Foam dressings for venous leg ulcers
Abstract
Background
Venous leg ulcers are a common and recurring type of chronic or complex wound that are associated with considerable cost to patients and to healthcare providers. Primary wound contact dressings are usually applied beneath compression devices with the aim of aiding healing. Foam dressings are used frequently, and a variety of foam products is available on the market. The evidence base to guide dressing choice, however, is sparse.
Objectives
To determine the effects of foam dressings on the healing of venous leg ulcers.
Search methods
In October 2012 we searched The Cochrane Wounds Group Specialised Register; the Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library); the Database of Abstracts of Reviews of Effects (DARE) (The Cochrane Library); the Economic Evaluation Database (The Cochrane Library); Ovid MEDLINE; Ovid MEDLINE (In‐Process & Other Non‐Indexed Citations); Ovid EMBASE; EBSCO CINAHL. There were no restrictions based on language or date of publication.
Selection criteria
We included published or unpublished randomised controlled trials (RCTs) that evaluated the effects of any type of foam dressing in the treatment of venous ulcers.
Data collection and analysis
Two review authors independently performed study selection, data extraction and risk of bias assessment. Meta‐analysis was undertaken when deemed feasible and appropriate.
Main results
Twelve RCTs (1023 participants) reporting 14 comparisons were included in this review. There was no difference in healing outcomes between hydrocellular foam dressings and polyurethane foam dressings (three RCTs). Pooled data across five RCTs (418 participants) showed no statistically significant difference between foam dressings and hydrocolloid dressings in the proportion of ulcers healed at 12 to 16 weeks (risk ratio (RR) 1.00, 95% confidence interval (CI) 0.81 to 1.22). No statistically significant between‐group differences in healing outcomes were detected when foam dressings were compared with: paraffin gauze (two RCTs); hydrocapillary dressing (one RCT); knitted viscose dressing (one RCT); and protease modulating matrix (one RCT). No statistically significant between‐group differences in the proportion of participants experiencing adverse events were detected when hydrocellular foam dressings were compared with polyurethane foam dressings, or when foam dressings were compared with hydrocapillary, hydrocolloid, or knitted viscose dressings (one RCT for each comparison). Six RCTs were considered as being at overall high risk of bias, and the remaining six RCTs were considered to be at overall unclear risk of bias. No included RCT had an overall low risk of bias.
Authors' conclusions
The current evidence base does not suggest that foam dressings are more effective in the healing of venous leg ulcers than other wound dressing treatments. The evidence in this area is of low quality. Further evidence is required from well‐designed and rigorously‐conducted RCTs, that employ methods to minimise bias and report them clearly, before any definitive conclusions can be made regarding the efficacy of foam dressings in the management of venous leg ulcers.
Plain language summary
Foam dressings for venous leg ulcers
Venous leg ulcers are a common and recurring type of chronic wound. Compression therapy (bandages or stockings) is used to treat venous leg ulcers. Dressings that aim to protect the wound and provide a moist environment to aid ulcer healing are applied beneath compression devices. Foam dressings are one of several types of dressing available. We evaluated the evidence from 12 randomised controlled trials that either compared different types of foam dressings, or compared foam dressings with other types of wound dressings. We found no evidence to suggest that polyurethane foam dressings are significantly better or worse than hydrocellular foam dressings in venous leg ulcer healing. Similarly, we found no evidence to suggest that foam dressings are significantly better or worse than other types of dressings (paraffin‐impregnated gauze dressings, hydrocapillary dressings, hydrocolloid dressings, knitted viscose dressings, or protease‐modulating matrix dressings), for the healing of venous leg ulcers. We found insufficient evidence to draw any conclusions regarding: adverse events, quality of life, costs, pain, or dressing performance. Overall, the current evidence is of low or unclear methodological quality. This limits the making of any specific recommendations regarding the use of foam dressings. Further, good quality evidence is required before definitive conclusions can be made regarding the role of foam dressings in the management of venous leg ulcers.
Authors' conclusions
Summary of findings
| polyurethane foam dressing compared to hydrocellular foam dressing for venous leg ulceration | ||||||
| Patient or population: patients with venous leg ulceration | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No of Participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| Hydrocellular foam dressing | Polyurethane foam dressing | |||||
| Time to healing | Study population1 | RR 1.24 | 156 | ⊕⊕⊝⊝ | ||
| 617 per 1000 | 765 per 1000 | |||||
| Low1 | ||||||
| 465 per 1000 | 577 per 1000 | |||||
| High1 | ||||||
| 757 per 1000 | 939 per 1000 | |||||
| Proportion of participants with healed ulcers | Study population1 | RR 1.03 | 118 | ⊕⊕⊝⊝ | ||
| 300 per 1000 | 309 per 1000 | |||||
| Low1 | ||||||
| 198 per 1000 | 204 per 1000 | |||||
| High1 | ||||||
| 507 per 1000 | 522 per 1000 | |||||
| Proportion of participants with healed ulcers | Study population1 | RR 1.60 | 18 | ⊕⊕⊝⊝ | ||
| 500 per 1000 | 800 per 1000 | |||||
| Low1 | ||||||
| 465 per 1000 | 744 per 1000 | |||||
| High1 | ||||||
| 757 per 1000 | 1000 per 1000 | |||||
| Proportion of participants with healed ulcers | Study population1 | RR 1.08 | 156 | ⊕⊕⊝⊝ | ||
| 617 per 1000 | 667 per 1000 | |||||
| Low1 | ||||||
| 465 per 1000 | 502 per 1000 | |||||
| High1 | ||||||
| 757 per 1000 | 818 per 1000 | |||||
| Mean change in wound size, with adjustment for baseline size | See comment | See comment | Not estimable | 0 | See comment | Outcome not reported (2 RCTs reported mean change in ulcer area, with no variance estimate, and no adjustment for baseline area).<BR/> |
| Proportion of participants experiencing adverse events | 272 per 10008 | 307 per 1000 | RR 1.13 | 156 | ⊕⊕⊝⊝ | |
| Health‐related quality of life | See comment | See comment | Not estimable | 0 | See comment | Outcome not reported. |
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (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). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1 Note: lower risk of the outcome is less favourable (i.e. lower risk of healing) than higher risk. Estimates for baseline low and high risks of healing at 60 days (8 weeks) and 120 days (17 weeks) have been taken from a meta‐analysis of RCTs evaluating different types of compression. The low risk estimate is based on a subset of participants with larger baseline ulcer area (greater than 5 cm squared). The high risk estimate is based on a subset of participants with smaller baseline ulcer surface area (5 cm squared or smaller). Most participants received a simple, low‐adherent dressing plus four‐layer bandage (O'Meara 2007). | ||||||
| foam dressing compared to paraffin gauze dressing for venous leg ulceration | ||||||
| Patient or population: patients with venous leg ulceration | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No of Participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| Paraffin gauze dressing | Foam dressing | |||||
| Time to healing | See comment | See comment | Not estimable | 0 | See comment | Outcome not reported. |
| Proportion of participants with healed ulcers | Study population1 | RR 1.34 | 71 | ⊕⊕⊝⊝ | ||
| 229 per 1000 | 306 per 1000 | |||||
| Low1 | ||||||
| 465 per 1000 | 623 per 1000 | |||||
| High1 | ||||||
| 757 per 1000 | 1000 per 1000 | |||||
| Mean change in wound size, with adjustment for baseline size | See comment | See comment | Not estimable | 0 | See comment | Outcome not reported (2 RCTs reported change in mean or median ulcer area, with no adjustment for baseline area).<BR/> |
| Adverse events | See comment | See comment | Not estimable | 0 | See comment | Limited data reported. |
| Health‐related quality of life | See comment | See comment | Not estimable | 0 | See comment | Outcome not reported. |
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (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). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1 Note: lower risk of the outcome is less favourable (i.e. lower risk of healing) than higher risk. Estimates for baseline low and high risks of healing at 120 days (17 weeks) have been taken from a meta‐analysis of RCTs evaluating different types of compression. The low risk estimate is based on a subset of participants with larger baseline ulcer area (greater than 5 cm squared). The high risk estimate is based on a subset of participants with smaller baseline ulcer surface area (5 cm squared or smaller). Most participants received a simple, low‐adherent dressing plus four‐layer bandage (O'Meara 2007). | ||||||
