	Dichotomous	Continous
Oedema	RR 0.53 (0.28 to 0.99)	SMD ‐0.17 (‐0.61 to 0.28)
Ulcer cured	NS	‐
Trophic disorder	NS	‐
Pain	RR 0.43 (0.23 to 0.79)	‐
Cramps	RR 0.56 (0.31 to 0.99)	‐
Itching	RR 0.53 (0.31 to 0.91)	‐
Heaviness	RR 0.32 (0.17 to 0.60)	‐
Quality of live	‐	MD ‐10.00 (‐17.01 to ‐2.99)
Adverse events	NS	‐
Note: Only 1 study was analyzed

MD: mean difference
NS: non‐significant
RR: risk ratio
SMD: standardized mean difference

Table 3. Results by pharmacological group: calcium dobesilate

Variables	Dichotomous	Continuous
Oedema	‐	‐
Oedema (mm)	‐	NS
Oedema (volume)	‐	SMD ‐0.38 (‐0.51 to ‐0.24)
Ulcer cured	NS	‐
Pain	RR 0.53 (0.35 to 0.82)	NS
Cramps	RR 0.65 (0.50 to 0.84)	‐
Restless legs	RR 0.73 (0.59 to 0.91)	NS
Itching	‐	NS
Heaviness	NS	NS
Swelling	RR 0.19 (0.08 to 0.41)	NS
Paraesthesia	NS	‐
Quality of life	‐	NS
Global assessment by the participant	‐	SMD ‐0.52 (‐0.71 to ‐0.33)
Adverse events	RR 1.22 (1.0 to 1.49)	‐
Note: No measures of effect are specified when I² was > 75% for the subgroup

NS: non‐significant
RR: risk ratio
SMD: standardized mean difference

Table 4. Results by pharmacological group: Centella asiatica

Variables	Dichotomous	Continuous
Heaviness	NS	‐
Global assessment by the participant	RR 0.28 (0.14 to 0.57)	‐
Adverse events	NS	‐
Note: Only 1 study was analyzed

NS: non‐significant
RR: risk ratio

Table 5. Results by pharmacological group: diosmine, hidrosmine

Variables	Dichotomous	Continuous
Oedema	RR 0.63 (0.46 to 0.86)	‐
Oedema (mm)	‐	MD ‐5.98 (‐7.78 to ‐4.18)
Ulcer cured	NS	‐
Trophic disorder	RR 0.87 (0.81 to 0.94)	‐
Pain	NS	SMD ‐0.23 (‐0.41 to ‐0.05)
Cramps	RR 0.83 (0.70 to 0.98)	SMD ‐0.46 (‐0.78 to ‐0.14)
Restless legs	NS	‐
Itching	NS	‐
Heaviness	NS	SMD ‐0.69 (‐1.02 to ‐0.36)
Swelling	RR 0.70 (0.52 to 0.94)	SMD ‐0.92 (‐1.26 to ‐0.58)
Paraesthesia	NS	NS
Quality of life	‐	NS
Global assessment by the participant	‐	SMD ‐0.81 (‐1.14 to ‐0.47)
Adverse events	NS	‐
Note: No measures of effect are specified when I² was > 75% for the subgroup

MD: mean difference
NS: non‐significant
RR: risk ratio
SMD: standardized mean difference

Table 6. Results by pharmacological group: French maritime pine bark extract

Variables	Dichotomous	Continuous
Pain	RR 0.66 (0.48 to 0.91)	SMD ‐1.39 (‐2.09 to ‐0.69)
Heaviness	NS	SMD ‐1.50 (‐2.21 to ‐0.79)
Swelling	NS	SMD ‐1.65 (‐2.38 to ‐0.92)
Note: Only 1 study was analyzed

NS: non‐significant
RR: risk ratio
SMD: standardized mean difference

Table 7. Results by pharmacological group: grape seed extract

Variables	Dichotomous	Continuous
Oedema	NS	‐
Adverse events	NS	NS
Note: Only 1 study was analyzed

NS: non‐significant

Table 8. Results by pharmacological group: rutosides

Variables	Dichotomous	Continuous
Oedema	RR 0.72 (0.64 to 0.81)	‐
Oedema (mm)	‐	NS
Oedema (volume)	‐	SMD ‐0.15 (‐0.16 to ‐0.03)
Ulcer cured	NS	‐
Trophic disorder	NS	‐
Pain	‐	SMD ‐0.71 (‐1.23 to ‐0.19)
Cramps	RR ‐0.83 (‐1.50 to ‐0.16)	NS
Restless legs	NS	‐
Itching	‐	SMD ‐0.58 (‐1.10 to ‐0.06)
Heaviness	RR 0.60 (0.48 to 0.74)	‐
Swelling	RR 0.67 (0.50 to 0.88)	NS
Paraesthesias	RR 0.55 (0.37 to 0.83)	NS
Global assessment by the participant	‐	‐
Adverse events	RR 1.22 (1.04 to 1.43)	‐
Note: No measures of effect are specified when I² was > 75%

NS: non‐significant
RR: risk ratio
SMD: standardized mean difference

Of the 69 included studies, we excluded 13 studies from the efficacy analysis for the reasons explained under Included studies (Bergqvist 1981; Cloarec 1994; Jongste 1986; Mann 1981; Nocker 1990; Padrós 1972; Pecchi 1990; Petrassi 2000; Prerovsky 1972; Renton 1994; Rose 1970; Rudofsky 1989; Sentou 1984). Belczak 2014 compared three different interventions with placebo. For the analysis, we included only the comparison of aminaftone with placebo because the other two interventions were combinations of different drugs (micronised diosmine and hesperidin; coumarin and troxerutin).

Primary outcomes

Oedema in the lower limb (dichotomous variable)

We included 13 trials in the meta‐analysis: seven corresponding to rutosides (Cauwenberge 1972; Cauwenberge 1978; Cloarec 1996; Ihme 1996; Kriner 1985; MacLennan 1994; Welch 1985), two to calcium dobesilate (Casley‐Smith 1988; Labs 2004), two to hidrosmine and diosmine (Fermoso 1992; Planchon 1990), one to grape seed extract (Thebaut 1985) and one to aminaftone (Lazzarini 1982), with a total of 626 participants in the active treatment group and 619 in the placebo group. The median time to follow‐up was 49 days. Phebotonics probably reduce oedema in the lower limb compared to placebo (RR 0.70, 95% CI 0.63 to 0.78; 13 studies; 1245 participants; moderate‐certainty evidence; Analysis 1.1). The certainty of the evidence was downgraded by one step to moderate because of overall risk of bias (10 studies had an unclear risk of bias and two had a high risk of bias) (summary of findings Table 1). No differences between the subgroups was detected (test for subgroup differences: P = 0.74).

Oedema in the lower limb (continuous variables)

Ankle perimeter circumference

We included 15 studies in the meta‐analysis: seven corresponding to rutosides (Cloarec 1996; Cornu‐Thenard 1985; Jongste 1989; MacLennan 1994; Parrado 1999; Vin 1994; Welch 1985), five to calcium dobesilate (Flota‐Cervera 2008; Labs 2004; Martinez‐Zapata 2008; Rabe 2011; Widmer 1990), and three to diosmine (Gilly 1994; Planchon 1990; Tsouderos 1989), with a total of 1001 participants given active treatment and 1009 given placebo. The median time to follow‐up was 60 days. Phlebotonics probably slightly reduce ankle perimeter circumference compared to placebo (MD ‐4.27 mm; 95% CI ‐5.61 to ‐2.93; 15 studies; 2010 participants; moderate‐certainty evidence; Analysis 1.2). The certainty of the evidence was downgraded by one step to moderate because of overall risk of bias (11 studies had an unclear risk of bias and one had a high risk of bias). Differences between the subgroups was detected (test for subgroup differences: P = 0.02) due to a larger effect of diosmin‐hidrosmin.

Volume of the leg

We included 11 studies in the analysis: six corresponding to rutosides (Burnand 1989; Diebschlag 1994; Ihme 1996; Kiesewetter 1997; NCT01848210; Vanscheidt 2002a), four to calcium dobesilate (Casley‐Smith 1988; Rabe 2011;Rabe 2016; Widmer 1990) and one to aminaftone (Belczak 2014), with a total of 686 participants treated with phlebotonics and 706 with placebo. Phlebotonics probably slightly reduce volume of the leg compared to placebo (SMD ‐0.24 mL; 95% CI ‐0.33 to ‐0.15; 11 studies; 2072 participants; moderate‐certainty evidence; Analysis 1.3). The certainty of the evidence was downgraded (1 level) to moderate because of overall risk of bias (seven studies had an unclear risk of bias and one had a high risk of bias). Differences between the subgroups was detected (test for subgroup differences: P = 0.04) due to a larger effect of calcium dobesilate.

QoL

Seven studies evaluated QoL (Belczak 2014; Martinez‐Zapata 2008; Rabe 2011; Rabe 2015; Rabe 2016; Vanscheidt 2002a; Vanscheidt 2002b). Vanscheidt 2002a and Vanscheidt 2002b assessed QoL by using a questionnaire (EuroQol Measure of Health‐Related QoL and Freiburg Life Quality Assessment, respectively) and therefore did not provide quantifiable results. Martinez‐Zapata 2008, Rabe 2011, Rabe 2015 and Rabe 2016 evaluated QoL via the Chronic Venous Insufficiency International Questionnaire (CIVIQ). Belczak 2014 used a specific questionnaire for chronic venous disease adapted from Cesarone 2006. Phebotonics probably make little or no difference to QoL compared with placebo (SMD ‐0.06, 95% CI ‐0.22 to 0.10; five studies; 1639 participants; moderate‐certainty evidence; Analysis 1.4). The certainty of the evidence was downgraded by one step to moderate because of overall risk of bias (one study had an unclear risk of bias and two had a high risk of bias). Differences between the subgroups was detected (test for subgroup differences: P = 0.02) due to a larger effect of aminaftone .

Secondary outcomes

Assessment of CVI by objective signs: skin manifestations

Ulcer healing (dichotomous variable)

We included six trials in the meta‐analysis: one on aminaftone (Lazzarini 1982), one on calcium dobesilate (DOBESILATO500/2), two on diosmine (Fermoso 1992; Guilhou 1997) and two on rutoside (MacLennan 1994; Schultz‐Ehrenburg 1993), with a total of 230 participants in the active treatment group and 231 in the placebo group. Phlebotonics may make little or no difference to dichotomous variable ulcer cured compared to placebo (RR 0.94; 95% CI 0.79 to 1.13; 6 studies; 461 participants; low‐certainty evidence; Analysis 1.5). The certainty of the evidence was downgraded by two levels to low because of overall risk of bias (four studies had an unclear risk of bias and two had a high risk of bias) and imprecision (low number of total events) (summary of findings Table 1). No differences between the subgroups was detected (test for subgroup differences: P = 0.21).

Trophic disorders (dichotomous variable)

We included six studies in the meta‐analysis: four on hidrosmine and diosmine (Fermoso 1992; Gilly 1994; Laurent 1988; Planchon 1990), one on aminaftone (Lazzarini 1982) and one on rutosides (MacLennan 1994), with a total of 355 participants in the phlebotonics group and 350 in the placebo group. Phlebotonics probably slightly improve trophic disorders compared to placebo (RR 0.87, 95% CI 0.81 to 0.95; 6 studies; 705 participants; moderate‐certainty evidence; Analysis 1.6). The certainty of the evidence was downgraded by one level to moderate because of overall risk of bias (five studies had an unclear risk of bias and one had a high risk of bias). No differences between the subgroups was detected (test for subgroup differences: P = 0.80).

Telangiectasia, reticular veins and varicose veins (dichotomous variable)

Included studies did not report data on improvement in skin signs such as telangiectasia, reticular veins and varicose veins. Only Fermoso 1992 reported results regarding varicose veins. Before treatment, 3/16 (18.8%) participants presented varicose veins in the hidrosmine group and 2/12 participants in the placebo group (16.7%). After treatment, one participant from the hidrosmine group was cured of varicose veins, and no participants from the placebo group were cured.

Assessment of CVI by subjective symptoms

Pain in the lower legs (dichotomous variable)

A total of 21 studies reported on this outcome as a dichotomous variable: 10 on rutosides (Balmer 1980; Cauwenberge 1972; Cauwenberge 1978; Jongste 1989; Klüken 1971; Languillat 1988; Pedersen 1992; Pulvertaft 1983; Vanscheidt 2002a; Welch 1985), five on calcium dobesilate (Casley‐Smith 1988; Flota‐Cervera 2008; Hachen 1982; Rabe 2016; Widmer 1990), four on diosmine and hidrosmine (Biland 1982; Dominguez 1992; Fermoso 1992; Planchon 1990), one on aminaftone (Lazzarini 1982), and one on French maritime pine bark extract (Arcangeli 2000), with a total of 1468 participants treated with phlebotonics and 1130 with placebo (Analysis 1.7). The analysis showed heterogeneity (I² = 77%); therefore, we did not pool the data.

Pain in the lower legs (continuous variable)

We included 12 studies in the meta‐analysis: five on calcium dobesilate (DOBESILATO500/2; Marinello 2002; Martinez‐Zapata 2008; Rabe 2011; Rabe 2016), three on rutosides (Cloarec 1996; Cornu‐Thenard 1985; Parrado 1999), three on diosmine (Gilly 1994; Planchon 1990; Rabe 2015) and one on French maritime pine bark extract (Arcangeli 2000), with a total of 1110 participants assigned to phlebotonics and 1122 to placebo (Analysis 1.8). Phlebotonics may reduce pain (measured as a continuous variable) in the lower legs compared to placebo (SMD ‐0.35, 95% CI ‐0.54 to ‐0.17; 12 studies; 2232 participants; low‐certainty evidence). The certainty of the evidence was downgraded by two levels to low because of overall risk of bias (seven studies had an unclear risk of bias and three had a high risk of bias) and imprecision. We used a random‐effects model as heterogeneity was detected (I² = 75%). Differences between the subgroups was detected (test for subgroup differences: P = 0.002) due to differences in results between the subgroup of French maritime pine bark extract compared to the other subgroups.

Cramps in the lower legs (dichotomous variable)

We included 14 studies in the meta‐analysis: eight on rutosides (Balmer 1980; Cauwenberge 1978; Jongste 1989; Languillat 1988; Pedersen 1992; Pulvertaft 1983; Vin 1994; Welch 1985), three on diosmine and hidrosmine (Biland 1982; Fermoso 1992; Planchon 1990), two on calcium dobesilate (Casley‐Smith 1988; Widmer 1990) and one on aminaftone (Lazzarini 1982), with a total of 1072 participants treated with phlebotonics and 721 with placebo. Phlebotonics probably reduce cramps (measured as a dichotomous variable) compared to placebo (RR 0.72, 95% CI 0.58 to 0.89; 14 studies; 1793 participants; moderate‐certainty evidence; Analysis 1.9) The certainty of the evidence was downgraded by one level to moderate because of overall risk of bias (11 studies had an unclear risk of bias and three had a high risk of bias). We used a random‐effects model as heterogeneity was detected (I² = 73%). No differences between the subgroups was detected (test for subgroup differences: P = 0.28).

Cramps in the lower legs (continuous variable)

We included four studies in the meta‐analysis: two on rutosides (Cloarec 1996; Parrado 1999), one on calcium dobesilate (Martinez‐Zapata 2008), and one on diosmine (Gilly 1994), with 363 participants treated with phlebotonics and 366 with placebo (Analysis 1.10). The analysis showed heterogeneity (I² = 86%); therefore, we did not pool the data.

Restless legs (dichotomous variable)

We included seven studies in the meta‐analysis: four on rutosides (Balmer 1980; Cauwenberge 1978; Jongste 1989; Pedersen 1992), two on calcium dobesilate (Casley‐Smith 1988; Widmer 1990), and one on diosmine (Biland 1982). A total of 329 participants were treated with phlebotonics and 323 with placebo (Analysis 1.11). Phebotonics probably slightly reduce restless legs (measured as a dichotomous variable) compared to placebo (RR 0.81, 95% CI 0.72 to 0.91; 7 studies; 652 participants; moderate‐certainty evidence). The certainty of the evidence was downgraded by one level to moderate because of overall risk of bias (five studies had an unclear risk of bias and two had a high risk of bias). No differences between the subgroups was detected (test for subgroup differences: P = 0.41).

Itching in the lower legs (dichotomous variable)

We included four studies in the analysis: two on rutoside (Pedersen 1992; Vanscheidt 2002a), one on hidrosmine (Fermoso 1992), and one on aminaftone (Lazzarini 1982). A total of 206 participants were included in the active treatment group and 199 in the placebo group (Analysis 1.12). The analysis showed heterogeneity (I² = 92%); therefore, we did not pool the data.

Itching in the lower legs (continuous variable)

We included two studies in the analysis: one on calcium dobesilate (Martinez‐Zapata 2008), and one on rutosides (Parrado 1999). A total of 234 participants were treated with phlebotonics and 242 with placebo (Analysis 1.13). The analysis showed heterogeneity (I² = 82%), and we did not pool the data.

Feeling of heaviness in the lower legs (dichotomous variable)

We included 19 studies in the analysis: nine on rutosides (Cauwenberge 1972; Cauwenberge 1978; Jongste 1989; Languillat 1988; Pedersen 1992; Pulvertaft 1983; Vanscheidt 2002a; Vin 1994; Welch 1985), four on diosmine and hidrosmine (Dominguez 1992; Fermoso 1992; Planchon 1990; Tsouderos 1989), three on calcium dobesilate (Casley‐Smith 1988; Hachen 1982; Widmer 1990), one on aminaftone (Lazzarini 1982), one on Centella asiatica (Pointel 1986), and one on French maritime pine bark extract (Arcangeli 2000). A total of 1257 participants were included in the active treatment group and 909 in the placebo group (Analysis 1.14). The analysis showed heterogeneity (I² = 80%), and we did not pool the data.

Feeling of heaviness in the lower legs (continuous variable)

We included 10 studies in the analysis: six on rutosides (Alterkamper 1987; Cloarec 1996; Cornu‐Thenard 1985; Diebschlag 1994; Parrado 1999; Unkauf 1996), two on calcium dobesilate (Marinello 2002; Martinez‐Zapata 2008), one on diosmine (Gilly 1994), and one on French maritime pine bark extract (Arcangeli 2000). A total of 557 participants were included in the active treatment group and 557 in the placebo group (Analysis 1.15). The analysis showed heterogeneity (I² = 91%); therefore, we did not pool the data.

Swelling in the lower legs (dichotomous variable)

We included 14 studies in the analysis: nine on rutosides (Balmer 1980; Cauwenberge 1978; Jongste 1989; Kriner 1985; Languillat 1988; Pedersen 1992; Vanscheidt 2002a; Vin 1994; Welch 1985), two on calcium dobesilate (Casley‐Smith 1988; Hachen 1982), two on diosmine and hidrosmine (Biland 1982; Fermoso 1992), and one on French maritime pine bark extract (Arcangeli 2000), with 544 participants included in the active treatment group and 528 in the placebo group. Phebotonics probably reduce swelling in the lower leg (measured as a dichotomous variable) compared to placebo (RR 0.63, 95% CI 0.50 to 0.80; 14 studies; 1072 participants; moderate‐certainty evidence; Analysis 1.16). The certainty of the evidence was downgraded by one level to moderate because of overall risk of bias (11 studies had an unclear risk of bias and two had a high risk of bias). We used a random‐effects model as heterogeneity was detected (I² = 69%). Differences between the subgroups was detected (test for subgroup differences: P = 0.007) due to a larger effect of calcium dobesilate.

Swelling in the lower legs (continuous variable)

We included six studies in the analysis: three on rutosides (Cloarec 1996; Diebschlag 1994; Unkauf 1996), one on diosmine (Gilly 1994), one on calcium dobesilate (Martinez‐Zapata 2008), and one on French maritime pine bark extract (Arcangeli 2000), with 436 participants assigned to active treatment and 435 to placebo (Analysis 1.17). The analysis showed heterogeneity (I² = 95%), and we did not pool the data.

Paraesthesia in the lower legs (dichotomous variable)

We included nine studies in the analysis: four on rutosides (Balmer 1980; Cauwenberge 1978; Pulvertaft 1983; Welch 1985), three on calcium dobesilate (Casley‐Smith 1988; Hachen 1982; Widmer 1990) and two on diosmine and hidrosmine (Fermoso 1992; Planchon 1990), with 896 participants assigned to active treatment and 560 to placebo (Analysis 1.18). Phlebotonics probably reduce paraesthesia in the lower legs (measured as a dichotomous variable) compared to placebo (RR 0.67, 95% CI 0.50 to 0.88; 9 studies; 1456 participants; moderate‐certainty evidence). The certainty of the evidence was downgraded by one level) to moderate because of overall risk of bias (eight studies had an unclear risk of bias and one had a high risk of bias). We used a random‐effects model as heterogeneity was detected (I² = 72%). No differences between the subgroups was detected (test for subgroup differences: P = 0.32).

Paraesthesia in the lower legs (continuous variable)

We included two studies in the analysis: one on diosmine (Gilly 1994), and one on rutoside (Cornu‐Thenard 1985), with 97 participants assigned to active treatment and 91 to placebo (Analysis 1.19). It is uncertain whether phlebotonics reduce continuous variable paraesthesia because the certainty of this evidence is very low (SMD ‐0.15, 95% CI ‐0.44 to 0.13; 2 studies; 188 participants). The certainty of the evidence was downgraded by three levels to very low because of risk of bias (one level) and the sample size was small with a high imprecision in the results (two levels). No differences between the subgroups was detected (test for subgroup differences: P = 0.59).

Participant satisfaction (dichotomous variable)

We included 16 studies in the analysis: eight on rutosides (Burnand 1989; Cloarec 1996; Jongste 1989; Languillat 1988; Parrado 1999; Pedersen 1992; Pulvertaft 1983; Welch 1985), three on calcium dobesilate (Casley‐Smith 1988; Labs 2004; Rabe 2011), four on diosmine (Biland 1982; Chassignolle 1994; Danielsson 2002; Laurent 1988), and one on Centella asiatica (Allegra 1981), with a total of 1265 participants treated with phlebotonics and 939 with placebo (Analysis 1.20). The analysis showed heterogeneity (I² = 86%), and we did not pool the data.

Participant satisfaction (continuous variable)

We included seven studies in the analysis: four on rutosides (Cesarone 2002; Cloarec 1996; Ihme 1996; Kiesewetter 1997), two on calcium dobesilate (Rabe 2011; Widmer 1990), and one on diosmine (Gilly 1994), with 440 participants treated with phlebotonics and 441 with placebo (Analysis 1.21). The analysis showed heterogeneity (I² = 85%), and we did not pool the data.

Adverse events

Thirty‐seven studies reported on adverse events. These included 17 trials considering rutosides (Alterkamper 1987; Balmer 1980; Diebschlag 1994; Jongste 1989; Koscielnny 1996; Kriner 1985; Languillat 1988; MacLennan 1994; NCT01848210; Parrado 1999; Serralde 1990; Unkauf 1996; Vanscheidt 2002a; Vanscheidt 2002b; Vin 1994; Welch 1985; Zucarelli 1987), nine on hidrosmine‐diosmine (Biland 1982; Danielsson 2002; Dominguez 1992; Fermoso 1992; Gilly 1994; Guilhou 1997; Laurent 1988; Planchon 1990; Rabe 2015), eight on calcium dobesilate (Flota‐Cervera 2008; Hachen 1982; Labs 2004; Marinello 2002; Martinez‐Zapata 2008; Rabe 2011; Rabe 2016; Widmer 1990), one on aminaftone (Belczak 2014), one on grape seed extract (Thebaut 1985), and one on Centella asiatica (Pointel 1986).

We included in the meta‐analysis a total of 2944 participants treated with phlebotonics and 2845 with placebo. Phlebotonics probably increase adverse events slightly, compared to placebo (RR 1.14, 95% CI 1.02 to 1.27; 37 studies; 5789 participants; moderate‐certainty evidence; Analysis 1.22). The certainty of the evidence was downgraded by one level to moderate because of overall risk of bias (28 RCTs had unclear risk of bias and four RCTs had high risk of bias) (summary of findings Table 1). No differences between the subgroups was detected (test for subgroup differences: P = 0.36).

Adverse events analyzed by active agent

Aminaftone

Only one trial reported adverse events (Belczak 2014). One participant presented with headache in the group given Aminaftone, and two in the placebo group dropped out as the result of subjective worsening of leg pain. It is uncertain whether aminaftone reduces adverse events because the certainty of this evidence is very low (RR 0.60, 95% CI 0.06 to 6.32; 79 participants; Analysis 1.22). The certainty of the evidence was downgraded by three levels to very low because Belczak 2014 had unclear risk of bias (one level) and the sample size was very small with a high imprecision in the results (two levels).

Calcium dobesilate

In total, eight trials evaluated adverse events with calcium dobesilate use (Flota‐Cervera 2008; Hachen 1982; Labs 2004; Marinello 2002; Martinez‐Zapata 2008; Rabe 2011; Rabe 2016; Widmer 1990). Nineteen per cent of participants in the calcium dobesilate group (179/932) experienced an adverse event and 15% (133/892) in the placebo group. Calcium dobesilate may make little or no difference to adverse events compared with placebo (RR 1.22, 95% CI 1.00 to 1.49; 8 studies 1824 participants; low‐certainty evidence). The certainty of the evidence was downgraded by 2 levels to low because of overall risk of bias (four RCTs had unclear risk of bias and one RCT had high risk of bias). The most common adverse event was a gastrointestinal event (epigastric discomfort, vomiting). No agranulocytosis or white blood cell disorders were identified. Twenty‐five participants were withdrawn from the calcium dobesilate group and 17 from the placebo group as the result of adverse events.

Centella asiatica

One study reported information on adverse events with Centalla asiatica (Pointel 1986). Thirty‐one per cent of participants in the Centella asiatica group (19/61) suffered from adverse events and 27.3% (9/33) in the placebo group. It is uncertain whether Centella asiatica reduces adverse events because the certainty of this evidence is very‐low (RR 1.14, 95% CI 0.58 to 2.23; 94 participants). The certainty of the evidence was downgraded by three levels to very low because Pointel 1986 had unclear risk of bias (one level) and the sample size was very small with a high imprecision in the results (two levels). Two participants who took Centella asiatica 120 mg withdrew ‐ one because of gastralgia (gastric colic) and the other because of neurological absence (absence of nerve activity). One participant taking placebo discontinued the study because of cyanosis of the extremities (bluish discolouration caused by lack of oxygen in the blood).

Diosmine and hidrosmine

Nine studies reported the number of participants who experienced adverse events (Biland 1982; Danielsson 2002; Dominguez 1992; Fermoso 1992; Gilly 1994; Guilhou 1997; Laurent 1988; Planchon 1990; Rabe 2015). Ninety‐nine adverse events occurred in the hidrosmine and diosmine group (99/720) and 106 (106/709) in the placebo group. Diosmine and hidrosmine may make little or no difference to adverse events compared with placebo (RR 0.93, 95% CI 0.72 to 1.19; 9 studies; 1429 participants; low‐certainty evidence). The certainty of the evidence was downgraded by two levels to low because of overall risk of bias (eight RCTs had unclear risk of bias and one RCT had high risk of bias).

Gastrointestinal disorders were the most reported adverse events (heartburn and nausea): 14 cases were reported in the hidrosmine and diosmine group and 11 in the placebo group.

Thirteen participants withdrew from the hidrosmine group and 12 from the placebo group as the result of adverse events.

Grape seed extract

One study reported information regarding adverse events (Thebaut 1985). Eleven per cent of participants (4/35) receiving active treatment reported adverse effects (three withdrew): two participants had gastralgia, one participant had a headache and one had an allergic reaction. Twenty per cent of participants in the placebo group (8/40) experienced adverse effects (one withdrew); these included constipation, gastralgia, tiredness, dry mouth and discomfort. It is uncertain whether grape seed extract reduces adverse events because of the certainty of this evidence is very low (RR 0.57, 95% CI 0.19 to 1.74; 75 participants). The certainty of the evidence was downgraded by three levels to very low because of Thebaut 1985 had unclear risk of bias (one level) and the sample size was very small with a high imprecision in the results (two levels).

Rutoside

Sixteen trials reported information regarding the number of participants who experienced adverse events (Alterkamper 1987; Balmer 1980; Diebschlag 1994; Jongste 1989; Koscielnny 1996; Kriner 1985; Languillat 1988; MacLennan 1994; Parrado 1999; Serralde 1990; Unkauf 1996; Vanscheidt 2002a; Vanscheidt 2002b; Vin 1994; Welch 1985; Zucarelli 1987).Twenty per cent of participants (233/1160) in the rutoside group suffered from adverse events and 16% (181/1128) in the placebo group. Rutosids may slightly increase adverse events compared to placebo (RR 1.41, 95% CI 1.08 to 1.83; 16 studies; 2288 participants; low‐certainty evidence). The certainty of the evidence was downgraded by two levels to low because of the overall risk of bias (13 RCTs had unclear risk of bias and two had high risk of bias). The most common adverse events were gastrointestinal in nature (constipation, dry mouth, epigastric discomfort, vomiting): 127 in the rutoside group and 81 in the placebo group, followed by headache (23 in the rutoside group, 21 in the placebo group) and tiredness (17 in the rutoside group, nine in the placebo group).

Thirteen participants withdrew from the rutoside group and 22 from the placebo group as the result of adverse events.