| foam dressing compared to hydrocapillary dressing for venous leg ulceration | ||||||
| Patient or population: patients with venous leg ulceration | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No of Participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| Hydrocapillary dressing | Foam dressing | |||||
| Time to healing | See comment | See comment | Not estimable | 0 | See comment | Limited information reported. |
| Proportion of participants with healed ulcers | Study population1 | RR 0.78 | 97 | ⊕⊕⊝⊝ | ||
| 510 per 1000 | 398 per 1000 | |||||
| Low1 | ||||||
| 465 per 1000 | 363 per 1000 | |||||
| High1 | ||||||
| 757 per 1000 | 590 per 1000 | |||||
| Mean change in wound size, with adjustment for baseline size | See comment | See comment | Not estimable | 0 | See comment | Outcome not reported. |
| Proportion of participants experiencing adverse events | 163 per 10006 | 229 per 1000 | RR 1.40 | 97 | ⊕⊕⊝⊝ | |
| Health‐related quality of life | See comment | See comment | Not estimable | 0 | See comment | Limited information provided. |
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (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). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1 Note: lower risk of the outcome is less favourable (i.e. lower risk of healing) than higher risk. Estimates for baseline low and high risks of healing at 120 days (17 weeks) have been taken from a meta‐analysis of RCTs evaluating different types of compression. The low risk estimate is based on a subset of participants with larger baseline ulcer area (greater than 5 cm squared). The high risk estimate is based on a subset of participants with smaller baseline ulcer surface area (5 cm squared or smaller). Most participants received a simple, low‐adherent dressing plus four‐layer bandage (O'Meara 2007). | ||||||
| foam dressing compared to hydrocolloid dressing for venous leg ulceration | ||||||
| Patient or population: patients with venous leg ulceration | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No of Participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| Hydrocolloid dressing | Foam dressing | |||||
| Time to healing | See comment | See comment | Not estimable | 0 | See comment | Limited data reported. |
| Proportion of participants with healed ulcers | Study population1 | RR 1.0 | 418 | ⊕⊕⊕⊝ | ||
| 471 per 1000 | 471 per 1000 | |||||
| Low1 | ||||||
| 465 per 1000 | 465 per 1000 | |||||
| High1 | ||||||
| 757 per 1000 | 757 per 1000 | |||||
| Mean change in wound size, with adjustment for baseline size | See comment | See comment | Not estimable | 0 | See comment | Two RCTs reported change in wound area, but not with baseline adjustment.<BR/> |
| Proportion of participants experiencing adverse events | 236 per 10004 | 288 per 1000 | RR 1.22 | 107 | ⊕⊕⊝⊝ | |
| Health‐related quality of life | See comment | See comment | Not estimable | 0 | See comment | Outcome not reported. |
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (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). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1 Note: lower risk of the outcome is less favourable (i.e. lower risk of healing) than higher risk. Estimates for baseline low and high risks of healing at 120 days (17 weeks) have been taken from a meta‐analysis of RCTs evaluating different types of compression. The low risk estimate is based on a subset of participants with larger baseline ulcer area (greater than 5 cm squared). The high risk estimate is based on a subset of participants with smaller baseline ulcer surface area (5 cm squared or smaller). Most participants received a simple, low‐adherent dressing plus four‐layer bandage (O'Meara 2007). | ||||||
| foam dressing compared to knitted viscose dressing for venous leg ulceration | ||||||
| Patient or population: patients with venous leg ulceration | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No of Participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| Knitted viscose dressing | Foam dressing | |||||
| Time to healing | See comment | See comment | Not estimable | 0 | See comment | Limited information provided. |
| Proportion of participants with healed ulcers | Study population1 | RR 1.35 | 132 | ⊕⊕⊝⊝ | ||
| 348 per 1000 | 470 per 1000 | |||||
| Low1 | ||||||
| 311 per 1000 | 420 per 1000 | |||||
| High1 | ||||||
| 696 per 1000 | 940 per 1000 | |||||
| Mean change in wound size, with adjustment for baseline size | See comment | See comment | Not estimable | 0 | See comment | Limited information provided. |
| Proportion of participants experiencing adverse events | RR 5.00 | 132 | ⊕⊕⊝⊝ | Assumed and corresponding risks not estimated because of zero event rate in knitted viscose group (3% in foam dressing group). | ||
| Health‐related quality of life | See comment | See comment | Not estimable | 0 | See comment | Outcome not reported. |
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (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). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1 Note: lower risk of the outcome is less favourable (i.e. lower risk of healing) than higher risk. Estimates for baseline low and high risks of healing at 90 days (12‐13 weeks) have been taken from a meta‐analysis of RCTs evaluating different types of compression. The low risk estimate is based on a subset of participants with larger baseline ulcer area (greater than 5 cm squared). The high risk estimate is based on a subset of participants with smaller baseline ulcer surface area (5 cm squared or smaller). Most participants received a simple, low‐adherent dressing plus four‐layer bandage (O'Meara 2007). | ||||||
Background
For definitions of terminology see Glossary of terms (Appendix 1).
Description of the condition
Venous leg ulcers are a common and recurring type of chronic, or complex, wound. They are usually caused by venous insufficiency (impaired venous blood flow) brought about by venous hypertension. Predisposing factors for venous hypertension include history of deep vein thrombosis (DVT), thrombophlebitis, leg trauma, arthritis, obesity, pregnancy and a sedentary lifestyle. These factors can result in damage to the valves in the leg veins allowing pathological (disease‐causing) two‐way blood flow instead of the normal one‐way movement. A related issue is diminished calf muscle pump action. Both valvular and calf‐muscle pump impairment can result in reduced venous blood flow leading to venous hypertension. This causes distension of the leg veins, oedema of the lower limb and leakage of circulatory fluids from the capillaries into the surrounding tissues. This, in turn, induces irritation and increased fragility of the epidermis (the outer layer of skin) leading to ulceration (Doughty 2007). The duration of venous leg ulceration ranges from a matter of weeks to more than 10 years and some people never heal (Moffatt 1995; Ruckley 1998; Vowden 2009a). Older patient age, longer wound duration and larger ulcer surface area have been reported as independent risk factors for delayed ulcer healing (Margolis 2004; Gohel 2005).
A review of 11 venous leg ulceration prevalence studies conducted in Australia and Europe estimated point prevalence as 0.1% to 0.3% (Nelzen 2008). Surveys undertaken in the UK estimated prevalence of venous leg ulceration as 0.023% in Wandsworth, London (Moffatt 2004); 0.044% in Hull and East Yorkshire (Srinivasaiah 2007); and 0.039% in Bradford and Airedale (Vowden 2009a; Vowden 2009b). The lower estimates reported in the UK surveys relative to the worldwide literature might be explained by differences in disease management or case definition, or both. We were unable to identify contemporary prevalence data for non‐western countries. The epidemiological data have consistently suggested that prevalence increases with age, and is higher among women (Margolis 2002; Graham 2003; Lorimer 2003; Moffatt 2004; Vowden 2009a).
Diagnosis of venous leg ulceration can be made according to the appearance and location of the ulcer. Clinical practice guidelines recommend the use of clinical history, physical examination, laboratory tests and haemodynamic assessment (RCN 2006; SIGN 2011a). The latter typically includes an assessment of arterial supply to the leg using the ankle‐brachial pressure index (ABPI), measured using a hand‐held Doppler ultrasound device. An ABPI measurement of more than 0.8 is generally used to rule out the co‐existence of clinically significant peripheral arterial disease in a leg ulcer that has been diagnosed as due to venous insufficiency (Moffatt 2007).
Leg ulcers are associated with considerable cost to patients and to healthcare providers. Two systematic reviews summarised the literature on health‐related quality of life in patients with leg ulcers (Persoon 2004; Herber 2007). Both included qualitative and quantitative evaluations and reported that the presence of leg ulceration was associated with pain, restriction of work and leisure activities, impaired mobility, sleep disturbance, reduced psychological well‐being and social isolation.