Sensitivity analysis

Exclusion of studies using compression measures (elastic stockings)

We carried out sensitivity analysis by re‐analysing the data excluding studies that allowed the use of elastic stockings (Balmer 1980; DOBESILATO500/2; Guilhou 1997; Laurent 1988; MacLennan 1994; Martinez‐Zapata 2008; Rabe 2011; Schultz‐Ehrenburg 1993; Zucarelli 1987). We found that generally, results did not change, except for the following variables.

Phlebotonics may reduce dichotomous variable pain (RR 0.70, 95% CI 0.60 to 0.82; 18 studies; 1818 participants, low‐certainty evidence; Analysis 2.7). The certainty of evidence was downgraded by two levels to low because of overall risk of bias (18 studies had unclear risk of bias and five had high risk of bias). No differences between the subgroups was detected (test for subgroup differences: P = 0.12). In the overall analysis, the results were very heterogeneous, so we did not pool them.
Phlebotonics may reduce dichotomous variable participant satisfaction (RR 0.69, 95% CI 0.53 to 0.90; 12 studies; 1193 participants; low‐certainty evidence; Analysis 2.20). The certainty of evidence was downgraded by two levels to low because of overall risk of bias (10 studies had unclear risk of bias and one had high risk of bias). No differences between the subgroups was detected (test for subgroup differences: P = 0.33). In the overall analysis, the results were very heterogeneous, so we did not pool them.

Exclusion of unpublished data

Only one study, which focused on rutosides, was not published (Welch 1985). When we re‐analysed the data while excluding this study, we found results very similar to those of the main analysis for all outcomes.

Exclusion of studies at high or unclear risk of bias

In judging quality levels based on the aforementioned criteria, we identified only four studies with low risk of bias (Labs 2004; Martinez‐Zapata 2008; Rabe 2011; Vanscheidt 2002a). Consequently, limited sensitivity analyses for the included variables were possible.

Results changed only for the following variables:

For the dichotomous variable of oedema, only one study on calcium dobesilate was included with a low risk of bias (Labs 2004). Phlebotonics may make little or no difference in the dichotomous variable oedema compared to placebo (RR 0.99, 95% CI 0.63 to 1.55; 1 study; 260 participants; low‐certainty evidence; Analysis 4.1). The certainty of the evidence was downgraded by two levels to low because Labs 2004 had a small sample size and imprecision.
For the continuous variable of oedema (measure of ankle circumference in mm), three studies on calcium dobesilate were included with a low risk of bias (Labs 2004; Martinez‐Zapata 2008; Rabe 2011). Based on their results, phlebotonics probably make little or no difference in the continuous variable oedema (measure of ankle circumference in mm) compared to placebo (MD ‐2.34 mm, 95% CI ‐8.79 to 4.11; 3 studies; 867 participants; moderate‐certainty evidence; Analysis 4.2). The certainty of the evidence was downgraded by one level for imprecision.
For the dichotomous variable of itching, only one study on rutoside had a low risk of bias (Vanscheidt 2002a). Phlebotonics probably reduce dichotomous variable itching, compared with placebo (RR 0.44, 95% CI 0.32 to 0.62; 231 participants; moderate‐certainty evidence; Analysis 4.7). The certainty of the evidence was downgraded by one level for imprecision.
For the continuous variable of itching, one study on calcium dobesilate was at low risk of bias (Martinez‐Zapata 2008). Phlebotonics probably make little or no difference to the continuous variable itching (MD 4.60 cm, 95% CI ‐5.66 to 14.86; 416 participants; moderate‐certainty evidence; Analysis 4.8). The certainty of the evidence was downgraded by one level by imprecision.
For the dichotomous variable of heaviness, one study on rutoside was at low risk of bias (Vanscheidt 2002a). Phlebotonics probably reduce the dichotomous variable heaviness compared to placebo (RR 0.62, 95% CI 0.47 to 0.82; 231 participants; moderate‐certainty evidence; Analysis 4.9). The certainty of the evidence was downgraded by one level by imprecision.
For the continuous variable of heaviness, one study on calcium dobesilate was at low risk of bias (Martinez‐Zapata 2008). Phlebotonics probably make little or no difference on the continuous variable heaviness compared with placebo (MD ‐2.40 cm, 95% CI ‐7.89 to 3.09; 417 participants; moderate‐certainty evidence; Analysis 4.10). The certainty of the evidence was downgraded by one level for imprecision.
For the continuous variable of swelling, one study on calcium dobesilate was at low risk of bias (Martinez‐Zapata 2008). Phlebotonics probably make little or no difference on the continuous variable swelling, compared to placebo (MD ‐1.30 cm, 95% CI ‐6.72 to 4.12; 417 participants; moderate‐certainty evidence; Analysis 4.12). The certainty of the evidence was downgraded by one level for imprecision.
For the dichotomous variable of participant satisfaction, two studies on calcium dobesilate were at low risk of bias (Labs 2004; Rabe 2011). Phlebotonics probably make little or no difference on the dichotomous variable participant satisfaction compared to placebo (RR 1.04, 95% CI 0.81 to 1.32; 2 studies; 476 participants; moderate‐certainty evidence; Analysis 4.13). The certainty of the evidence was downgraded by one level for imprecision.
For the continuous variable of participant satisfaction, one study on calcium dobesilate had a low risk of bias (Rabe 2011). Phlebotonics probably improve the continuous variable participant satisfaction (MD ‐5.64, 95% CI ‐8.85 to ‐2.43; 223 participants; moderate‐ certainty evidence; Analysis 4.14). The certainty of the evidence was downgraded by one level for imprecision.
For the dichotomous variable of adverse events, four studies were at low risk of bias (Labs 2004; Martinez‐Zapata 2008; Rabe 2011; Vanscheidt 2002a). Phlebotonics probably make little or no difference on the dichotomous variable adverse events compared to placebo (RR 1.59, 95% CI 0.97 to 2.63; four studies; 1257 participants; moderate‐certainty evidence; Analysis 4.15). The certainty of the evidence was downgraded by one level for imprecision (low number of events). No differences between the subgroups was detected (test for subgroup differences: P = 0.70)

Discussion

Summary of main results

We evaluated the efficacy and safety of phlebotonics in the treatment of CVI. Only analyses of studies using oral phlebotonics were possible because no identified study of topical phlebotonics met the inclusion criteria of this Cochrane Review. This Cochrane Review included 69 RCTs, and analyzed data from 56 trials involving 7690 participants. Studies included in the review generally provided objective measurement of ankle and calf oedema reduction, as well as a subjective assessment of other signs and symptoms of CVI.

According to the ITT analysis, there was moderate‐certainty evidence of a probable beneficial effect on the dichotomous variable oedema. Analyses with continuous variables also showed a probable beneficial effect of phlebotonics on oedema (moderate‐certainty evidence).

However, there was moderate‐certainty evidence of little or no difference in QoL with phlebotonic use compared to placebo; and little or no difference to dichotomous variable ulcer healing (low‐certainty evidence).

Phlebotonic use probably slightly improves trophic disorders compared to placebo (moderate‐certainty evidence). Data on telangiectasia, reticular veins and varicose veins were very limited. Heterogeneity prevented meta‐analysis on dichotomous variable pain but showed phlebotonic use may reduce pain (continuous variable) compared to placebo (low‐certainty evidence). Similarly, a possible benefit was seen in cramps (moderate‐certainty evidence), restless legs (moderate‐certainty evidence); swelling in lower legs (moderate‐certainty evidence); and paraesthesia (dichotomous variable) in the lower legs (moderate‐certainty evidence). However, based on very low‐certainty evidence, it is uncertain whether phlebotonics reduce continuous variable paraesthesia.

Heterogeneity prevented meta‐analysis on the outcomes of itching and feeling of heaviness in the lower legs.

There was moderate‐certainty evidence that the incidence of adverse events was probably slightly increased in the phlebotonics group compared to placebo group. Gastrointestinal disorders were the most frequently reported adverse events among studies that provided this information (rutosides, calcium dobesilate, diosmine‐hidrosmine). Our review did not report agranulocytosis associated with calcium dobesilate, although this adverse effect was described in a previous case‐control study that detected potential risk of agranulocytosis, with an incidence rate of 1.21 cases per 10,000 patient‐years of treatment (Ibañez 2000; Ibáñez 2005). This could be explained by the small number of participants in the included RCTs and the short period of participant follow‐up provided.

The results by type of active drug showed that It is uncertain whether aminaftone reduces dichotomous variables oedema, pain, cramps, itching and heaviness (very low‐certainty evidence). There was low‐certainty evidence that aminaftone may slightly improve the continuous variables oedema (volume) and QoL, and may make little or no difference to adverse events.

Calcium dobesilate may reduce continuous volume of the leg and may slightly improve continuous variable participant satisfaction (low‐certainty evidence); it may reduce dichotomous variable swelling and may slightly reduce the dichotomous variables pain, cramps and restless legs. Meanwhile, based on moderate‐certainty evidence, calcium dobesilate probably makes little or no difference on QoL and dichotomous variables assessment by the participant and adverse events. Furthermore, calcium dobesilate may make little or no differences on the continuous variables ankle perimeter circumference, pain and heaviness; and for the dichotomous variables heaviness and paraesthesia (low‐certainty evidence). Calcium dobesilate does not have an important effect on the continuous variables swelling, cramps and itching (high‐certainty evidence). We do not know if calcium dobesilate improves dichotomous variable oedema and ulcer healing as the certainty of the evidence is very low.

Based on very low‐certainty evidence, it is uncertain whether Centella asiatica compared to placebo reduces dichotomous variable heaviness (Pointel 1986), improves the dichotomous variable participant satisfaction (Allegra 1981) or increases adverse events (Pointel 1986).

Diosmine and hidrosmine, based on a moderate‐certainty evidence, probably slightly reduces ankle perimeter, trophic disorders and cramps; and, probably make little or no differences in dichotomous variable pain, QoL, dichotomous variable participant satisfaction and adverse events, Additionally, based on low‐certainty evidence, they may slightly reduce oedema, cramps, heaviness, swelling and dichotomous variable participant satisfaction. Furthermore, diosmine and hidrosmine, based on low‐certainty evidence, may make little or no difference in paraesthesia, ulcer healing, continuous variable pain and dichotomous variable heaviness. We do not know if these drugs improve itching because the certainty of the evidence is very low.

Based on very low‐certainty evidence, it is uncertain whether French maritime pine bark extract reduces pain, heaviness and swelling. For grape seed extract it is also uncertain if it reduces the dichotomous variable oedema (very low‐certainty evidence).

Rutosides were included in the greatest number of clinical trials. Based on moderate‐certainty evidence, rutosides probably improve oedema, volume of the leg and continuous variable pain; and they probably make little or no difference in ankle perimeter and ulcer healing. Based on low‐certainty evidence, rutosides may slightly reduce heaviness, participant satisfaction, continuous variables cramps and itching, dichotomous variables pain, and paraesthesia. Furthermore, rutosides may make little or no difference in continuous variables swelling and paraesthesia and dichotomous variables trophic disorders, cramps, restless and itching. Rutosides may slightly increase adverse events (low‐certainty evidence).

No evidence was found regarding the efficacy of disodium flavodate, naftazone, chromocarbe or topical phlebotonics.

Sensitivity analyses did not alter the results of this review. Whether elastic stockings were used did not influence pooled results, supporting the view that an appropriate randomisation method results in a homogeneous distribution of the groups under comparison.

Overall completeness and applicability of evidence

Several limitations were identified in the included studies. Only one of five studies specified standard diagnostic criteria for CVI, and different studies applied different criteria. Only eleven studies used the currently accepted CEAP classification (Porter 1995) (Belczak 2014; Danielsson 2002; DOBESILATO500/2; Labs 2004; Marinello 2002; Martinez‐Zapata 2008; Rabe 2011; Rabe 2015; Rabe 2016; Vanscheidt 2002a; Vanscheidt 2002b). Therefore, homogeneity in diagnostic criteria is limited, and potential misclassification bias cannot be ruled out. Furthermore, we were unable to perform subgroup analysis by CVI stage because the severity of CVI was variable.

In most RCTs, the way in which participants were included is heterogeneous, and this may have led to differences in response to treatment. In addition, too few participants were included in the studies with the limitations of imprecision in the results and lack of statistical power to detect a difference between phlebotonics and placebo, when an effect could have occurred (beta error, or type II error). Different instruments were used to measure signs and symptoms, and sometimes results were inconclusive. Only seven RCTs assessed the variable QoL using a standardized questionnaire (Belczak 2014; Martinez‐Zapata 2008; Rabe 2011; Rabe 2015; Rabe 2016;Vanscheidt 2002a; Vanscheidt 2002b), but two studies did not provide quantifiable information (Vanscheidt 2002a; Vanscheidt 2002b). Although some studies favoured phlebotonics, the clinical relevance of these findings remains questionable.

Although infrequent, important signs such as venous ulcers have been poorly evaluated. Only six studies included participants with venous ulcers and when pooled, showed none that yielded a difference in ulcer healing (DOBESILATO500/2; Fermoso 1992; Guilhou 1997; Lazzarini 1982; MacLennan 1994; Schultz‐Ehrenburg 1993).

Only two studies addressing trophic disorders defined this term (MacLennan 1994; Planchon 1990), and four did not (Fermoso 1992; Gilly 1994; Laurent 1988; Lazzarini 1982). However, in two studies, trophic disorders were assessed subjectively as present or absent (Fermoso 1992; MacLennan 1994), or as reported on semi‐quantitative four‐item scales (Gilly 1994; Lazzarini 1982; Planchon 1990). Therefore, although data from the examination of trophic alterations were analyzed, these results should be interpreted with caution.

Most studies provided short‐term results (one to three months). Specificaly, for the primary outcome 'oedema in the lower limb,' the median time of follow‐up was 49 days for oedema measured as a dichotomous outcome and 60 days for oedema measures as a continuous outcome. Given the chronic nature of the disease, more long‐term data on the efficacy and safety of phlebotonics are needed (at least one‐year follow‐up). To achieve homogeneous data collection and to specify evidence on the efficacy of phlebotonics, measurement of signs and symptoms should be standardized. Although we have done a subgroup analysis by drugs, we noted that different doses were involved, and we are unable to comment on which is the optimal dose.

Quality of the evidence

Risk of bias of the included studies is somewhat unclear regarding randomisation and blinding because only a limited number of studies specifically reported details regarding these issues. It is difficult to determine whether this is a result of poor design or publication restrictions. As a result, among the 69 RCTs included in this review, only 39 explained the double‐blinding procedure in detail, 18 provided data on randomisation and 10 explained blinding of the randomisation. Furthermore, seven studies had attrition bias and nine selective reporting with high risk of bias. These issues were not addressed in the remaining included studies, and this adds uncertainty to the evidence. Only four studies were graded as having an overall low risk of bias (Labs 2004; Martinez‐Zapata 2008; Rabe 2011; Vanscheidt 2002a).

In the clinical area of CVI, results lack reliability if the RCT did not include a placebo group because of seasonal exacerbations (spring and summer) that might be self limiting and highly subjective symptoms. Consequently, an adequate control group is needed, and both randomisation and treatment should be appropriately blinded (preferably double‐blinded). For this reason, studies that did not include a control group and single‐blinded studies were excluded from the review. Among studies identified as double‐blinded, those with inappropriate blinding of treatments or randomisation were excluded from the meta‐analyses.

We adopted a conservative approach in our review, which prioritised the ITT analysis in terms of both treatment losses and failures. On the other hand, we used change measures only if conditions of the compared groups at baseline were the same, to avoid bias in the assessment of results related to participants' baseline differences.

We evaluated the certainty of the body of evidence using the GRADE approach (Schünemann 2011), which is based on five considerations including risk of bias (study limitations), directness of the evidence, heterogeneity in the data, precision of effect estimates and additional considerations (including risk of publication bias) to assess the certainty of the body of evidence for a priori selected outcomes (in our review, these included the dichotomous variable of oedema in the lower legs and the continuous variables of oedema in the lower legs, QoL, participants with ulcer healing and participants with adverse events) (summary of findings Table 1).

With this approach, the overall certainty of evidence is ranked from very low (paraesthesia (continuous variable) to moderate (dichotomous and continuous outcomes of oedema and adverse events; cramps, restless legs, swelling and paraesthesia (dichotomous variable).

Reasons for downgrading the certainty of evidence for the outcome ulcer healing include the presence of selective reporting and incomplete outcome data; for the outcome QoL, we downgraded for incomplete outcome data and imprecision (wide confidence intervals); for the dichotomous variables measurement of oedema we downgraded for incomplete outcome data; and for the continuous variable oedema we downgraded for unclear risk of bias of one trial; for the outcome adverse events we downgraded for incomplete outcome data and indirectness (moderate heterogeneity). See summary of findings Table 1.

Potential biases in the review process

Any systematic review is influenced by the quality of included studies and reports. In this respect, we classified only four RCTs as having low risk of bias, and we considered most included studies to have moderate risk of potential bias. We excluded RCTs with high risk of bias. Therefore, conclusions about the results of these studies should be interpreted with caution.

The heterogeneity of several analysis variables may be due to the following:

Different diagnosis classification criteria have been applied; therefore, characteristics of the included population in terms of degree of progression of CVI might vary among studies.
No standardisation is involved in measuring variables, given the different scales that have been used, some of which are not validated. Although the same criteria were applied to the data dichotomisation (participants without symptoms/signs or with mild symptoms/sign versus participants with moderate to severe symptoms/signs), these may not be equally relevant, as they result from the application of different scales.
On the other hand, the same subjectivity of collected variables may represent differences among individuals and may influence the variability of results.
In addition, efficacy of evaluated treatments may not be the same because different active principles were used. This explains observed differences among treatments in the subgroup analysis.

All these considerations limit the validity of included clinical trials and the conclusions of this review. The existence of such heterogeneity restricts the importance of its detection in the process of generating hypotheses (i.e. phlebotonics could be effective for treatment of the pain, cramps, heaviness and swelling of CVI).

Only 54% of included studies reported information on adverse events. However, to adequately assess adverse events related to phlebotonics, it is necessary to include observational study designs that were excluded from our review.

Agreements and disagreements with other studies or reviews

Several reviews have tried to evaluate the clinical benefit of phlebotonics. Some of these used poor methods, which did not include information on search strategies and data collection sources, extraction and statistical treatment (diosmine, escin and rutosides (Diehm 1996b); flavonoids, tribenosides, escin and calcium dobesilate (Markwardt 1996); rutosides (Wadworth 1992); flavonoids (Rabe 2013)). Other reviews are more elaborate and were developed systematically (global phlebotonics (Boada 1999); calcium dobesilate (Ciapponi 2004); escin (Pittler 1998); rutosides (Aziz 2015; Poynard 1994); flavonoids (Kakkos 2018)). Five reviews pursued data meta‐analysis (Aziz 2015; Boada 1999; Ciapponi 2004; Kakkos 2018; Poynard 1994).

One review specifically evaluated hydroxyethyl rutosides and the review authors included 15 randomised studies and applied a per‐protocol (PP) analysis. They stated that rutosides were better than control for controlling symptoms of pain, cramps and heaviness (Aziz 2015).

Another review analyzed rutosides and the review authors included 12 randomised, double‐blind, placebo‐controlled studies and applied an ITT analysis. They stated that rutosides were better than placebo for controlling symptoms of pain, cramps, heaviness, swelling and tiredness of affected legs. They did not mention CVI signs (Poynard 1994).

Boada 1999 reviewed all drugs that have been evaluated for CVI through randomised, double‐blind, placebo‐controlled trials without concomitant compression procedures. These included traditional agents such as hidrosmine, diosmine, escin, rutosides and calcium dobesilate, along with other, less usual ones such as extract of Centella asiatica, benzarone, tribenoside, flunarizine, dihydroergotamine mesylate and mucopolysaccharide sulphate. The conclusion of the Boada 1999 review was that phlebotonics might improve leg heaviness in patients with CVI. Review authors presented no conclusive data regarding other signs or symptoms, performed PP rather than ITT analysis and provided no information on individual phlebotonics (Boada 1999).

The review led by Ciapponi 2004 analyzed calcium dobesilate and the review authors included 10 double‐blind, randomised, placebo‐controlled studies and applied a PP analysis. They stated that calcium dobesilate was better than placebo for controlling cramps and discomfort. Subgroup analysis showed greater efficacy in more severe cases of the disease in terms of improving symptoms (pain, heaviness and swelling) and signs (leg volume). Sensitivity analysis based on the ITT analysis did not influence these results (Ciapponi 2004).

Kakkos 2018 evaluated the efficacy of a micronized purified flavonoid fraction (diosmine) in CVD. The systematic review included seven double‐blind, randomised, placebo‐controlled studies. Their results showed a significant improvement for diosmine with respect to signs and symptoms related to CVD, QoL and treatment assessment by the physician. Although our review presents some differences because we pooled studies assessing diosmine and hidrosmine, except for QoL, the general results are in agreement with Kakkos 2018.

With the exception of Aziz 2015 and Kakkos 2018, the above‐cited reviews were published some time ago and have not been updated. Our review provides an update regarding evidence on phlebotonics in general and by drug group.

Figure 1

Funnel plot of comparison: 1 Phlebotonics vs placebo, outcome: 1.1 Oedema in the lower legs (dichotomous variable).

Figure 2

Study flow diagram.

Figure 3

Methodological quality graph: review authors' judgements about each methodological quality item presented as percentages across all included studies.

Figure 4

Methodological quality summary: review authors' judgements about each methodological quality item for each included study.

Analysis 1.1

Comparison 1: Phlebotonics versus placebo, Outcome 1: Oedema in the lower legs (dichotomous variable)

Analysis 1.2

Comparison 1: Phlebotonics versus placebo, Outcome 2: Ankle perimeter circumference (mm)

Analysis 1.3

Comparison 1: Phlebotonics versus placebo, Outcome 3: Volume of the leg (mL)

Analysis 1.4

Comparison 1: Phlebotonics versus placebo, Outcome 4: Quality of life

Analysis 1.5

Comparison 1: Phlebotonics versus placebo, Outcome 5: Ulcer healing

Analysis 1.6

Comparison 1: Phlebotonics versus placebo, Outcome 6: Trophic disorders (dichotomous variable)

Analysis 1.7

Comparison 1: Phlebotonics versus placebo, Outcome 7: Pain in the lower legs (dichotomous variable)

Analysis 1.8

Comparison 1: Phlebotonics versus placebo, Outcome 8: Pain in the lower legs (continuous variable)

Analysis 1.9

Comparison 1: Phlebotonics versus placebo, Outcome 9: Cramps in the lower legs (dichotomous variable)

Analysis 1.10

Comparison 1: Phlebotonics versus placebo, Outcome 10: Cramps in the lower legs (continuous variable)

Analysis 1.11

Comparison 1: Phlebotonics versus placebo, Outcome 11: Restless legs (dichotomous variable)

Analysis 1.12

Comparison 1: Phlebotonics versus placebo, Outcome 12: Itching in the lower legs (dichotomous variable)

Analysis 1.13

Comparison 1: Phlebotonics versus placebo, Outcome 13: Itching in the lower legs (continuous variable)

Analysis 1.14

Comparison 1: Phlebotonics versus placebo, Outcome 14: Heaviness in the lower legs (dichotomous variable)

Analysis 1.15

Comparison 1: Phlebotonics versus placebo, Outcome 15: Heaviness in the lower legs (continuous variable)

Analysis 1.16

Comparison 1: Phlebotonics versus placebo, Outcome 16: Swelling in the lower legs (dichotomous variable)

Analysis 1.17

Comparison 1: Phlebotonics versus placebo, Outcome 17: Swelling in the lower legs (continuous variable)

Analysis 1.18

Comparison 1: Phlebotonics versus placebo, Outcome 18: Paraesthesia in the lower legs (dichotomous variable)

Analysis 1.19

Comparison 1: Phlebotonics versus placebo, Outcome 19: Paraesthesia in the lower legs (continuous variable)

Analysis 1.20

Comparison 1: Phlebotonics versus placebo, Outcome 20: Participant satisfaction (dichotomous variable)

Analysis 1.21

Comparison 1: Phlebotonics versus placebo, Outcome 21: Participant satisfaction (continuous variable)

Analysis 1.22

Comparison 1: Phlebotonics versus placebo, Outcome 22: Adverse events

Analysis 2.1

Comparison 2: Sensitivity analysis excluding studies that allowed the use of elastic stockings, Outcome 1: Oedema in the lower legs (dichotomous variable)

Analysis 2.2

Comparison 2: Sensitivity analysis excluding studies that allowed the use of elastic stockings, Outcome 2: Ankle perimeter circumference (mm)

Analysis 2.3

Comparison 2: Sensitivity analysis excluding studies that allowed the use of elastic stockings, Outcome 3: Volume of the leg (mL)

Analysis 2.4

Comparison 2: Sensitivity analysis excluding studies that allowed the use of elastic stockings, Outcome 4: Quality of life

Analysis 2.5

Comparison 2: Sensitivity analysis excluding studies that allowed the use of elastic stockings, Outcome 5: Ulcer healing

Analysis 2.6

Comparison 2: Sensitivity analysis excluding studies that allowed the use of elastic stockings, Outcome 6: Trophic disorders (dichotomous variable)

Analysis 2.7

Comparison 2: Sensitivity analysis excluding studies that allowed the use of elastic stockings, Outcome 7: Pain in the lower legs (dichotomous variable)

Analysis 2.8

Comparison 2: Sensitivity analysis excluding studies that allowed the use of elastic stockings, Outcome 8: Pain in the lower legs (continuous variable)

Analysis 2.9

Comparison 2: Sensitivity analysis excluding studies that allowed the use of elastic stockings, Outcome 9: Cramps in the lower legs (dichotomous variable)

Analysis 2.10

Comparison 2: Sensitivity analysis excluding studies that allowed the use of elastic stockings, Outcome 10: Cramps in the lower legs (continuous variable)

Analysis 2.11

Comparison 2: Sensitivity analysis excluding studies that allowed the use of elastic stockings, Outcome 11: Restless legs (dichotomous variable)

Analysis 2.12

Comparison 2: Sensitivity analysis excluding studies that allowed the use of elastic stockings, Outcome 12: Itching in the lower legs (dichotomous variable)

Analysis 2.13

Comparison 2: Sensitivity analysis excluding studies that allowed the use of elastic stockings, Outcome 13: Itching in the lower legs (continuous variable)

Analysis 2.14

Comparison 2: Sensitivity analysis excluding studies that allowed the use of elastic stockings, Outcome 14: Heaviness in the lower legs (dichotomous variable)

Analysis 2.15

Comparison 2: Sensitivity analysis excluding studies that allowed the use of elastic stockings, Outcome 15: Heaviness in the lower legs (continuous variable)

Analysis 2.16

Comparison 2: Sensitivity analysis excluding studies that allowed the use of elastic stockings, Outcome 16: Swelling in the lower legs (dichotomous variable)

Analysis 2.17

Comparison 2: Sensitivity analysis excluding studies that allowed the use of elastic stockings, Outcome 17: Swelling in the lower legs (continuous variable)

Analysis 2.18

Comparison 2: Sensitivity analysis excluding studies that allowed the use of elastic stockings, Outcome 18: Paraesthesias in the lower legs (dichotomous variable)

Analysis 2.19

Comparison 2: Sensitivity analysis excluding studies that allowed the use of elastic stockings, Outcome 19: Paraesthesias in the lower legs (continuous variable)

Analysis 2.20

Comparison 2: Sensitivity analysis excluding studies that allowed the use of elastic stockings, Outcome 20: Participant satisfaction (dichotomous variable)

Analysis 2.21

Comparison 2: Sensitivity analysis excluding studies that allowed the use of elastic stockings, Outcome 21: Participant satisfaction (continuous variable)

Analysis 2.22

Comparison 2: Sensitivity analysis excluding studies that allowed the use of elastic stockings, Outcome 22: Adverse events

Analysis 3.1

Comparison 3: Sensitivity analysis of published studies only, Outcome 1: Oedema in the lower legs (dichotomous variable)

Analysis 3.2

Comparison 3: Sensitivity analysis of published studies only, Outcome 2: Ankle perimeter circumference (mm)

Analysis 3.3

Comparison 3: Sensitivity analysis of published studies only, Outcome 3: Volume of the leg (mL)

Analysis 3.4

Comparison 3: Sensitivity analysis of published studies only, Outcome 4: Quality of life

Analysis 3.5

Comparison 3: Sensitivity analysis of published studies only, Outcome 5: Patients with ulcer (dichotomous variable)

Analysis 3.6

Comparison 3: Sensitivity analysis of published studies only, Outcome 6: Trophic disorders (dichotomous variable)

Analysis 3.7

Comparison 3: Sensitivity analysis of published studies only, Outcome 7: Pain in the lower legs (dichotomous variable)

Analysis 3.8

Comparison 3: Sensitivity analysis of published studies only, Outcome 8: Pain in the lower legs (continuous variable)

Analysis 3.9

Comparison 3: Sensitivity analysis of published studies only, Outcome 9: Cramps in the lower legs (dichotomous variable)

Analysis 3.10

Comparison 3: Sensitivity analysis of published studies only, Outcome 10: Cramps in the lower legs (continuous variable)

Analysis 3.11

Comparison 3: Sensitivity analysis of published studies only, Outcome 11: Restless legs (dichotomous variable)

Analysis 3.12

Comparison 3: Sensitivity analysis of published studies only, Outcome 12: Itching in the lower legs (dichotomous variable)

Analysis 3.13

Comparison 3: Sensitivity analysis of published studies only, Outcome 13: Itching in the lower legs (continuous variable)