The cost of treating an unhealed leg ulcer in the UK has been estimated to be around GBP 1300 per year at 2001 prices (Iglesias 2004). Another evaluation estimated the average cost of treating a venous leg ulcer in the UK (based on costs for material for dressing changes) as lying between EUR 814 and EUR 1994, and, in Sweden, as lying between EUR 1332 and EUR 2585 (price year 2002), with higher costs associated with larger and more chronic wounds (Ragnarson Tennvall 2005). This reflected findings from a more recent evaluation conducted in Hamburg, Germany, recruiting 502 community based adult patients with any type of leg ulcer. The total mean annual cost of illness for leg ulcers was estimated as EUR 9060 per patient (price year 2006), taking account of direct, indirect and intangible costs from a societal perspective. Direct costs included all expenses directly related to leg ulcer care (dressings, bandages, topical agents, systemic treatment, diagnostic procedures, clinician fees, in‐patient treatment costs and transport); indirect costs related to loss of productivity; and intangible costs included impact on health‐related quality of life. Estimates ranged from zero cost (i.e. no treatment) to EUR 44,462, with higher costs associated with arterial aetiology of the ulcer, larger wound size and no history of wound closure (Augustin 2012). In Bradford, in the UK, GBP 1.69 million was spent on dressings and compression bandages, and GBP 3.08 million on nursing time (estimates derived from resource use data for all wound types, not just venous leg ulcers) during the financial year 2006‐2007 (Vowden 2009c). We were not able to identify further contemporary international cost data.
Compression therapy (bandages or stockings) is now considered to be the cornerstone of venous leg ulcer management (Moffatt 2007; O'Meara 2009). Primary wound contact dressings (i.e. dressings in direct contact with the wound bed) are usually applied underneath compression devices. A range of other interventions may be used concurrently with compression, including debriding agents (Davies 2005), vasoactive drugs (Robson 2006), fibrinolytic therapy (Robson 2006), physical therapies (Flemming 2001; Ravaghi 2006; Al‐Kurdi 2008), and topical applications (Robson 2006).
Description of the intervention
Primary wound contact dressings are applied with the aim of aiding healing, providing comfort, controlling exudate (the fluid produced by wounds) and helping to prevent bandages and stockings from adhering to the wound bed. The ideal conditions required for wound healing in terms of dressing application have been explained as follows: maintenance of a moist wound environment without risk of maceration (excessive softening of skin because of being constantly wet); avoidance of toxic chemicals, particles or fibres in the dressing fabric; a minimal number of dressing changes; and maintenance of an optimum pH level (balanced acidity and alkalinity) (BNF 2013).
Several types of wound dressing are available and costs vary (Appendix 2). For example, there can be a six‐fold difference in the UK unit price of a 9.5 cm x 9.5 cm non‐adherent (knitted viscose) dressing compared with a 10 cm x 10 cm polyurethane foam dressing (BNF 2013).
Foam dressings were one of the first advanced (modern) dressings to be used in wound management and are available internationally. In the UK, foam dressings are one of a variety of advanced wound dressings that are currently available, and are used frequently, being common to many wound‐care formularies. One of the purported advantages of foam dressings is their exudate handling properties, not only through absorption of exudate into the fabric of the dressing, but also via transmission through a top sheet, keeping moisture away from the wound margin (and thereby potentially reducing the risk of maceration of peri‐ulcer skin). If necessary, foam dressings can be used in combination with a secondary dressing, with the latter absorbing excess exudate. It has been suggested that foam dressings can remain in place for longer than other dressing types, which provides potential to decrease the frequency of dressing changes and, therefore, reduce costs. Another suggested benefit is non‐traumatic dressing removal (Sussman 2010). There is a variety of foam products on the market, and these vary in their ability to absorb exudate. Some are suitable for only lightly‐ to moderately‐exuding wounds, whilst others have a greater fluid‐handling capacity.
Several potential disadvantages of foam dressings have been noted, namely, the possibility of saturation leading to maceration of healthy peri‐ulcer skin and also potential reduction in fluid‐handling capacity if used beneath compression devices.
Examples of foam dressings currently available in the UK include Allevyn® Non‐Adhesive (Smith and Nephew) and Biatain® Non‐Adhesive (Coloplast). Appendix 2 provides a description of all wound dressings categorised by the British National Formulary (BNF 2013).
How the intervention might work
Findings from research based on animal models suggest that acute wounds heal more quickly when the wound surface is kept moist, so that formation of a hard scab or eschar is prevented (Winter 1963). A moist environment is also thought to provide optimal conditions for promoting autolytic debridement (the breakdown of dead cells lying on top of the wound bed), which is sometimes considered to be an important part of the healing pathway (König 2005). It is purported that foam dressings manage exudate, provide a moist wound healing environment and promote healing (BNF 2013).
Why it is important to do this review
Wound dressings are a key part of the treatment pathway when caring for venous leg ulcers. Most will be used in combination with compression systems, and guidelines are necessary to help make decisions regarding the value, and best use, of the available dressings. Several types of wound dressing are available and costs vary considerably, however, the evidence base to guide dressing choice is sparse. A previous Cochrane review that evaluated different wound dressings for venous leg ulcers concluded that the type of dressing applied beneath compression had not been shown to affect ulcer healing (Palfreyman 2006). The review authors concluded that there was no evidence of benefit for foam dressings compared with non‐adherent dressings, nor for different types of foam dressings compared with one another.
This review will now update part of the previous Palfreyman 2006 Cochrane review and will be one of several Cochrane reviews investigating the use of dressings in the treatment of venous leg ulcers. Each review will focus on a particular dressing type, which, in this review, will be foam dressings. These reviews will eventually be summarised in an overview of reviews (Becker 2011), which will draw together all existing Cochrane review evidence regarding the use of dressings in the treatment of venous leg ulcers.
Objectives
To determine the effects of foam dressings compared with alternative dressings, non‐dressing treatments or no dressing, with or without concurrent compression therapy, on the healing of venous leg ulcers; and to determine the comparative effects of foam dressings with alternatives on health‐related quality of life, costs, pain, dressing performance (management of wound exudate and ease of removal) and adverse effects.
Methods
Criteria for considering studies for this review
Types of studies
We included randomised controlled trials (RCTs), either published or unpublished, that evaluated the effects of any type of foam dressing in the treatment of venous leg ulcers, irrespective of publication status or language. RCTs reported in abstract form only were eligible for inclusion, provided adequate information was presented in the abstract, or was available from the authors. Studies using quasi‐randomisation were excluded.
Types of participants
RCTs recruiting people described in the primary report as having venous leg ulcers, managed in any setting, were eligible for inclusion. As the method of diagnosis of venous ulceration may vary, we accepted definitions as used in the RCTs. We included RCTs that recruited samples comprising people with venous leg ulcers and people with other types of wounds (e.g. arterial ulcers, diabetic foot ulcers) if the results for people with venous ulcers were presented separately (or separate data available from the authors), or if the majority of participants (75% or more in each arm) had leg ulcers of venous aetiology (origin).
Types of interventions
The primary intervention of interest was foam wound dressings. For ease of comparison we categorised dressings according to the British National Formulary (BNF 2013). We have presented generic names where possible, also providing trade names and manufacturers where available. It is important, however, to note that manufacturers and distributors of dressings may vary from country to country, and dressing names may also differ. We did not include RCTs evaluating foam dressings impregnated with antimicrobial, antiseptic or analgesic agents, as these interventions are evaluated in other Cochrane reviews (Briggs 2012; O'Meara 2010), and will be captured in the proposed overview of reviews (see Why it is important to do this review). RCTs evaluating wound dressing pads, hydrocolloid dressings, hydrogels and alginate dressings will be covered in other, separate Cochrane reviews and are included in this review only if they are comparators to foam dressings. Forthcoming, related Cochrane reviews on dressings for venous leg ulcers will, along with this review, update the review by Palfreyman 2006.
We included any RCT in which the presence or absence of a specific foam dressing was the only systematic difference between treatment groups; and in which a foam dressing was compared with other wound dressings (including alternative foam dressings), non‐dressing treatments (for example, topical applications) or no dressing. We included RCTs of foam dressings, irrespective of whether compression therapy was reported as a concurrent treatment.
Types of outcome measures
Primary outcomes
The primary outcome for the review was complete wound healing.
Wound healing is measured and reported by trialists in many different ways, including time to complete wound healing, the proportion of wounds healed during follow up and rates of change of wound size. For this review we regarded RCTs that reported one or more of the following as providing the best measures of outcome in terms of relevance and rigour:
-
time to complete wound healing (correctly analysed using survival, time‐to‐event approaches, ideally with adjustment for relevant covariates such as baseline size);
-
the proportion of ulcers healed during follow up (frequency of complete healing);
-
and change (and rate of change) in wound size, with adjustment for baseline size.
We considered evidence from RCTs that reported mean or median time to healing without survival analysis (i.e. they regarded time to healing as a continuous measure without censoring), and those that measured and reported change or rate of change in wound size without adjustment for baseline size, as less rigorous assessments of these outcomes, and did not use data reported in this manner to populate the Summary of Findings tables (see Summary of Findings tables).