Analysis 3.14

Comparison 3: Sensitivity analysis of published studies only, Outcome 14: Heaviness in the lower legs (dichotomous variable)

Analysis 3.15

Comparison 3: Sensitivity analysis of published studies only, Outcome 15: Heaviness in the lower legs (continuous variable)

Analysis 3.16

Comparison 3: Sensitivity analysis of published studies only, Outcome 16: Swelling in the lower legs (dichotomous variable)

Analysis 3.17

Comparison 3: Sensitivity analysis of published studies only, Outcome 17: Swelling in the lower legs (continuous variable)

Analysis 3.18

Comparison 3: Sensitivity analysis of published studies only, Outcome 18: Paraesthesias in the lower legs (dichotomous variable)

Analysis 3.19

Comparison 3: Sensitivity analysis of published studies only, Outcome 19: Paraesthesias in the lower legs (continuous variable)

Analysis 3.20

Comparison 3: Sensitivity analysis of published studies only, Outcome 20: Participant satisfaction (dichotomous variable)

Analysis 3.21

Comparison 3: Sensitivity analysis of published studies only, Outcome 21: Participant satisfaction (continuous variable)

Analysis 3.22

Comparison 3: Sensitivity analysis of published studies only, Outcome 22: Adverse events

Analysis 4.1

Comparison 4: Sensitivity analysis based on low risk of bias, Outcome 1: Oedema in the lower legs (dichotomous variable)

Analysis 4.2

Comparison 4: Sensitivity analysis based on low risk of bias, Outcome 2: Ankle perimeter circumference (mm)

Analysis 4.3

Comparison 4: Sensitivity analysis based on low risk of bias, Outcome 3: Quality of life

Analysis 4.4

Comparison 4: Sensitivity analysis based on low risk of bias, Outcome 4: Pain in the lower legs (dichotomous variable)

Analysis 4.5

Comparison 4: Sensitivity analysis based on low risk of bias, Outcome 5: Pain in the lower legs (continuous variable)

Analysis 4.6

Comparison 4: Sensitivity analysis based on low risk of bias, Outcome 6: Cramps in the lower legs (continuous variable)

Analysis 4.7

Comparison 4: Sensitivity analysis based on low risk of bias, Outcome 7: Itching in the lower legs (dichotomous variable)

Analysis 4.8

Comparison 4: Sensitivity analysis based on low risk of bias, Outcome 8: Itching in the lower legs (continuous variable)

Analysis 4.9

Comparison 4: Sensitivity analysis based on low risk of bias, Outcome 9: Heaviness in the lower legs (dichotomous variable)

Analysis 4.10

Comparison 4: Sensitivity analysis based on low risk of bias, Outcome 10: Heaviness in the lower legs (continuous variable)

Analysis 4.11

Comparison 4: Sensitivity analysis based on low risk of bias, Outcome 11: Swelling in the lower legs (dichotomous variable)

Analysis 4.12

Comparison 4: Sensitivity analysis based on low risk of bias, Outcome 12: Swelling in the lower legs (continuous variable)

Analysis 4.13

Comparison 4: Sensitivity analysis based on low risk of bias, Outcome 13: Participant satisfaction (dichotomous variable)

Analysis 4.14

Comparison 4: Sensitivity analysis based on low risk of bias, Outcome 14: Participant satisfaction (continuous variable)

Analysis 4.15

Comparison 4: Sensitivity analysis based on low risk of bias, Outcome 15: Adverse events

Summary of findings 1. Do phlebotonics improve signs and symptoms of venous insufficiency when compared with placebo?

Outcomes	Anticipated absolute effects *		Relative effect (95% CI)	Number of participants (RCTs)	Certainty of the evidence (GRADE)	Comments
Phlebotonics compared with placebo for venous insufficiency
Patient or population: patients with venous insufficiency Settings: hospital and ambulatory settings Intervention: phlebotonics Comparison: placebo
Outcomes	Risk with placebo	Risk with phlebotonics	Relative effect (95% CI)	Number of participants (RCTs)	Certainty of the evidence (GRADE)	Comments
Oedema in the lower legs (dichotomous variable) Follow‐up: 1‐6 months	575 per 1000	403 per 1000 (362 to 449)	RR 0.70 (0.63 to 0.78)	1245 (13 studies)	⊕⊕⊕⊝ Moderate^a	Phebotonics probably slightly reduce oedema in the lower limb compared to placebo
Oedema in the lower legs (ankle circumference, mm) Follow‐up: 1‐12 months	‐	Mean ankle circumference in the lower legs in the phlebotonic groups was 4.27 mm lower (5.61 to 2.93 lower) than in the placebo groups	‐	2010 (15 studies)	⊕⊕⊕⊝ Moderate^b	Phlebotonics probably slightly reduce ankle perimeter circumference compared to placebo
Quality of life (CIVIQ and other questionnaires) Follow‐up: mean 2‐12 months	‐	The QoL in the phlebotonic groups was 0.06 SMD lower (0.22 lower to 0.1 higher) than in the placebo groups	‐	1639 (5 studies)	⊕⊕⊕⊝ Moderate^c	Phebotonics probably make little or no difference to QoL compared with placebo
Ulcer healing (dichotomous variable) Follow‐up: 1‐12 months	381 per 1000	358 per 1000 (301 to 430)	RR 0.94 (0.79 to 1.13)	461 (6 studies)	⊕⊕⊝⊝ Low^d	Phlebotonics may make little or no difference to ulcer healing compared to placebo
Adverse events Follow‐up: 1‐12 months	158 per 1000	180 per 1000 (161 to 200)	RR 1.14 (1.02 to 1.27)	5789 (37 studies)	⊕⊕⊕⊝ Moderate^e	Phlebotonics probably slightly increase adverse events compared to placebo
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; CIVIQ: Chronic Venous Insufficiency International Questionnaire; QoL: quality of life; RCT: randomised controlled trial; RR: risk ratio
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
^aThe certainty of the evidence was downgraded (1 level) to moderate because of overall risk of bias (10 studies had an unclear risk of bias and two had a high risk of bias) ^bThe certainty of the evidence was downgraded (1 level) to moderate because of overall risk of bias (11 studies had an unclear risk of bias and one had a high risk of bias) ^cThe certainty of the evidence was downgraded (1 level) to moderate because of overall risk of bias (one study had an unclear risk of bias and two had a high risk of bias) ^dThe certainty of the evidence was downgraded (2 levels) to low because of overall risk of bias (four studies had an unclear risk of bias and two had a high risk of bias) and imprecision (low number of events) ^eThe certainty of the evidence was downgraded (1 level) to moderate because of overall risk of bias (28 RCTs had unclear risk of bias and four RCTs had high risk of bias)

Summary of findings 1. Do phlebotonics improve signs and symptoms of venous insufficiency when compared with placebo?

Table 1. Results of all outcomes analysed (all phlebotonics)

Variables	Dichotomous	Continuous
Oedema	RR 0.70 (0.63 to 0.78)	‐
Oedema (mm)	‐	MD ‐4.27 (‐5.61 to ‐2.93)
Oedema (volume)	‐	SMD ‐0.24 (‐0.33 to ‐0.15)
Ulcer cured	NS	‐
Trophic disorders	RR 0.87 (0.81 to 0.95)	‐
Pain	‐	SMD ‐0.35 (‐0.54, ‐0.17)
Cramps	RR 0.72 (0.58 to 0.89)	‐
Restless legs	RR 0.81 (0.72 to 0.91)	‐
Itching	‐	‐
Heaviness	‐	‐
Swelling	RR 0.63 (0.50 to 0.80)	‐
Paraesthesia	RR 0.67 (0.50 to 0.88)	NS
Quality of life	‐	NS
Global assessment by the participant	‐	‐
Adverse events	RR 1.14 (1.02 to 1.27)	‐
Note: No measures of effect are specified when I² was > 75% for the subgroup
RR: risk ratio MD: mean difference NS: non‐significant RR: risk ratio SMD: standardized mean difference

Table 1. Results of all outcomes analysed (all phlebotonics)

Table 2. Results by pharmacological group: aminaftone

Variables	Dichotomous	Continous
Oedema	RR 0.53 (0.28 to 0.99)	SMD ‐0.17 (‐0.61 to 0.28)
Ulcer cured	NS	‐
Trophic disorder	NS	‐
Pain	RR 0.43 (0.23 to 0.79)	‐
Cramps	RR 0.56 (0.31 to 0.99)	‐
Itching	RR 0.53 (0.31 to 0.91)	‐
Heaviness	RR 0.32 (0.17 to 0.60)	‐
Quality of live	‐	MD ‐10.00 (‐17.01 to ‐2.99)
Adverse events	NS	‐
Note: Only 1 study was analyzed
MD: mean difference NS: non‐significant RR: risk ratio SMD: standardized mean difference

Table 2. Results by pharmacological group: aminaftone

Table 3. Results by pharmacological group: calcium dobesilate

Variables	Dichotomous	Continuous
Oedema	‐	‐
Oedema (mm)	‐	NS
Oedema (volume)	‐	SMD ‐0.38 (‐0.51 to ‐0.24)
Ulcer cured	NS	‐
Pain	RR 0.53 (0.35 to 0.82)	NS
Cramps	RR 0.65 (0.50 to 0.84)	‐
Restless legs	RR 0.73 (0.59 to 0.91)	NS
Itching	‐	NS
Heaviness	NS	NS
Swelling	RR 0.19 (0.08 to 0.41)	NS
Paraesthesia	NS	‐
Quality of life	‐	NS
Global assessment by the participant	‐	SMD ‐0.52 (‐0.71 to ‐0.33)
Adverse events	RR 1.22 (1.0 to 1.49)	‐
Note: No measures of effect are specified when I² was > 75% for the subgroup
NS: non‐significant RR: risk ratio SMD: standardized mean difference

Table 3. Results by pharmacological group: calcium dobesilate

Table 4. Results by pharmacological group: Centella asiatica

Variables	Dichotomous	Continuous
Heaviness	NS	‐
Global assessment by the participant	RR 0.28 (0.14 to 0.57)	‐
Adverse events	NS	‐
Note: Only 1 study was analyzed
NS: non‐significant RR: risk ratio

Table 4. Results by pharmacological group: Centella asiatica

Table 5. Results by pharmacological group: diosmine, hidrosmine

Variables	Dichotomous	Continuous
Oedema	RR 0.63 (0.46 to 0.86)	‐
Oedema (mm)	‐	MD ‐5.98 (‐7.78 to ‐4.18)
Ulcer cured	NS	‐
Trophic disorder	RR 0.87 (0.81 to 0.94)	‐
Pain	NS	SMD ‐0.23 (‐0.41 to ‐0.05)
Cramps	RR 0.83 (0.70 to 0.98)	SMD ‐0.46 (‐0.78 to ‐0.14)
Restless legs	NS	‐
Itching	NS	‐
Heaviness	NS	SMD ‐0.69 (‐1.02 to ‐0.36)
Swelling	RR 0.70 (0.52 to 0.94)	SMD ‐0.92 (‐1.26 to ‐0.58)
Paraesthesia	NS	NS
Quality of life	‐	NS
Global assessment by the participant	‐	SMD ‐0.81 (‐1.14 to ‐0.47)
Adverse events	NS	‐
Note: No measures of effect are specified when I² was > 75% for the subgroup
MD: mean difference NS: non‐significant RR: risk ratio SMD: standardized mean difference

Table 5. Results by pharmacological group: diosmine, hidrosmine

Table 6. Results by pharmacological group: French maritime pine bark extract

Variables	Dichotomous	Continuous
Pain	RR 0.66 (0.48 to 0.91)	SMD ‐1.39 (‐2.09 to ‐0.69)
Heaviness	NS	SMD ‐1.50 (‐2.21 to ‐0.79)
Swelling	NS	SMD ‐1.65 (‐2.38 to ‐0.92)
Note: Only 1 study was analyzed
NS: non‐significant RR: risk ratio SMD: standardized mean difference

Table 6. Results by pharmacological group: French maritime pine bark extract

Table 7. Results by pharmacological group: grape seed extract

Variables	Dichotomous	Continuous
Oedema	NS	‐
Adverse events	NS	NS
Note: Only 1 study was analyzed
NS: non‐significant

Table 7. Results by pharmacological group: grape seed extract

Table 8. Results by pharmacological group: rutosides

Variables	Dichotomous	Continuous
Oedema	RR 0.72 (0.64 to 0.81)	‐
Oedema (mm)	‐	NS
Oedema (volume)	‐	SMD ‐0.15 (‐0.16 to ‐0.03)
Ulcer cured	NS	‐
Trophic disorder	NS	‐
Pain	‐	SMD ‐0.71 (‐1.23 to ‐0.19)
Cramps	RR ‐0.83 (‐1.50 to ‐0.16)	NS
Restless legs	NS	‐
Itching	‐	SMD ‐0.58 (‐1.10 to ‐0.06)
Heaviness	RR 0.60 (0.48 to 0.74)	‐
Swelling	RR 0.67 (0.50 to 0.88)	NS
Paraesthesias	RR 0.55 (0.37 to 0.83)	NS
Global assessment by the participant	‐	‐
Adverse events	RR 1.22 (1.04 to 1.43)	‐
Note: No measures of effect are specified when I² was > 75%
NS: non‐significant RR: risk ratio SMD: standardized mean difference

Table 8. Results by pharmacological group: rutosides

Comparison 1. Phlebotonics versus placebo

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1.1 Oedema in the lower legs (dichotomous variable) Show forest plot	13	1245	Risk Ratio (M‐H, Fixed, 95% CI)	0.70 [0.63, 0.78]

1.1.1 Aminaftone	1	82	Risk Ratio (M‐H, Fixed, 95% CI)	0.53 [0.28, 0.99]
1.1.2 Calcium dobesilate	2	290	Risk Ratio (M‐H, Fixed, 95% CI)	0.72 [0.48, 1.07]
1.1.3 Diosmine, Hidrosmine	2	144	Risk Ratio (M‐H, Fixed, 95% CI)	0.63 [0.46, 0.86]
1.1.4 Grape seed extract	1	75	Risk Ratio (M‐H, Fixed, 95% CI)	0.79 [0.58, 1.06]
1.1.5 Rutosides	7	654	Risk Ratio (M‐H, Fixed, 95% CI)	0.72 [0.64, 0.81]
1.2 Ankle perimeter circumference (mm) Show forest plot	15	2010	Mean Difference (IV, Fixed, 95% CI)	‐4.27 [‐5.61, ‐2.93]

1.2.1 Calcium dobesilate	5	1122	Mean Difference (IV, Fixed, 95% CI)	‐1.69 [‐4.84, 1.47]
1.2.2 Diosmine, Hidrosmine	3	286	Mean Difference (IV, Fixed, 95% CI)	‐5.98 [‐7.78, ‐4.18]
1.2.3 Rutosides	7	602	Mean Difference (IV, Fixed, 95% CI)	‐2.45 [‐5.06, 0.15]
1.3 Volume of the leg (mL) Show forest plot	11	2072	Std. Mean Difference (IV, Fixed, 95% CI)	‐0.24 [‐0.33, ‐0.15]

1.3.1 Aminaftone	1	79	Std. Mean Difference (IV, Fixed, 95% CI)	‐0.17 [‐0.61, 0.28]
1.3.2 Calcium dobesilate	4	826	Std. Mean Difference (IV, Fixed, 95% CI)	‐0.38 [‐0.51, ‐0.24]
1.3.3 Rutosides	6	1167	Std. Mean Difference (IV, Fixed, 95% CI)	‐0.15 [‐0.26, ‐0.03]
1.4 Quality of life Show forest plot	5	1639	Std. Mean Difference (IV, Random, 95% CI)	‐0.06 [‐0.22, 0.10]

1.4.1 Aminaftone	1	79	Std. Mean Difference (IV, Random, 95% CI)	‐0.64 [‐1.10, ‐0.19]
1.4.2 Calcium dobesilate at 3 months of treatment	3	968	Std. Mean Difference (IV, Random, 95% CI)	‐0.03 [‐0.16, 0.10]
1.4.3 Diosmine, Hidrosmine	1	592	Std. Mean Difference (IV, Random, 95% CI)	0.04 [‐0.12, 0.20]
1.5 Ulcer healing Show forest plot	6	461	Risk Ratio (M‐H, Random, 95% CI)	0.94 [0.79, 1.13]

1.5.1 Aminaftone	1	100	Risk Ratio (M‐H, Random, 95% CI)	0.75 [0.18, 3.18]
1.5.2 Calcium dobesilate	1	69	Risk Ratio (M‐H, Random, 95% CI)	1.09 [0.69, 1.74]
1.5.3 Diosmine, Hidrosmine	2	133	Risk Ratio (M‐H, Random, 95% CI)	0.84 [0.69, 1.01]
1.5.4 Rutosides	2	159	Risk Ratio (M‐H, Random, 95% CI)	1.28 [0.87, 1.86]
1.6 Trophic disorders (dichotomous variable) Show forest plot	6	705	Risk Ratio (M‐H, Fixed, 95% CI)	0.87 [0.81, 0.95]

1.6.1 Aminaftone	1	97	Risk Ratio (M‐H, Fixed, 95% CI)	0.77 [0.41, 1.44]
1.6.2 Diosmine, Hidrosmine	4	504	Risk Ratio (M‐H, Fixed, 95% CI)	0.87 [0.81, 0.94]
1.6.3 Rutosides	1	104	Risk Ratio (M‐H, Fixed, 95% CI)	0.94 [0.71, 1.25]
1.7 Pain in the lower legs (dichotomous variable) Show forest plot	21		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only

1.7.1 Aminaftone	1	97	Risk Ratio (M‐H, Random, 95% CI)	0.43 [0.23, 0.79]
1.7.2 Calcium dobesilate	5	705	Risk Ratio (M‐H, Random, 95% CI)	0.53 [0.35, 0.82]
1.7.3 Diosmine, Hidrosmine	4	271	Risk Ratio (M‐H, Random, 95% CI)	0.82 [0.63, 1.08]
1.7.4 French maritime pine bark extract	1	40	Risk Ratio (M‐H, Random, 95% CI)	0.66 [0.48, 0.91]
1.7.5 Rutosides	10	1485	Risk Ratio (M‐H, Random, 95% CI)	0.63 [0.48, 0.83]
1.8 Pain in the lower legs (continuous variable) Show forest plot	12	2232	Std. Mean Difference (IV, Random, 95% CI)	‐0.35 [‐0.54, ‐0.17]

1.8.1 Calcium dobesilate	5	1127	Std. Mean Difference (IV, Random, 95% CI)	‐0.14 [‐0.31, 0.03]
1.8.2 Diosmine, Hidrosmine	3	846	Std. Mean Difference (IV, Random, 95% CI)	‐0.23 [‐0.41, ‐0.05]
1.8.3 French maritime pine bark extract	1	40	Std. Mean Difference (IV, Random, 95% CI)	‐1.39 [‐2.09, ‐0.69]
1.8.4 Rutosides	3	219	Std. Mean Difference (IV, Random, 95% CI)	‐0.71 [‐1.23, ‐0.19]
1.9 Cramps in the lower legs (dichotomous variable) Show forest plot	14	1793	Risk Ratio (M‐H, Random, 95% CI)	0.72 [0.58, 0.89]

1.9.1 Aminaftone	1	97	Risk Ratio (M‐H, Random, 95% CI)	0.56 [0.31, 0.99]
1.9.2 Calcium dobesilate	2	255	Risk Ratio (M‐H, Random, 95% CI)	0.65 [0.50, 0.84]
1.9.3 Diosmine, Hidrosmine	3	214	Risk Ratio (M‐H, Random, 95% CI)	0.83 [0.70, 0.98]
1.9.4 Rutosides	8	1227	Risk Ratio (M‐H, Random, 95% CI)	0.70 [0.47, 1.02]
1.10 Cramps in the lower legs (continuous variable) Show forest plot	4		Std. Mean Difference (IV, Random, 95% CI)	Subtotals only

1.10.1 Calcium dobesilate	1	415	Std. Mean Difference (IV, Random, 95% CI)	‐0.10 [‐0.29, 0.09]
1.10.2 Diosmine, Hidrosmine	1	150	Std. Mean Difference (IV, Random, 95% CI)	‐0.46 [‐0.78, ‐0.14]
1.10.3 Rutosides	2	164	Std. Mean Difference (IV, Random, 95% CI)	‐0.83 [‐1.50, ‐0.16]
1.11 Restless legs (dichotomous variable) Show forest plot	7	652	Risk Ratio (M‐H, Fixed, 95% CI)	0.81 [0.72, 0.91]

1.11.1 Calcium dobesilate	2	255	Risk Ratio (M‐H, Fixed, 95% CI)	0.73 [0.59, 0.91]
1.11.2 Diosmine, Hidrosmine	1	70	Risk Ratio (M‐H, Fixed, 95% CI)	0.90 [0.70, 1.15]
1.11.3 Rutosides	4	327	Risk Ratio (M‐H, Fixed, 95% CI)	0.85 [0.72, 1.01]
1.12 Itching in the lower legs (dichotomous variable) Show forest plot	4		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only

1.12.1 Aminaftone	1	97	Risk Ratio (M‐H, Random, 95% CI)	0.53 [0.31, 0.91]
1.12.2 Diosmine, Hidrosmine	1	34	Risk Ratio (M‐H, Random, 95% CI)	1.63 [0.51, 5.25]
1.12.3 Rutosides	2	274	Risk Ratio (M‐H, Random, 95% CI)	0.68 [0.21, 2.21]
1.13 Itching in the lower legs (continuous variable) Show forest plot	2		Std. Mean Difference (IV, Random, 95% CI)	Subtotals only

1.13.1 Calcium dobesilate	1	416	Std. Mean Difference (IV, Random, 95% CI)	0.09 [‐0.11, 0.28]
1.13.2 Rutosides	1	60	Std. Mean Difference (IV, Random, 95% CI)	‐0.58 [‐1.10, ‐0.06]
1.14 Heaviness in the lower legs (dichotomous variable) Show forest plot	19		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only

1.14.1 Aminaftone	1	97	Risk Ratio (M‐H, Random, 95% CI)	0.32 [0.17, 0.60]
1.14.2 Calcium dobesilate	3	305	Risk Ratio (M‐H, Random, 95% CI)	0.33 [0.08, 1.42]
1.14.3 Centella asiatica	1	63	Risk Ratio (M‐H, Random, 95% CI)	0.62 [0.32, 1.19]
1.14.4 Diosmine, Hidrosmine	4	241	Risk Ratio (M‐H, Random, 95% CI)	0.60 [0.35, 1.05]
1.14.5 French maritime pine bark extract	1	40	Risk Ratio (M‐H, Random, 95% CI)	0.90 [0.76, 1.07]
1.14.6 Rutosides	9	1420	Risk Ratio (M‐H, Random, 95% CI)	0.60 [0.48, 0.74]
1.15 Heaviness in the lower legs (continuous variable) Show forest plot	10		Std. Mean Difference (IV, Random, 95% CI)	Subtotals only

1.15.1 Calcium dobesilate	2	483	Std. Mean Difference (IV, Random, 95% CI)	‐0.05 [‐0.23, 0.13]
1.15.2 Diosmine, Hidrosmine	1	150	Std. Mean Difference (IV, Random, 95% CI)	‐0.69 [‐1.02, ‐0.36]
1.15.3 French maritime pine bark extract	1	40	Std. Mean Difference (IV, Random, 95% CI)	‐1.50 [‐2.21, ‐0.79]
1.15.4 Rutosides	6	441	Std. Mean Difference (IV, Random, 95% CI)	‐1.11 [‐1.87, ‐0.36]
1.16 Swelling in the lower legs (dichotomous variable) Show forest plot	14	1072	Risk Ratio (M‐H, Random, 95% CI)	0.63 [0.50, 0.80]

1.16.1 Calcium dobesilate	2	80	Risk Ratio (M‐H, Random, 95% CI)	0.19 [0.08, 0.41]
1.16.2 Diosmine, Hidrosmine	2	104	Risk Ratio (M‐H, Random, 95% CI)	0.70 [0.52, 0.94]
1.16.3 French maritime pine bark extract	1	40	Risk Ratio (M‐H, Random, 95% CI)	0.80 [0.64, 1.02]
1.16.4 Rutosides	9	848	Risk Ratio (M‐H, Random, 95% CI)	0.67 [0.50, 0.88]
1.17 Swelling in the lower legs (continuous variable) Show forest plot	6		Std. Mean Difference (IV, Random, 95% CI)	Subtotals only

1.17.1 Calcium dobesilate	1	417	Std. Mean Difference (IV, Random, 95% CI)	‐0.05 [‐0.24, 0.15]
1.17.2 Diosmine, Hidrosmine	1	150	Std. Mean Difference (IV, Random, 95% CI)	‐0.92 [‐1.26, ‐0.58]
1.17.3 French maritime pine bark extract	1	40	Std. Mean Difference (IV, Random, 95% CI)	‐1.65 [‐2.38, ‐0.92]
1.17.4 Rutosides	3	264	Std. Mean Difference (IV, Random, 95% CI)	‐1.73 [‐3.50, 0.04]
1.18 Paraesthesia in the lower legs (dichotomous variable) Show forest plot	9	1456	Risk Ratio (M‐H, Random, 95% CI)	0.67 [0.50, 0.88]

1.18.1 Calcium dobesilate	3	305	Risk Ratio (M‐H, Random, 95% CI)	0.74 [0.51, 1.08]
1.18.2 Diosmine, Hidrosmine	2	144	Risk Ratio (M‐H, Random, 95% CI)	0.80 [0.62, 1.05]
1.18.3 Rutosides	4	1007	Risk Ratio (M‐H, Random, 95% CI)	0.55 [0.37, 0.83]
1.19 Paraesthesia in the lower legs (continuous variable) Show forest plot	2	188	Std. Mean Difference (IV, Fixed, 95% CI)	‐0.15 [‐0.44, 0.13]

1.19.1 Diosmine, Hidrosmine	1	150	Std. Mean Difference (IV, Fixed, 95% CI)	‐0.12 [‐0.44, 0.21]
1.19.2 Rutosides	1	38	Std. Mean Difference (IV, Fixed, 95% CI)	‐0.31 [‐0.96, 0.33]
1.20 Participant satisfaction (dichotomous variable) Show forest plot	16		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only

1.20.1 Calcium dobesilate	3	506	Risk Ratio (M‐H, Random, 95% CI)	0.72 [0.36, 1.46]
1.20.2 Diosmine, Hidrosmine	4	451	Risk Ratio (M‐H, Random, 95% CI)	0.66 [0.43, 1.02]
1.20.3 Centella asiatica	1	80	Risk Ratio (M‐H, Random, 95% CI)	0.28 [0.14, 0.57]
1.20.4 Rutosides	8	1167	Risk Ratio (M‐H, Random, 95% CI)	0.50 [0.30, 0.84]
1.21 Participant satisfaction (continuous variable) Show forest plot	7		Std. Mean Difference (IV, Random, 95% CI)	Subtotals only

1.21.1 Calcium dobesilate	2	448	Std. Mean Difference (IV, Random, 95% CI)	‐0.52 [‐0.71, ‐0.33]
1.21.2 Diosmine, Hidrosmine	1	150	Std. Mean Difference (IV, Random, 95% CI)	‐0.81 [‐1.14, ‐0.47]
1.21.3 Rutosides	4	283	Std. Mean Difference (IV, Random, 95% CI)	‐1.18 [‐1.96, ‐0.39]
1.22 Adverse events Show forest plot	37	5789	Risk Ratio (M‐H, Fixed, 95% CI)	1.14 [1.02, 1.27]

1.22.1 Aminaftone	1	79	Risk Ratio (M‐H, Fixed, 95% CI)	0.60 [0.06, 6.32]
1.22.2 Calcium dobesilate	8	1824	Risk Ratio (M‐H, Fixed, 95% CI)	1.22 [1.00, 1.49]
1.22.3 Centella asiatica	1	94	Risk Ratio (M‐H, Fixed, 95% CI)	1.14 [0.58, 2.23]
1.22.4 Diosmine, Hidrosmine	9	1429	Risk Ratio (M‐H, Fixed, 95% CI)	0.93 [0.72, 1.19]
1.22.5 Grape seed extract	1	75	Risk Ratio (M‐H, Fixed, 95% CI)	0.57 [0.19, 1.74]
1.22.6 Rutosides	17	2288	Risk Ratio (M‐H, Fixed, 95% CI)	1.22 [1.04, 1.43]

Comparison 1. Phlebotonics versus placebo

Comparison 2. Sensitivity analysis excluding studies that allowed the use of elastic stockings

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
2.1 Oedema in the lower legs (dichotomous variable) Show forest plot	12	1131	Risk Ratio (M‐H, Fixed, 95% CI)	0.68 [0.60, 0.76]

2.1.1 Aminaftone	1	82	Risk Ratio (M‐H, Fixed, 95% CI)	0.53 [0.28, 0.99]
2.1.2 Calcium dobesilate	2	290	Risk Ratio (M‐H, Fixed, 95% CI)	0.72 [0.48, 1.07]
2.1.3 Diosmine, Hidrosmine	2	144	Risk Ratio (M‐H, Fixed, 95% CI)	0.63 [0.46, 0.86]
2.1.4 Grape seed extract	1	75	Risk Ratio (M‐H, Fixed, 95% CI)	0.79 [0.58, 1.06]
2.1.5 Rutosides	6	540	Risk Ratio (M‐H, Fixed, 95% CI)	0.68 [0.58, 0.78]
2.2 Ankle perimeter circumference (mm) Show forest plot	10	1212	Mean Difference (IV, Fixed, 95% CI)	‐4.59 [‐6.02, ‐3.16]