Secondary outcomes
The secondary outcomes for the review were:
-
rates of all reported adverse events (e.g. infection, eczema, maceration);
-
health‐related quality of life (measured using a validated standardised generic questionnaire such as EQ‐5D, SF‐36, SF‐12 or SF‐6, or validated disease‐specific questionnaire) preferably with follow‐up estimates adjusted for baseline scores;
-
cost (including cost or cost‐effectiveness estimations, as well as measurements of resource use such as number of dressing changes, dressing wear time and nurse time);
-
pain (e.g. at dressing change, between dressing changes or over the course of treatment);
-
dressing performance (exudate management and ease of removal or adherence to the wound bed).
Search methods for identification of studies
Electronic searches
In October 2012 we searched the following electronic databases for potentially relevant RCTs:
-
the Cochrane Wounds Group Specialised Register (searched 25 October 2012);
-
the Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library 2012, Issue 10);
-
the Database of Abstracts of Reviews of Effects (DARE) (The Cochrane Library 2012, Issue 10);
-
the Economic Evaluation Database (The Cochrane Library 2012, Issue 10);
-
Ovid MEDLINE (1948 to October Week 3 2012);
-
Ovid MEDLINE (In‐Process & Other Non‐Indexed Citations, 25 October 2012);
-
Ovid EMBASE (1974 to 2012 Week 42);
-
EBSCO CINAHL (1982 to 19 October 2012).
We used the following search strategy in the Cochrane Central Register of Controlled Trials (CENTRAL):
#1 MeSH descriptor: [Occlusive Dressings] explode all trees 436
#2 foam*:ti,kw,ab 929
#3 (ActivHeal or Allevyn or Avazorb or Biatain or Copa or LyoFoam or PermaFoam or PolyMem or Suprasorb or Tegaderm or Tielle or Transobent or Trufoam or UrgoCell):ti,kw,ab 171
#4 #1 or #2 or #3 1465
#5 MeSH descriptor: [Leg Ulcer] explode all trees1064
#6 ((varicose next ulcer*) or (venous next ulcer*) or (leg next ulcer*) or (stasis next ulcer*) or (crural next ulcer*) or "ulcus cruris" or "ulcer cruris"):ti,ab,kw 1408
#7 #5 or #6 1830
#8 #4 and #7 126
The search strategies for Ovid MEDLINE, Ovid EMBASE and EBSCO CINAHL can be found in Appendix 3. We combined the Ovid MEDLINE search with the Cochrane Highly Sensitive Search Strategy for identifying randomised trials in MEDLINE: sensitivity‐ and precision‐maximising version (2008 revision) (Lefebvre 2011). We combined the EMBASE search with the Ovid EMBASE filter developed by the UK Cochrane Centre (Lefebvre 2011). We combined the CINAHL searches with the RCT filters developed by the Scottish Intercollegiate Guidelines Network (SIGN 2011b). We did not restrict searches with respect to language or date of publication.
We searched for ongoing RCTs in the World Health Organization International Trial Registry Platform (http://www.who.int/ictrp/en/ accessed 1 August 2012) and the ISRCTN (International Standard Randomised Controlled Trial Number) register (http://www.controlled‐trials.com/isrctn/ accessed 1 August 2012) using the search term 'leg ulcer'.
Searching other resources
We attempted to contact trialists to obtain unpublished data and information as required, along with manufacturers to request information about ongoing or as yet unpublished RCTs (for a list of manufacturers see Appendix 4). We also searched the reference lists of RCT reports identified for inclusion and of other systematic review articles.
Data collection and analysis
Selection of studies
Two review authors independently assessed the titles and abstracts retrieved by the searches for relevance. After this initial assessment, we obtained full text copies of all RCT reports felt to be potentially relevant. Two review authors then independently checked the full papers for eligibility, with disagreements resolved by discussion, and, where required, referral to the editorial base of the Cochrane Wounds Group. We recorded all reasons for exclusion.
We have presented our study selection process as a PRISMA (Preferred Reporting Items for Systematic Reviews and Meta‐Analyses) flow diagram (Liberati 2009) (see Results of the search).
Data extraction and management
We extracted and summarised details of the eligible RCTs using a data extraction sheet. We extracted the data from RCT reports using an Excel spreadsheet designed to capture the information detailed below. Initially, we piloted the spreadsheet with a sample of eligible RCTs, to explore any issues that might arise in relation to the data extraction process. We expanded and amended the spreadsheet as necessary after the piloting process. Two review authors performed independent data extraction of all included RCTs, after which both data extractions were compared, and a final version agreed. We resolved any disagreements by discussion. If data were missing from reports we attempted to contact the trial authors to obtain the missing information. We included RCTs published as duplicate reports (parallel publications) once, using all associated reports to extract a maximal amount of information, but ensuring that data were not duplicated in the review. We extracted the following information:
-
trial authors;
-
year of publication;
-
country where RCT performed;
-
setting of care;
-
unit of investigation – participant, leg or ulcer;
-
overall sample size and methods used to estimate statistical power (relates to the target number of participants to be recruited, the clinical difference to be detected and the ability of the RCT to detect this difference);
-
participant selection criteria;
-
number of participants randomised to each treatment arm;
-
baseline characteristics of participants per treatment arm (gender, age, baseline ulcer area, ulcer duration, prevalence of co‐morbidities such as diabetes, prevalence of clinically infected wounds or colonised wounds, previous history of ulceration, baseline levels of wound exudate, and participant mobility);
-
details of the dressing/treatment regimen prescribed for each arm including details of concomitant therapy, for example, compression;
-
duration of treatment;
-
duration of follow up;
-
statistical methods used for data analysis;
-
primary and secondary outcomes measured;
-
primary and secondary outcome data by treatment arm;
-
adverse effects of treatment (per treatment arm with numbers and type);
-
withdrawals (per treatment arm with numbers and reasons); and
-
source of trial funding.
Assessment of risk of bias in included studies
Two review authors independently assessed each included RCT report using the Cochrane Collaboration tool for assessing risk of bias (Higgins 2011a). This tool addresses specific domains, namely: sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data and selective outcome reporting (see Appendix 5 for details of the criteria on which judgements were based). For blinded outcome assessment we made separate judgements for primary and secondary outcomes. As the primary outcomes for this review (wound healing), whichever way measured, are subject to potential observer or measurement bias, blinding of outcome assessment is important. Similarly we made separate judgements for primary and secondary outcomes for the domain of incomplete outcome data. In order to assess selective outcome reporting, we sought protocols for all included RCTs. Where protocols were unavailable, we made a judgement based on congruence of information in methods and results sections of reports of RCTs. We classified RCTs as being at overall high risk of bias if they were rated as 'high' for any one of three key domains (allocation concealment, blinding of outcome assessors and completeness of outcome data). We classified RCTs as being at overall unclear risk of bias if any one of the three key domains was rated as unclear. RCTs were judged to be at overall low risk of bias only if all three key domains were rated as low risk.
We have presented our assessment of risk of bias findings using 'Risk of bias' summary figures. Figure 1 is a summary of information across all included RCTs and Figure 2 shows a cross‐tabulation of each individual RCT with each risk of bias item. This display of internal validity indicates the weight the reader may give the results of each RCT.
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included trials.
Risk of bias summary: review authors' judgements about each risk of bias item for each included trial.
Measures of treatment effect
We have presented a narrative overview of all included RCTs, with results grouped according to the comparator intervention. We undertook statistical pooling of outcome data on groups of RCTs considered to be sufficiently similar in terms of design and characteristics of participants, interventions and outcomes, using Cochrane RevMan software (version 5.2) (RevMan 2012). We reported estimates for dichotomous outcomes (e.g. number of ulcers healed) as risk ratios (RRs) with associated 95% confidence intervals (CI). We reported estimates for continuous data outcomes (e.g. absolute or relative change in ulcer area and healing rate) as a mean difference (MD) with 95% CI. We planned to report estimates of time to healing and plot hazard ratios where available from the included RCTs. Where hazard ratios were not reported we planned, where possible, to extrapolate these using other reported data relating to time to healing (Parmar 1998). Where RCTs reported adverse events in sufficient detail (e.g. the number of participants who experienced at least one adverse event) we analysed these as dichotomous data. Where it was unclear whether the denominator was the total number of adverse events, or the number of participants, we reported these data narratively. Where adverse events were reported as dressing‐related we analysed the data separately.
Unit of analysis issues
We recorded whether RCT reports specified participants, limbs or ulcers as the units of allocation and analysis. In cases where multiple limbs or ulcers on the same individual were studied, we planned to note whether the trialists' analysis was appropriate (i.e. correctly taking account of highly correlated data) or inappropriate (i.e. considering outcomes for multiple ulcers on the same participant as independent). Where the number of wounds appeared to equal the number of participants, we assumed that the participant was the unit of analysis, unless otherwise stated.
Dealing with missing data
Missing data are a common problem in RCTs. Excluding randomised participants from the analysis or ignoring those participants lost to follow up can compromise the process of randomisation and introduce bias. Where RCTs reported dichotomous complete healing outcomes for only those participants who completed the RCT (i.e. participants withdrawing and lost to follow up were excluded from the analysis), we treated the participants who were not included in the analysis as if their wound did not heal (that is, they were included in the denominator but not the numerator for healing outcomes). Where results were reported for participants who completed the RCT without specifying the numbers that were randomised per group initially, we presented only complete case data. For other outcomes we presented data for all participants randomised, where reported; otherwise we based estimates on complete cases only.
Assessment of heterogeneity
We considered clinical heterogeneity (that is the degree to which RCTs appear similar in terms of participants, intervention type and duration and outcome type) and statistical heterogeneity. We assessed statistical heterogeneity using the Chi² test (P value less than 0.10 was considered to indicate statistically significant heterogeneity) in conjunction with the I² statistic (Higgins 2003). The I² statistic examines the percentage of total variation across RCTs due to heterogeneity rather than chance (Higgins 2003). We considered that I² values of 40% or less indicated a low level of heterogeneity and values of 75% or more represented very high heterogeneity (Deeks 2011).
Assessment of reporting biases
Reporting biases arise when the dissemination of research findings is influenced by the nature and direction of results. Publication bias is one of a number of possible causes of 'small study effects' – a tendency for estimates of the intervention effect to be more beneficial in smaller trials. Funnel plots allow a visual assessment of whether small study effects may be present in a meta‐analysis. A funnel plot is a simple scatter plot of the intervention effect estimates from individual RCTs against some measure of each trial’s size or precision (Sterne 2011). We planned to present funnel plots for meta‐analyses comprising 10 RCTs or more using RevMan 5.2.
Data synthesis
We have presented a narrative overview of the included RCTs. Where appropriate, we have presented meta‐analyses of outcome data using RevMan 5.2. The decision to pool data in a meta‐analysis depended upon the availability of outcome data and the assessment of between‐trial heterogeneity. For comparisons where there was no apparent clinical heterogeneity and the I² value was 40% or less, we applied a fixed‐effect model. Where there was no apparent clinical heterogeneity and the I² value was greater than 40%, we planned to apply a random‐effects model. However, we planned not to pool data where heterogeneity was very high (I² values of 75% or above).
For the dichotomous outcomes we have presented the summary estimate as a RR with 95% confidence intervals (CI). For outcomes reported as count data (i.e. where the denominator was not the number of participants), a between‐group difference with 95% CI was not estimated. Where continuous outcomes were measured in the same way across RCTs, we have presented a mean difference (MD) with 95% CI. Where variance data were not available, a mean between‐group difference with 95% CI was not estimated. We planned to present a standardised mean difference (SMD) when RCTs measured the same outcome using different methods. For time‐to‐event data, we planned to plot (and if appropriate pool) estimates of hazard ratios and 95% CIs as they were presented in the RCT reports using the generic inverse variance method in RevMan 5.2. Where hazard ratios were not reported, and could not be extrapolated, a between‐group difference with 95% CI was not estimated.
Summary of Findings tables
We have presented the main results of the review in Summary of Findings tables. These tables present key information concerning the quality of the evidence, the magnitude of the effects of the interventions examined, and the sum of the available data for the main outcomes (Schünemann 2011a). The Summary of Findings tables also include an overall grading of the evidence related to each of the main outcomes using the GRADE (Grades of Recommendation, Assessment, Development and Evaluation) approach. The GRADE approach defines the quality of a body of evidence with regard to the extent to which one can be confident that an estimate of effect or association is close to the true quantity of specific interest. Quality of a body of evidence involves consideration of within‐trial risk of bias (methodological quality), directness of evidence, heterogeneity, precision of effect estimates and risk of publication bias (Schünemann 2011b). We planned to present the following outcomes in the Summary of Findings tables:
-
time to complete ulcer healing where analysed using appropriate survival analysis methods;
-
proportion of ulcers completely healing during the trial period;
-
change in wound size, when adjusted for baseline size;
-
adverse events; and
-
health‐related quality of life.
Subgroup analysis and investigation of heterogeneity
We planned to conduct subgroup analyses according to whether RCTs assessed the dressing of interest with or without the application of compression therapy. We planned to exclude RCTs in which the presence or absence of compression therapy was not clearly indicated from this subgroup analysis.
Sensitivity analysis
Where data permitted, we planned to undertake sensitivity analyses according to risk of bias, excluding RCTs that were judged as being at overall high or unclear risk of bias (i.e. rated as high or unclear respectively for any one of three key domains ‐ allocation concealment, blinding of outcome assessors and completeness of outcome data). We also planned to undertake a sensitivity analysis in which we excluded RCTs that reported complete healing outcomes for only those participants who completed the RCT, and RCTs that reported results for participants who completed the RCT without specifying the numbers that were randomised per treatment group initially.
Results
Description of studies
See Characteristics of included studies; Characteristics of excluded studies; Characteristics of studies awaiting classification; Characteristics of ongoing studies.
Results of the search
The search strategy identified 436 records in total. Of these, 256 were retrieved from electronic bibliographic databases, 174 were found from registers of ongoing trials, four were identified from examination of reference lists, and two were obtained as unpublished reports whilst requesting trial protocols for the risk of bias assessment. No references were obtained as a result of contact with wound dressing manufacturers (Appendix 4). Eight manufacturers out of 14 contacted confirmed that there were no ongoing or recently completed RCTs of foam dressings; no replies were received from the remainder. Three hundred and fifty‐five records were excluded because of irrelevance on the basis of information in titles and abstracts or details of ongoing trials.
Eighty‐one full text reports were retrieved, comprising: 76 published reports; three records of ongoing trials; and two unpublished reports. Following assessment of full text reports against the review's study selection criteria, 43 full text articles were excluded that reported 30 unique studies. Reasons for exclusion were as follows:
-
11 were not RCTs (Cherry 1984; van Rijswijk 1985; Berard 1986; Pottle 1987; Moody 1991; Kurring 1994; Anonoymous 1997; Wollina 1997; Larsen 2005; Verdu Soriano 2006; Bressieux 2007);
-
seven did not evaluate a treatment of interest for this review (Moffatt 1992; Scurr 1993; Winter 1996; Capillas Pérez 2000; Sironi 2003; Price 2004; Meaume 2012);
-
six indicated that the foam dressing was not the only systematic difference between treatment groups (Rubin 1990; Pessenhofer 1992; Colgan 1996; Dmochowska 1999; Schulze 2001; Reynolds 2004);
-
four recruited a group of participants in which less than 75% had venous leg ulcers, or separate data for participants with venous leg ulcers were not available, or both (Bale 1998; Collier 1992; Banks 1997; Romanelli 1999); and
-
two were reported only as abstracts with limited information (Mulder 1995; Brown‐Etris 2004).
See Characteristics of excluded studies for further details.
Twelve RCTs (reported in 32 articles) were included in the review (see next section for further details). In addition, two RCTs (reported in three articles) were classified as awaiting assessment (Jørgensen 2008; Romanelli 2008), and three RCTs (reported in three articles) were ongoing (Vas 2008; Bayer 2009; Badiavas 2011). See Characteristics of studies awaiting classification and Characteristics of ongoing studies for further details. The study selection process is shown in Figure 3.
Flow diagram of the study selection process
Included studies
We included 12 RCTs that recruited 1023 participants (see Characteristics of included studies). One trial took place entirely in North America (Weiss 1996); another jointly in North America and Europe (Vanscheidt 2004); while the remaining RCTs took place in EU countries. Six were multicentred RCTs (Callam 1992; Thomas 1997; Andersen 2002; Vanscheidt 2004; Norkus 2005; Franks 2007). The remainder were conducted at single centres. Two RCTs were of a factorial design that allowed an evaluation of two dressings and two compression bandage systems (Callam 1992; Franks 2007); and another two were described as pilot studies (Weiss 1996; Andriessen 2009). The included RCTs were reported from 1990 to 2009.