2.2.1 Calcium dobesilate	3	502	Mean Difference (IV, Fixed, 95% CI)	‐0.80 [‐4.95, 3.34]
2.2.2 Diosmine, Hidrosmine	2	246	Mean Difference (IV, Fixed, 95% CI)	‐5.90 [‐7.72, ‐4.07]
2.2.3 Rutosides	5	464	Mean Difference (IV, Fixed, 95% CI)	‐3.28 [‐6.06, ‐0.50]
2.3 Volume of the leg (mL) Show forest plot	9	1153	Std. Mean Difference (IV, Fixed, 95% CI)	‐0.30 [‐0.42, ‐0.19]

2.3.1 Aminaftone	1	79	Std. Mean Difference (IV, Fixed, 95% CI)	‐0.17 [‐0.61, 0.28]
2.3.2 Calcium dobesilate	3	587	Std. Mean Difference (IV, Fixed, 95% CI)	‐0.33 [‐0.49, ‐0.17]
2.3.3 Rutosides	5	487	Std. Mean Difference (IV, Fixed, 95% CI)	‐0.29 [‐0.47, ‐0.11]
2.4 Quality of life Show forest plot	3	1022	Std. Mean Difference (IV, Random, 95% CI)	‐0.13 [‐0.41, 0.15]

2.4.1 Aminaftone	1	79	Std. Mean Difference (IV, Random, 95% CI)	‐0.64 [‐1.10, ‐0.19]
2.4.2 Calcium dobesilate	1	351	Std. Mean Difference (IV, Random, 95% CI)	‐0.03 [‐0.23, 0.18]
2.4.3 Diosmine, Hidrosmine	1	592	Std. Mean Difference (IV, Random, 95% CI)	0.04 [‐0.12, 0.20]
2.5 Ulcer healing Show forest plot	2	128	Risk Ratio (M‐H, Random, 95% CI)	0.91 [0.27, 3.10]

2.5.1 Aminaftone	1	100	Risk Ratio (M‐H, Random, 95% CI)	0.75 [0.18, 3.18]
2.5.2 Diosmine, Hidrosmine	1	28	Risk Ratio (M‐H, Random, 95% CI)	1.50 [0.15, 14.68]
2.6 Trophic disorders (dichotomous variable) Show forest plot	5	601	Risk Ratio (M‐H, Fixed, 95% CI)	0.86 [0.79, 0.94]

2.6.1 Aminaftone	1	97	Risk Ratio (M‐H, Fixed, 95% CI)	0.77 [0.41, 1.44]
2.6.2 Diosmine, Hidrosmine	4	504	Risk Ratio (M‐H, Fixed, 95% CI)	0.87 [0.81, 0.94]
2.7 Pain in the lower legs (dichotomous variable) Show forest plot	18	1818	Risk Ratio (M‐H, Random, 95% CI)	0.70 [0.60, 0.82]

2.7.1 Aminaftone	1	97	Risk Ratio (M‐H, Random, 95% CI)	0.43 [0.23, 0.79]
2.7.2 Calcium dobesilate	5	705	Risk Ratio (M‐H, Random, 95% CI)	0.53 [0.35, 0.82]
2.7.3 Diosmine, Hidrosmine	4	271	Risk Ratio (M‐H, Random, 95% CI)	0.82 [0.63, 1.08]
2.7.4 Rutosides	8	745	Risk Ratio (M‐H, Random, 95% CI)	0.75 [0.61, 0.91]
2.8 Pain in the lower legs (continuous variable) Show forest plot	6		Std. Mean Difference (IV, Random, 95% CI)	Subtotals only

2.8.1 Calcium dobesilate	1	351	Std. Mean Difference (IV, Random, 95% CI)	‐0.21 [‐0.42, 0.00]
2.8.2 Diosmine, Hidrosmine	3	846	Std. Mean Difference (IV, Random, 95% CI)	‐0.23 [‐0.41, ‐0.05]
2.8.3 Rutosides	2	164	Std. Mean Difference (IV, Random, 95% CI)	‐0.96 [‐1.33, ‐0.59]
2.9 Cramps in the lower legs (dichotomous variable) Show forest plot	12	1603	Risk Ratio (M‐H, Random, 95% CI)	0.72 [0.57, 0.91]

2.9.1 Aminaftone	1	97	Risk Ratio (M‐H, Random, 95% CI)	0.56 [0.31, 0.99]
2.9.2 Calcium dobesilate	2	255	Risk Ratio (M‐H, Random, 95% CI)	0.65 [0.50, 0.84]
2.9.3 Diosmine, Hidrosmine	2	104	Risk Ratio (M‐H, Random, 95% CI)	0.87 [0.69, 1.09]
2.9.4 Rutosides	7	1147	Risk Ratio (M‐H, Random, 95% CI)	0.73 [0.50, 1.06]
2.10 Cramps in the lower legs (continuous variable) Show forest plot	3	314	Std. Mean Difference (IV, Random, 95% CI)	‐0.70 [‐1.15, ‐0.24]

2.10.1 Diosmine, Hidrosmine	1	150	Std. Mean Difference (IV, Random, 95% CI)	‐0.46 [‐0.78, ‐0.14]
2.10.2 Rutosides	2	164	Std. Mean Difference (IV, Random, 95% CI)	‐0.83 [‐1.50, ‐0.16]
2.11 Restless legs (dichotomous variable) Show forest plot	6	572	Risk Ratio (M‐H, Fixed, 95% CI)	0.81 [0.72, 0.91]

2.11.1 Calcium dobesilate	2	255	Risk Ratio (M‐H, Fixed, 95% CI)	0.73 [0.59, 0.91]
2.11.2 Diosmine, Hidrosmine	1	70	Risk Ratio (M‐H, Fixed, 95% CI)	0.90 [0.70, 1.15]
2.11.3 Rutosides	3	247	Risk Ratio (M‐H, Fixed, 95% CI)	0.86 [0.73, 1.01]
2.12 Itching in the lower legs (dichotomous variable) Show forest plot	4		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only

2.12.1 Aminaftone	1	97	Risk Ratio (M‐H, Random, 95% CI)	0.53 [0.31, 0.91]
2.12.2 Diosmine, Hidrosmine	1	34	Risk Ratio (M‐H, Random, 95% CI)	1.63 [0.51, 5.25]
2.12.3 Rutosides	2	274	Risk Ratio (M‐H, Random, 95% CI)	0.68 [0.21, 2.21]
2.13 Itching in the lower legs (continuous variable) Show forest plot	1		Std. Mean Difference (IV, Fixed, 95% CI)	Totals not selected

2.13.1 Rutosides	1		Std. Mean Difference (IV, Fixed, 95% CI)	Totals not selected
2.14 Heaviness in the lower legs (dichotomous variable) Show forest plot	16		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only

2.14.1 Aminaftone	1	97	Risk Ratio (M‐H, Random, 95% CI)	0.32 [0.17, 0.60]
2.14.2 Calcium dobesilate	3	305	Risk Ratio (M‐H, Random, 95% CI)	0.33 [0.08, 1.42]
2.14.3 Centella asiatica	1	63	Risk Ratio (M‐H, Random, 95% CI)	0.62 [0.32, 1.19]
2.14.4 Diosmine, Hidrosmine	3	201	Risk Ratio (M‐H, Random, 95% CI)	0.60 [0.29, 1.22]
2.14.5 Rutosides	8	531	Risk Ratio (M‐H, Random, 95% CI)	0.55 [0.38, 0.80]
2.15 Heaviness in the lower legs (continuous variable) Show forest plot	6		Std. Mean Difference (IV, Random, 95% CI)	Subtotals only

2.15.1 Diosmine, Hidrosmine	1	150	Std. Mean Difference (IV, Random, 95% CI)	‐0.69 [‐1.02, ‐0.36]
2.15.2 Rutosides	5	360	Std. Mean Difference (IV, Random, 95% CI)	‐1.27 [‐2.22, ‐0.32]
2.16 Swelling in the lower legs (dichotomous variable) Show forest plot	12	952	Risk Ratio (M‐H, Random, 95% CI)	0.66 [0.53, 0.82]

2.16.1 Calcium dobesilate	2	80	Risk Ratio (M‐H, Random, 95% CI)	0.19 [0.08, 0.41]
2.16.2 Diosmine, Hidrosmine	2	104	Risk Ratio (M‐H, Random, 95% CI)	0.70 [0.52, 0.94]
2.16.3 Rutosides	8	768	Risk Ratio (M‐H, Random, 95% CI)	0.72 [0.58, 0.89]
2.17 Swelling in the lower legs (continuous variable) Show forest plot	4		Std. Mean Difference (IV, Random, 95% CI)	Subtotals only

2.17.1 Diosmine, Hidrosmine	1	150	Std. Mean Difference (IV, Random, 95% CI)	‐0.92 [‐1.26, ‐0.58]
2.17.2 Rutosides	3	264	Std. Mean Difference (IV, Random, 95% CI)	‐1.73 [‐3.50, 0.04]
2.18 Paraesthesias in the lower legs (dichotomous variable) Show forest plot	7	716	Risk Ratio (M‐H, Fixed, 95% CI)	0.75 [0.64, 0.88]

2.18.1 Calcium dobesilate	3	305	Risk Ratio (M‐H, Fixed, 95% CI)	0.77 [0.58, 1.01]
2.18.2 Diosmine, Hidrosmine	2	144	Risk Ratio (M‐H, Fixed, 95% CI)	0.81 [0.61, 1.06]
2.18.3 Rutosides	2	267	Risk Ratio (M‐H, Fixed, 95% CI)	0.68 [0.51, 0.91]
2.19 Paraesthesias in the lower legs (continuous variable) Show forest plot	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected

2.19.1 Diosmine, Hidrosmine	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected
2.20 Participant satisfaction (dichotomous variable) Show forest plot	12	1193	Risk Ratio (M‐H, Random, 95% CI)	0.69 [0.53, 0.90]

2.20.1 Calcium dobesilate	3	515	Risk Ratio (M‐H, Random, 95% CI)	0.71 [0.43, 1.17]
2.20.2 Diosmine, Hidrosmine	2	171	Risk Ratio (M‐H, Random, 95% CI)	0.84 [0.68, 1.04]
2.20.3 Rutosides	7	507	Risk Ratio (M‐H, Random, 95% CI)	0.52 [0.28, 0.98]
2.21 Participant satisfaction (continuous variable) Show forest plot	4		Std. Mean Difference (IV, Random, 95% CI)	Totals not selected

2.21.1 Diosmine, Hidrosmine	1		Std. Mean Difference (IV, Random, 95% CI)	Totals not selected
2.21.2 Rutosides	3		Std. Mean Difference (IV, Random, 95% CI)	Totals not selected
2.22 Adverse events Show forest plot	27	3433	Risk Ratio (M‐H, Fixed, 95% CI)	1.12 [0.95, 1.33]

2.22.1 Aminaftone	1	79	Risk Ratio (M‐H, Fixed, 95% CI)	0.60 [0.06, 6.32]
2.22.2 Calcium dobesilate	5	935	Risk Ratio (M‐H, Fixed, 95% CI)	1.12 [0.82, 1.53]
2.22.3 Centella asiatica	1	94	Risk Ratio (M‐H, Fixed, 95% CI)	1.14 [0.58, 2.23]
2.22.4 Diosmine, Hidrosmine	7	1124	Risk Ratio (M‐H, Fixed, 95% CI)	0.97 [0.74, 1.27]
2.22.5 Grape seed extract	1	75	Risk Ratio (M‐H, Fixed, 95% CI)	0.57 [0.19, 1.74]
2.22.6 Rutosides	12	1126	Risk Ratio (M‐H, Fixed, 95% CI)	1.54 [1.08, 2.19]

Comparison 2. Sensitivity analysis excluding studies that allowed the use of elastic stockings

Comparison 3. Sensitivity analysis of published studies only

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
3.1 Oedema in the lower legs (dichotomous variable) Show forest plot	12	1088	Risk Ratio (M‐H, Fixed, 95% CI)	0.70 [0.63, 0.78]

3.1.1 Aminaftone	1	82	Risk Ratio (M‐H, Fixed, 95% CI)	0.53 [0.28, 0.99]
3.1.2 Calcium dobesilate	2	290	Risk Ratio (M‐H, Fixed, 95% CI)	0.72 [0.48, 1.07]
3.1.3 Diosmine, Hidrosmine	2	144	Risk Ratio (M‐H, Fixed, 95% CI)	0.63 [0.46, 0.86]
3.1.4 Grape seed extract	1	75	Risk Ratio (M‐H, Fixed, 95% CI)	0.79 [0.58, 1.06]
3.1.5 Rutosides	6	497	Risk Ratio (M‐H, Fixed, 95% CI)	0.72 [0.64, 0.81]
3.2 Ankle perimeter circumference (mm) Show forest plot	13	1796	Mean Difference (IV, Random, 95% CI)	‐3.61 [‐6.77, ‐0.45]

3.2.1 Calcium dobesilate	5	1122	Mean Difference (IV, Random, 95% CI)	‐3.17 [‐8.37, 2.02]
3.2.2 Diosmine, Hidrosmine	3	286	Mean Difference (IV, Random, 95% CI)	‐5.98 [‐7.78, ‐4.18]
3.2.3 Rutosides	5	388	Mean Difference (IV, Random, 95% CI)	‐2.18 [‐9.79, 5.43]
3.3 Volume of the leg (mL) Show forest plot	10	1392	Std. Mean Difference (IV, Fixed, 95% CI)	‐0.34 [‐0.44, ‐0.23]

3.3.1 Aminaftone	1	79	Std. Mean Difference (IV, Fixed, 95% CI)	‐0.17 [‐0.61, 0.28]
3.3.2 Calcium dobesilate	4	826	Std. Mean Difference (IV, Fixed, 95% CI)	‐0.38 [‐0.51, ‐0.24]
3.3.3 Rutosides	5	487	Std. Mean Difference (IV, Fixed, 95% CI)	‐0.29 [‐0.47, ‐0.11]
3.4 Quality of life Show forest plot	5	1639	Std. Mean Difference (IV, Random, 95% CI)	‐0.06 [‐0.22, 0.10]

3.4.1 Aminaftone	1	79	Std. Mean Difference (IV, Random, 95% CI)	‐0.64 [‐1.10, ‐0.19]
3.4.2 Calcium dobesilate	3	968	Std. Mean Difference (IV, Random, 95% CI)	‐0.03 [‐0.16, 0.10]
3.4.3 Diosmine, Hidrosmine	1	592	Std. Mean Difference (IV, Random, 95% CI)	0.04 [‐0.12, 0.20]
3.5 Patients with ulcer (dichotomous variable) Show forest plot	5	392	Risk Ratio (M‐H, Fixed, 95% CI)	0.95 [0.78, 1.15]