Sample sizes ranged from 12 to 159 participants. Two RCTs reported that a sample size calculation had been performed (Vanscheidt 2004; Franks 2007).
The mean age of participants ranged from 55 to 78 years. The proportion of female participants ranged from 33% to 89%. Where reported, the majority of RCTs recruited participants who were ambulatory.
With the exception of two RCTs that recruited participants with leg ulcers of various aetiologies (Andersen 2002; Norkus 2005), all participants had leg ulcers of venous aetiology. An APBI less than 0.80 was an exclusion criterion in the majority of RCTs, confirmed, where reported, using hand‐held Doppler or digital photoplethysmography.
The dressing comparisons evaluated by the included RCTs were as follows:
-
three RCTs compared hydrocellular foam dressings with polyurethane foam dressings (Weiss 1996; Andersen 2002; Franks 2007);
-
two RCTs compared paraffin‐impregnated gauze dressings with foam dressings (Banerjee 1990; Andriessen 2009), one of which was a three‐arm RCT also evaluating a protease‐modulating matrix dressing (Andriessen 2009);
-
one RCT compared a hydrocapillary dressing with a foam dressing (Norkus 2005);
-
five RCTs compared hydrocolloid dressings with foam dressings (Zuccarelli 1992; Bowszyc 1995; Thomas 1997; Charles 2002; Vanscheidt 2004);
-
one RCT compared a knitted viscose dressing with a foam dressing (Callam 1992).
All RCTs reported the use of compression therapy for eligible participants, with one exception, where no details were provided (Banerjee 1990). The length of treatment ranged from four weeks to twelve months. Treatment settings were mainly leg ulcer clinics, hospital outpatients clinics and the community.
With one exception (Andriessen 2009), all RCTs reported the proportion of ulcers completely healed at the end of treatment. Time to complete ulcer healing was assessed by eight RCTs (Callam 1992; Bowszyc 1995; Weiss 1996; Andersen 2002; Charles 2002; Vanscheidt 2004; Norkus 2005; Franks 2007), and change in ulcer size or healing rate was assessed by eight RCTs (Banerjee 1990; Callam 1992; Weiss 1996; Thomas 1997; Andersen 2002; Charles 2002; Vanscheidt 2004; Andriessen 2009).
Risk of bias in included studies
A summary of the risk of bias assessment is presented in Figure 1 and Figure 2.
Allocation
Generation of the randomisation sequence
One RCT reported using a computer‐generated programme (Andriessen 2009), and another used minimisation (Charles 2002). These two RCTs were considered to be at low risk of bias for generation of the randomisation sequence. A judgement of low risk of bias for this domain was also made for Franks 2007 following personal communication with the trial authors in connection with a separate review. The remaining nine RCTs did not describe the method used for random sequence generation clearly and were classified as being at unclear risk of bias for this domain.
Concealment of the allocation process
One RCT reported using sequentially‐numbered, sealed, opaque envelopes (Andersen 2002), and was considered to be at low risk of bias. A judgement of low risk of bias for this domain was also made for Franks 2007 following personal communication with the RCT authors. Two RCTs reported the use of envelopes for treatment group allocation, but did not specify whether the envelopes were opaque and sequentially‐numbered and, therefore, were classified as being at unclear risk of bias (Andriessen 2009; Thomas 1997). The remaining eight RCTs did not clearly describe the method of treatment group allocation and were also classified as being at unclear risk of bias for this domain.
Blinding
Blinding of participants and personnel
Two RCTs provided detail in the report that the RCT was not blinded (Weiss 1996; Andersen 2002), and one trialist provided detail via personal communication that the RCT was not blinded (Franks 2007). These RCTs were judged to be at a high risk of bias for blinding of participants and personnel. One report stated that the RCT was ‘open’ but did not provide further details regarding whether or not the participants or personnel were blinded, and was classified as having an unclear risk of bias for this domain (Thomas 1997). The remainder of the included RCT reports (eight RCTs) did not provide statements regarding blinding of participants or personnel, and were also judged to be at unclear risk of bias.
Blinding of outcome assessment
One RCT reported that the assessors were blinded to the treatment, and was judged to be at low risk of bias for blinded outcome assessment (Andriessen 2009). Two RCTs reported that the RCT was not blinded, and were judged to be at a high risk of bias for this domain (Weiss 1996; Andersen 2002), as was another following personal communication with the trial authors (Franks 2007). One RCT was described as ‘open’ but did not state whether the outcome assessment was blinded, and was classified as unclear risk of bias (Thomas 1997). The remainder of the included RCT reports (seven RCTs) did not provide details about blinded outcome assessment, and were also judged to be at unclear risk of bias for this domain.
Incomplete outcome data
One RCT reported that all randomised participants completed the RCT (Andriessen 2009), and another reported that although dropouts had occurred (reasons reported and numbers balanced across groups), analyses had been according to intention‐to‐treat (ITT) (Franks 2007). One RCT reported that some participants withdrew from both treatment groups, but all participants randomised were accounted for in the tabulated results (Bowszyc 1995). One RCT report did not include a statement regarding participant withdrawal, however, it appeared that all participants were included in the final analysis (Zuccarelli 1992). These RCTs were all considered to be at a low risk of attrition bias.
Andersen 2002 reported an imbalance of participants withdrawing across groups, and that the participants who withdrew were not included in the final analysis. Banerjee 1990 did not explain why some participants, other than those reported as withdrawing, were not included at follow up. Callam 1992 reported reasons for withdrawal, that withdrawals were balanced across groups, and that the analysis was undertaken as ITT, however, the report suggested that the missing participant data was not included in the final analysis. Although all participants were included in the healing analyses (ITT) of Charles 2002, not all recruited participants were included in the analyses of secondary outcomes. These four RCTs were all judged to be at a high risk of bias for this domain.
One RCT reported the use of ITT, but did not define the ITT population and presented some outcomes that did not include all of the randomised participants (Norkus 2005). The RCT by Thomas 1997 did not provide any statement regarding participant withdrawal, and, whilst all patients randomised were accounted for in tabulated healing outcomes, not all participants were accounted for in the text that reported proportions of participants whose ulcers reduced in size. The RCT by Vanscheidt 2004 reported that 29% of all participants withdrew before 9.6 weeks, but did not report the numbers withdrawing according to their group, or state whether any participants had withdrawn after 9.6 weeks. The RCT report by Weiss 1996 suggested that all participants were included in the final analysis of healing, but it was not clear whether pain and other outcomes included all participants randomised. These four RCTs were all judged to be at an unclear risk of bias for this domain.
Selective reporting
Six RCTs were considered to be at low risk of bias for this domain as, although the RCT protocols were not available, all RCT outcomes described in the methods section of the report were included in the results section (Banerjee 1990; Zuccarelli 1992; Weiss 1996; Thomas 1997; Norkus 2005; Andriessen 2009).
Two RCTs were considered to be at high risk of bias for this domain as the RCT reports included results for outcomes that were not described as being evaluated in the methods section of the report (number of weekly dressing changes, Bowszyc 1995; and adverse events, Charles 2002).
RCT protocols were provided for four of the included RCTs (Callam 1992; Andersen 2002; Vanscheidt 2004; Franks 2007). The RCT by Callam 1992 was considered to be at low risk of bias for this domain as all outcomes mentioned in the RCT protocol were presented in the RCT report. The RCT by Andersen 2002 was considered to be at unclear risk of bias as, although all RCT outcomes described in the published report were in the supplied RCT protocol, it was unclear from the published report what the primary outcome was (maceration in the protocol). A secondary outcome of ‘ability to adapt’ in the protocol (translated from Danish) was not identifiable in the published report. The RCT by Franks 2007 did not report the health‐related quality of life outcomes that were pre‐specified in the protocol and was classified as being at high risk of bias. The RCT by Vanscheidt 2004 was considered to be at high risk of selective reporting bias, as the published report indicated the use of scales to assess some outcomes that were not detailed in the protocol, the pre‐specified outcome of patient‐reported dressing leakage was not reported, and the cost of managing complications specified in the protocol was not included in the RCT report.
Overall risk of bias
Six of the included RCTs were considered as being at overall high risk of bias (Banerjee 1990; Callam 1992; Weiss 1996; Andersen 2002; Charles 2002; Franks 2007), as they were each rated as 'high' for one or more of three key domains (allocation concealment, blinding of outcome assessors and completeness of outcome data). The remaining six RCTs were classified as being at unclear risk of bias overall (Zuccarelli 1992; Bowszyc 1995; Thomas 1997; Vanscheidt 2004; Norkus 2005; Andriessen 2009), with three having 'unclear' ratings for all three key domains (Thomas 1997; Vanscheidt 2004; Norkus 2005). None of the RCTs was rated as having a low risk of bias overall.