3.5.1 Aminaftone	1	100	Risk Ratio (M‐H, Fixed, 95% CI)	0.75 [0.18, 3.18]
3.5.2 Diosmine, Hidrosmine	2	133	Risk Ratio (M‐H, Fixed, 95% CI)	0.85 [0.70, 1.03]
3.5.3 Rutosides	2	159	Risk Ratio (M‐H, Fixed, 95% CI)	1.25 [0.84, 1.87]
3.6 Trophic disorders (dichotomous variable) Show forest plot	6	705	Risk Ratio (M‐H, Fixed, 95% CI)	0.87 [0.81, 0.95]

3.6.1 Aminaftone	1	97	Risk Ratio (M‐H, Fixed, 95% CI)	0.77 [0.41, 1.44]
3.6.2 Diosmine, Hidrosmine	4	504	Risk Ratio (M‐H, Fixed, 95% CI)	0.87 [0.81, 0.94]
3.6.3 Rutosides	1	104	Risk Ratio (M‐H, Fixed, 95% CI)	0.94 [0.71, 1.25]
3.7 Pain in the lower legs (dichotomous variable) Show forest plot	19		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only

3.7.1 Aminaftone	1	97	Risk Ratio (M‐H, Random, 95% CI)	0.43 [0.23, 0.79]
3.7.2 Calcium dobesilate	5	705	Risk Ratio (M‐H, Random, 95% CI)	0.53 [0.35, 0.82]
3.7.3 Diosmine, Hidrosmine	4	271	Risk Ratio (M‐H, Random, 95% CI)	0.82 [0.63, 1.08]
3.7.4 French maritime pine bark extract	1	40	Risk Ratio (M‐H, Random, 95% CI)	0.66 [0.48, 0.91]
3.7.5 Rutosides	8	1318	Risk Ratio (M‐H, Random, 95% CI)	0.61 [0.45, 0.84]
3.8 Pain in the lower legs (continuous variable) Show forest plot	11		Std. Mean Difference (IV, Random, 95% CI)	Subtotals only

3.8.1 Calcium dobesilate	4	1075	Std. Mean Difference (IV, Random, 95% CI)	‐0.15 [‐0.34, 0.05]
3.8.2 Diosmine, Hidrosmine	3	846	Std. Mean Difference (IV, Random, 95% CI)	‐0.23 [‐0.41, ‐0.05]
3.8.3 French maritime pine bark extract	1	40	Std. Mean Difference (IV, Random, 95% CI)	‐1.39 [‐2.09, ‐0.69]
3.8.4 Rutosides	3	219	Std. Mean Difference (IV, Random, 95% CI)	‐0.71 [‐1.23, ‐0.19]
3.9 Cramps in the lower legs (dichotomous variable) Show forest plot	12		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only

3.9.1 Aminaftone	1	97	Risk Ratio (M‐H, Random, 95% CI)	0.56 [0.31, 0.99]
3.9.2 Calcium dobesilate	2	255	Risk Ratio (M‐H, Random, 95% CI)	0.65 [0.50, 0.84]
3.9.3 Diosmine, Hidrosmine	3	214	Risk Ratio (M‐H, Random, 95% CI)	0.83 [0.70, 0.98]
3.9.4 Rutosides	6	1060	Risk Ratio (M‐H, Random, 95% CI)	0.69 [0.45, 1.05]
3.10 Cramps in the lower legs (continuous variable) Show forest plot	4		Std. Mean Difference (IV, Random, 95% CI)	Subtotals only

3.10.1 Calcium dobesilate	1	415	Std. Mean Difference (IV, Random, 95% CI)	‐0.10 [‐0.29, 0.09]
3.10.2 Diosmine, Hidrosmine	1	150	Std. Mean Difference (IV, Random, 95% CI)	‐0.46 [‐0.78, ‐0.14]
3.10.3 Rutosides	2	164	Std. Mean Difference (IV, Random, 95% CI)	‐0.83 [‐1.50, ‐0.16]
3.11 Restless legs (dichotomous variable) Show forest plot	7	652	Risk Ratio (M‐H, Fixed, 95% CI)	0.81 [0.72, 0.91]

3.11.1 Calcium dobesilate	2	255	Risk Ratio (M‐H, Fixed, 95% CI)	0.73 [0.59, 0.91]
3.11.2 Diosmine, Hidrosmine	1	70	Risk Ratio (M‐H, Fixed, 95% CI)	0.90 [0.70, 1.15]
3.11.3 Rutosides	4	327	Risk Ratio (M‐H, Fixed, 95% CI)	0.85 [0.72, 1.01]
3.12 Itching in the lower legs (dichotomous variable) Show forest plot	4		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only

3.12.1 Aminaftone	1	97	Risk Ratio (M‐H, Random, 95% CI)	0.53 [0.31, 0.91]
3.12.2 Diosmine, Hidrosmine	1	34	Risk Ratio (M‐H, Random, 95% CI)	1.63 [0.51, 5.25]
3.12.3 Rutosides	2	274	Risk Ratio (M‐H, Random, 95% CI)	0.68 [0.21, 2.21]
3.13 Itching in the lower legs (continuous variable) Show forest plot	2		Std. Mean Difference (IV, Random, 95% CI)	Subtotals only

3.13.1 Calccium dobesilate	1	416	Std. Mean Difference (IV, Random, 95% CI)	0.09 [‐0.11, 0.28]
3.13.2 Rutosides	1	60	Std. Mean Difference (IV, Random, 95% CI)	‐0.58 [‐1.10, ‐0.06]
3.14 Heaviness in the lower legs (dichotomous variable) Show forest plot	17		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only

3.14.1 Aminaftone	1	97	Risk Ratio (M‐H, Random, 95% CI)	0.32 [0.17, 0.60]
3.14.2 Calcium dobesilate	3	305	Risk Ratio (M‐H, Random, 95% CI)	0.33 [0.08, 1.42]
3.14.3 Centella asiatica	1	63	Risk Ratio (M‐H, Random, 95% CI)	0.62 [0.32, 1.19]
3.14.4 Diosmine, Hidrosmine	4	241	Risk Ratio (M‐H, Random, 95% CI)	0.60 [0.35, 1.05]
3.14.5 French maritime pine bark extract	1	40	Risk Ratio (M‐H, Random, 95% CI)	0.90 [0.76, 1.07]
3.14.6 Rutosides	7	1253	Risk Ratio (M‐H, Random, 95% CI)	0.62 [0.49, 0.78]
3.15 Heaviness in the lower legs (continuous variable) Show forest plot	9		Std. Mean Difference (IV, Random, 95% CI)	Subtotals only

3.15.1 Calcium dobesilate	2	483	Std. Mean Difference (IV, Random, 95% CI)	‐0.05 [‐0.23, 0.13]
3.15.2 Diosmine, Hidrosmine	1	150	Std. Mean Difference (IV, Random, 95% CI)	‐0.69 [‐1.02, ‐0.36]
3.15.3 French maritime pine bark extract	1	40	Std. Mean Difference (IV, Random, 95% CI)	‐1.50 [‐2.21, ‐0.79]
3.15.4 Rutosides	5	360	Std. Mean Difference (IV, Random, 95% CI)	‐1.27 [‐2.22, ‐0.32]
3.16 Swelling in the lower legs (dichotomous variable) Show forest plot	12	905	Risk Ratio (M‐H, Random, 95% CI)	0.63 [0.49, 0.81]

3.16.1 Calcium dobesilate	2	80	Risk Ratio (M‐H, Random, 95% CI)	0.19 [0.08, 0.41]
3.16.2 Diosmine, Hidrosmine	2	104	Risk Ratio (M‐H, Random, 95% CI)	0.70 [0.52, 0.94]
3.16.3 French maritime pine bark extract	1	40	Risk Ratio (M‐H, Random, 95% CI)	0.80 [0.64, 1.02]
3.16.4 Rutosides	7	681	Risk Ratio (M‐H, Random, 95% CI)	0.67 [0.49, 0.91]
3.17 Swelling in the lower legs (continuous variable) Show forest plot	6		Std. Mean Difference (IV, Random, 95% CI)	Subtotals only

3.17.1 Calcium dobesilate	1	417	Std. Mean Difference (IV, Random, 95% CI)	‐0.05 [‐0.24, 0.15]
3.17.2 Diosmine, Hidrosmine	1	150	Std. Mean Difference (IV, Random, 95% CI)	‐1.15 [‐1.50, ‐0.80]
3.17.3 French maritime pine bark extract	1	40	Std. Mean Difference (IV, Random, 95% CI)	‐1.65 [‐2.38, ‐0.92]
3.17.4 Rutosides	3	264	Std. Mean Difference (IV, Random, 95% CI)	‐1.73 [‐3.50, 0.04]
3.18 Paraesthesias in the lower legs (dichotomous variable) Show forest plot	8	1309	Risk Ratio (M‐H, Random, 95% CI)	0.63 [0.48, 0.84]

3.18.1 Calcium dobesilate	3	305	Risk Ratio (M‐H, Random, 95% CI)	0.74 [0.51, 1.08]
3.18.2 Diosmine, Hidrosmine	2	144	Risk Ratio (M‐H, Random, 95% CI)	0.80 [0.62, 1.05]
3.18.3 Rutosides	3	860	Risk Ratio (M‐H, Random, 95% CI)	0.48 [0.35, 0.66]
3.19 Paraesthesias in the lower legs (continuous variable) Show forest plot	1		Std. Mean Difference (IV, Fixed, 95% CI)	Totals not selected

3.19.1 Diosmine, Hidrosmine	1		Std. Mean Difference (IV, Fixed, 95% CI)	Totals not selected
3.20 Participant satisfaction (dichotomous variable) Show forest plot	15		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only

3.20.1 Calcium dobesilate	4	758	Risk Ratio (M‐H, Random, 95% CI)	0.85 [0.61, 1.19]
3.20.2 Centella asiatica	1	80	Risk Ratio (M‐H, Random, 95% CI)	0.28 [0.14, 0.57]
3.20.3 Diosmine, Hidrosmine	4	451	Risk Ratio (M‐H, Random, 95% CI)	0.75 [0.59, 0.96]
3.20.4 Rutosides	6	1000	Risk Ratio (M‐H, Random, 95% CI)	0.50 [0.26, 0.97]
3.21 Participant satisfaction (continuous variable) Show forest plot	7		Std. Mean Difference (IV, Random, 95% CI)	Subtotals only

3.21.1 Calcium dobesilate	2	448	Std. Mean Difference (IV, Random, 95% CI)	‐0.52 [‐0.71, ‐0.33]
3.21.2 Diosmine, Hidrosmine	1	150	Std. Mean Difference (IV, Random, 95% CI)	‐0.81 [‐1.14, ‐0.47]
3.21.3 Rutosides	4	283	Std. Mean Difference (IV, Random, 95% CI)	‐1.18 [‐1.96, ‐0.39]
3.22 Adverse events Show forest plot	34	4830	Risk Ratio (M‐H, Fixed, 95% CI)	1.13 [0.99, 1.29]

3.22.1 Aminaftone	1	79	Risk Ratio (M‐H, Fixed, 95% CI)	0.60 [0.06, 6.32]
3.22.2 Calcium dobesilate	8	1824	Risk Ratio (M‐H, Fixed, 95% CI)	1.22 [1.00, 1.49]
3.22.3 Centella asiatica	1	94	Risk Ratio (M‐H, Fixed, 95% CI)	1.14 [0.58, 2.23]
3.22.4 Diosmine, Hidrosmine	9	1429	Risk Ratio (M‐H, Fixed, 95% CI)	0.93 [0.72, 1.19]
3.22.5 Grape seed extract	1	75	Risk Ratio (M‐H, Fixed, 95% CI)	0.57 [0.19, 1.74]
3.22.6 Rutosides	14	1329	Risk Ratio (M‐H, Fixed, 95% CI)	1.34 [1.02, 1.76]

Comparison 3. Sensitivity analysis of published studies only

Comparison 4. Sensitivity analysis based on low risk of bias

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
4.1 Oedema in the lower legs (dichotomous variable) Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

4.1.1 Calcium dobesilate	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
4.2 Ankle perimeter circumference (mm) Show forest plot	3	867	Mean Difference (IV, Random, 95% CI)	‐2.34 [‐8.79, 4.11]

4.2.1 Calcium dobesilate	3	867	Mean Difference (IV, Random, 95% CI)	‐2.34 [‐8.79, 4.11]
4.3 Quality of life Show forest plot	2	617	Mean Difference (IV, Fixed, 95% CI)	‐0.60 [‐2.15, 0.95]

4.3.1 Calcium dobesilate at 3 months of treatment	2	617	Mean Difference (IV, Fixed, 95% CI)	‐0.60 [‐2.15, 0.95]
4.4 Pain in the lower legs (dichotomous variable) Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

4.4.1 Rutosides	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
4.5 Pain in the lower legs (continuous variable) Show forest plot	2		Mean Difference (IV, Random, 95% CI)	Totals not selected

4.5.1 Calcium dobesilate	2		Mean Difference (IV, Random, 95% CI)	Totals not selected
4.6 Cramps in the lower legs (continuous variable) Show forest plot	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected

4.6.1 Calcium dobesilate	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected
4.7 Itching in the lower legs (dichotomous variable) Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

4.7.1 Rutosides	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
4.8 Itching in the lower legs (continuous variable) Show forest plot	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected

4.8.1 Calcium dobesilate	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected
4.9 Heaviness in the lower legs (dichotomous variable) Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

4.9.1 Rutosides	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
4.10 Heaviness in the lower legs (continuous variable) Show forest plot	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected

4.10.1 Calcium dobesilate	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected
4.11 Swelling in the lower legs (dichotomous variable) Show forest plot	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected

4.11.1 Rutosides	1		Risk Ratio (M‐H, Fixed, 95% CI)	Totals not selected
4.12 Swelling in the lower legs (continuous variable) Show forest plot	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected

4.12.1 Calcium dobesilate	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected
4.13 Participant satisfaction (dichotomous variable) Show forest plot	2	476	Risk Ratio (M‐H, Fixed, 95% CI)	1.04 [0.81, 1.32]

4.13.1 Calcium dobesilate	2	476	Risk Ratio (M‐H, Fixed, 95% CI)	1.04 [0.81, 1.32]
4.14 Participant satisfaction (continuous variable) Show forest plot	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected

4.14.1 Calcium dobesilate	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected
4.15 Adverse events Show forest plot	4	1257	Risk Ratio (M‐H, Random, 95% CI)	1.59 [0.97, 2.63]

4.15.1 Calcium dobesilate	3	1026	Risk Ratio (M‐H, Random, 95% CI)	1.53 [0.76, 3.09]
4.15.2 Rutosides	1	231	Risk Ratio (M‐H, Random, 95% CI)	1.83 [1.00, 3.34]

Comparison 4. Sensitivity analysis based on low risk of bias