Effects of interventions
See: Summary of findings for the main comparison polyurethane foam dressing compared to hydrocellular foam dressing for venous leg ulceration; Summary of findings 2 foam dressing compared to paraffin gauze dressing for venous leg ulceration; Summary of findings 3 foam dressing compared to hydrocapillary dressing for venous leg ulceration; Summary of findings 4 foam dressing compared to hydrocolloid dressing for venous leg ulceration; Summary of findings 5 foam dressing compared to knitted viscose dressing for venous leg ulceration
Twelve RCTs that evaluated foam dressings were included in this review. The results are presented according to the type of comparator dressing, starting with hydrocellular foam dressings compared with polyurethane foam dressings. This is followed by comparisons of foam dressings with: paraffin gauze dressings, hydrocapillary dressings, hydrocolloid dressings, knitted viscose dressings, and protease‐modulating matrix dressings. Details of primary and secondary outcome data are presented in Table 1 and Table 2.
| Time to healing weeks ‐ mean (SD): Group 1 (hydrocellular foam): 5.0 (1.7); Group 2 (polyurethane foam): 5.2 (1.9).
Number (%) of ulcers healed at 8 weeks: Group 1 (hydrocellular foam): 18/60 (30); Group 2 (polyurethane foam): 18/58 (31).
Change in ulcer size – mean change in area cm2: Group 1 (hydrocellular foam): ‐3.77; Group 2 (polyurethane foam): ‐3.18. P value for between‐group difference not reported by trialists.
Change in ulcer size – mean percentage change in area (read from graph): Group 1 (hydrocellular foam): ‐62%; Group 2 (polyurethane foam): ‐64%. P value for between‐group difference not reported by trialists. | |
| Change in ulcer size – mean (range) percentage change in area at 4 weeks: Group 1 (paraffin gauze): ‐17.2% (‐31% to 4.6%); Group 2: (protease‐modulating matrix ): ‐31.8% (‐28% to ‐34%); Group 3 (polyurethane foam): ‐26.4% (‐17.3% to ‐32%). Comments: trialists reported that all participants were compliant with treatment. | |
| Number (%) of ulcers healed at 17 weeks: Group 1 (paraffin gauze): 8/35 (23); Group 2 (polyurethane foam): 11/36 (31).
Ulcer area cm2 at 17 weeks median (range): Group 1 (paraffin gauze): 4.4 (0.2 to 22), n = 12; Group 2 (polyurethane foam): 6.3 (0.2 to 108.2), n = 14. Trialists did not report the change in ulcer size.
Change in ulcer area at 17 weeks cm2‐ difference in medians (calculated by review authors): Group 1 (paraffin gauze): ‐7.0; Group 2 (polyurethane foam): ‐5.9. | |
| Time to healing weeks: Group 1 (hydrocolloid): 9.34; Group 2 (polyurethane foam): 10.50. Not reported whether values are mean or median. Trialists reported P value = 0.35 for between‐group difference.
Number (%) of legs healed at 16 weeks: Group 1 (hydrocolloid): 24/41 legs (59); Group 2 (polyurethane foam): 24/41 legs (59).
Comments: 7 participants were excluded due to protocol violations, 5 due to heavily‐exuding wounds and 2 due to streptococcal cellulitis. Breakdown by group not reported, and not reported whether exclusion was pre‐ or post‐randomisation. There were 8 cases of streptococcal cellulitis (4 in each group) that required antibiotic treatment; 2 participants were withdrawn from the trial as a result. | |
| Time to healing: The reported P value for between‐group difference in time to healing from a Cox survival model including the covariate of initial ulcer size was 0.08. No other data were reported.
Number (%) of ulcers healed at 12 weeks: Group 1 (Knitted viscose): 23/66 (35); Group 2 (hydrocellular foam): 31/66 (47).
Change in ulcer size: Trialists reported a greater reduction in percentage ulcer area in Group 2 compared with Group 1, but the difference was not statistically significant at P value = 0.051. No average data were presented by group.
Percentage of ulcers with change in ulcer area ‐ > 100% increase vs 50%‐100% increase vs 0‐50% increase vs 0‐50% decrease vs 50%‐100% decrease vs healed vs not known (read from graph): Group 1 (Knitted viscose): 8% vs 8% vs 14% vs 15% vs 19% vs 35% vs 1%; Group 2 (hydrocellular foam): 4% vs 4% vs 10% vs 12% vs 21% vs 47% vs 2%. P value = 0.051 reported by trialists for between‐group difference in ulcer area reduction. Comments: trialists reported that initial or final ulcer tracing was not available for 3 patients. | |
| Mean time to healing weeks: Group 1 (hydrocolloid, Comfeel): 6.0; Group 2 (hydrocolloid, Granuflex): 7.5; Group 3 (polyurethane foam): 7.0.
Number (%) of ulcers healed at 12 weeks: Group 1 (hydrocolloid, Comfeel): 17/29 (59); Group 2 (hydrocolloid, Granuflex): 17/31 (55); Group 3 (polyurethane foam): 18/31 (58).
Change in ulcer size – mean percentage change in ulcer area: Group 1 (hydrocolloid Comfeel): ‐82.1%; Group 2 (hydrocolloid Granuflex): ‐83.6%; Group 3 (polyurethane foam): ‐67.3%. Trialists reported the percentage of ulcer area remaining, stating that the between‐group difference was not statistically significant (P value not provided). The percentage change values presented here were extrapolated by the review authors. | |
| Number (%) of ulcers healed at 24 weeks: Group 1 (polyurethane foam): 50/75 (66.7); Group 2 (hydrocellular foam): 50/81 (61.7).
Cumulative healing rates at 12 weeks derived from Kaplan‐Meier survival curves: Group 1 (polyurethane foam): 50.7%; Group 2 (hydrocellular foam): 47.5%.
Hazard ratio (95% CI) estimation from Cox proportional hazards model: Initial model adjusted for dressings, bandages and study centres HR 1.48 (0.87 to 2.54), P value = 0.15. Further model with additional adjustment for baseline covariates HR 1.50 (0.86 to 2.62), P value = 0.16. Comment: in cases where the original ulcer closed but a new area developed on the same limb while the original ulcer was still present, the limb was considered to be open until this new area of ulceration had also closed (Franks 2004, see references listed under Franks 2007). | |
| Time to healing: Trialists reported that no statistically significant difference was found between Group 1 and Group 2 for time to healing, but no data by group, or P value for difference were presented. Unclear what analysis method was used (described as Kaplan‐Meier principle).
Number (%) of ulcers healed at 12 months: Group 1 (hydrocapillary): 25/49 (51); Group 2 (polyurethane foam): 19/48 (40). | |
| Number (%) of ulcers healed at 13 weeks: Group 1 (hydrocolloid): 19/50 (38); Group 2 (polyurethane foam): 17/50 (34).
Change in ulcer area ‐ mean percentage change in ulcer area (read from graph): Group 1 (hydrocolloid): ‐78%; Group 2 (polyurethane foam): ‐88%. Trial authors reported no statistically significant difference between groups, but no P value presented.
Number (%) participants with ulcers that reduced in area during trial period: Group 1 (hydrocolloid): 34/50 (68); Group 2 (polyurethane foam): 44/50 (88). | |
| Time to healing days ‐ mean (SD): Group 1 (hydrocolloid): 66 (3.4); Group 2 (hydrocellular foam): 72.6 (3.1). Trialists reported that there were no statistically significant between‐group differences (P value = 0.47).
Number (%) of ulcers healed at 12 weeks: Group 1 (hydrocolloid): 20/55 (36); Group 2 (hydrocellular foam): 20/52 (38). Slight discrepancy between text and table in trial report (values above taken from table). Text of report states proportion of ulcers healed in Group 1 38.2%, and Group 2 38.5%.
Healing rate ‐ median change per week in cm2: Group 1 (hydrocolloid): ‐0.41; Group 2 (hydrocellular foam): ‐0.43. Trialists reported that there were no statistically significant between‐group differences (P value = 0.13).
Healing rate ‐ median percentage change per week: Group 1 (hydrocolloid): ‐7.3%; Group 2 (hydrocellular foam): ‐6.1%, Trialists reported that there were no statistically significant between‐group differences (P value = 0.27). | |
| Mean time to healing ‐ weeks: Group 1 (polyurethane foam): 5.6; Group 2 (hydrocellular foam): 6.5.
Number (%) of ulcers healed at 16 weeks: Group 1 (polyurethane foam): 8/10 (80); Group 2 (hydrocellular foam): 4/8 (50).
Change in ulcer size ‐percentage change (dimension not stated) at week 3: Group 1 (polyurethane foam): ‐27%; Group 2 (hydrocellular foam): ‐13%. P value for between‐group difference not reported by trialists. | |
| Number (%) of ulcers healed at 12 weeks: Group 1 (hydrocolloid): 9/19 (47); Group 2 (hydrocellular foam): 9/19 (47). Comment: trialists reported that the mean ulcer size in both groups reduced steadily throughout the trial, but no statistically significant between‐group difference was observed (P value = 0.86). |
| Number (%) participants experiencing adverse events: Group 1 (hydrocellular foam): Maceration/erythema/secondary infected dermatitis, 1 (2). Erysipelas, 1 (2). Serious adverse event: hospitalisation (pain), 1 (2); hospitalisation (cardiovascular accident), 1 (2); death (acute myocardial infarction), 1 (2). Group 2 (polyurethane foam): Maceration/erythema/secondary infected dermatitis, 1 (2). Pain after dressing change, 1 (2). Allergic reaction (dermatitis), 1 (2). Serious adverse event: no serious events reported. The trialists reported that they could not verify a relationship between the dressings and the adverse events (resulting in withdrawal from the study).
Average material costs per participant per week: Group 1 (hydrocellular foam): $ 18.99 ; Group 2 (polyurethane foam): $ 10.87. Included costs of dressings, unclear whether bandages and other treatment materials included, did not include nursing time, price year not stated, presumed currency is USD. Mean number of dressing changes per week: Group 1 (hydrocellular foam): 3.34; Group 2 (polyurethane foam): 2.14. The trialists reported a statistically significant difference (P value < 0.0005).
Pain at weekly assessments: Group 1 (hydrocellular foam): no pain was recorded at 169/319 (53%) weekly assessments; Group 2 (polyurethane foam): no pain was recorded at 226/379 (60%) weekly assessments. P value for between‐group differences not reported by trialists.
Pain ‐ comfort of the dressing: Participant ratings of comfort during treatment were reported as not significantly different between groups, but no data were presented.
Dressing performance ‐ absorbency: Group 1 (hydrocellular foam): dressing absorbency was rated as excellent in 12/170 (7%) of dressing changes; Group 2 (polyurethane foam): dressing absorbency was rated as excellent in 124/163 (76%) of dressing changes. The trialists reported a statistically significant difference (P value < 0.0005).
Dressing performance ‐ leakage: Group 1 (hydrocellular foam): leakage of exudate was observed in 198/309 (64%) of weekly assessments; Group 2 (polyurethane foam): leakage of exudate was observed in 172/355 (48%) of weekly assessments. The trialists reported a statistically significant difference (P value < 0.0005). | |
| Adverse events: The trialists stated that no adverse events were reported.
Pain at dressing removal: Group 1 (paraffin gauze): participants reported either moderate pain (mean VAS score of 4 to 6), or severe (mean VAS score of 7 to 9); Group 2: (protease‐modulating matrix dressing): participants reported little or no pain (mean VAS score of 1 to 2); Group 3 (polyurethane foam): participants reported little or no pain (mean VAS score of 1 to 2). Trialists report that the paraffin gauze dressing stuck to the wound bed in 72% of the dressing changes (no report of this outcome for other treatment groups). | |
| Adverse events ‐ number (%) participants died: Group 1 (paraffin gauze): 3/35 (9); Group 2 (polyurethane foam): 7/36 (19). Adverse events ‐ ulcer infection: Trialists reported that there were no statistically significant differences between groups in the proportion of patients with infected ulcers at weeks 1, 4, 8, 12 or 17, but further details were not provided. Pain: Trialists reported that the ulcer pain score was similar across groups with no statistically significant difference, but no data were presented.
Number of dressing changes per week: Trialists reported that more dressings and pads were used in Group 2 (polyurethane foam), but no data by group or P value for between‐group differences were reported.
Number of visits for healed/unhealed ulcers ‐ median: Group 1: (paraffin gauze) 6.5/25.0; Group 2: (polyurethane foam) 16.0/40.0. The trialists reported the following P values for between‐group differences: P < 0.05 for healed ulcers; and P < 0.002 for unhealed ulcers.
Nursing time spent per patient per week ‐ median: Group 1 (paraffin gauze): 4 hours 10 minutes Group 2 (polyurethane foam): 5 hours 40 minutes Trialists reported P value < 0.05 for between‐group difference. | |
| Number (%) patients experiencing adverse events: Not reported in full. Trialists reported that complications of secondary infection and maceration occurred: there were 8 cases of streptococcal cellulitis which required antibiotic treatment (2 patients withdrew because of this). The same number of patients were affected in each group.
Costs: UK Drug Tariff price per dressing (at October 1993) was reported as GBP 0.92 for 10 cm x 10 cm piece of foam dressing (Lyofoam) and GBP 2.08 for 10 cm x 10 cm piece of hydrocolloid dressing (Granuflex).
Pain ‐ dressing comfort score, mean (SD): Group 1 (hydrocolloid): 8.2 (2.5); Group 2 (polyurethane foam): 8.7 (2.2). Trialists reported no statistically significant between‐group difference (P value not provided).
Pain score at dressing removal, mean (SD): Group 1 (hydrocolloid): 3.63 (0.83); Group 2 (polyurethane foam): 3.72 (0.55). Trialists reported no statistical between‐group difference (P value not reported).
Mean number of dressing changes per week in first month: Group 1 (hydrocolloid): 1.5; Group 2 (polyurethane foam): 1.6. Trialists reported no statistically significant between‐group difference (P value = 0.49).
Nurse‐rated ease of removal – mean (SD) score from 11 point scale: Group 1: (hydrocolloid) 1.42 (0.54); Group 2 (polyurethane foam): 1.15 (0.42). Trialists reported P value = 0.016 for between‐group difference. | |
| Number (%) participants experiencing adverse events: Group 1 (knitted viscose): no adverse reactions; Group 2 (hydrocellular foam): local skin reaction, 2 (4). Number (%) participants experiencing skin complications ‐ sensitivity reaction to the dressing vs dry eczema vs wet eczema vs maceration vs other dressing complications (from supplied RCT report): Group 1 (knitted viscose): 3/66 (5) vs 47/66 (71) vs 16/66 (24) vs 3/66 (5) vs 5/66 (8); Group 2 (hydrocellular foam): 9/66 (14) vs 37/66 (56) vs 23/66 (35) vs 11/66 (17) vs 12/66 (18).
Pain at clinic visits: Number (%) participants experiencing pain at 0% vs 1%‐25% vs 26%‐50% vs 51%‐75% vs 76%‐99% and 100% of clinic visits: Group 1 (knitted viscose, n = 66 participants): 7 (11%) vs 8 (12%) vs 5 (8%) vs 5 (8%) vs 7 (11%) vs 34 (52%); Group 2 (hydrocellular foam, n = 66 participants): 13 (20%) vs 6 (9%) vs 9 (14%) vs 12 (18%) vs 9 (14%) vs 17 (26%). P value = 0.01 reported by trialists for between‐group difference in patients experiencing pain during 100% of clinic visits. Pain disturbing sleep: Number (%) participants reporting pain disturbing sleep at 0% vs 1%‐25% vs 26%‐50% vs 51%‐75% vs 76%‐99% and 100% of clinic visits: Group 1 (knitted viscose, n = 66 participants): 35 (53%) vs 16 (24%) vs 7 (11%) vs 1 (2%) vs 5 (8%) vs 2 (3%); Group 2 (hydrocellular foam, n = 66 participants): 34 (52%) vs 17 (26%) vs 6 (9%) vs 4 (6%) vs 0 (0%) vs 5 (8%). | |
| Adverse events: Trialists reported that no serious dressing‐related adverse events were observed during the trial.
Pain – percentage of participants reporting ulcer pain at 12 weeks (read from graph): Group 1 (hydrocolloid Comfeel): 8%; Group 2 (hydrocolloid Granuflex): 4%; Group 3 (polyurethane foam): 10%.
Pain ‐ mean ulcer‐associated VAS pain score at week 12: Group 1 (hydrocolloid, Comfeel): 0.64, n = 28; Group 2 (hydrocolloid, Granuflex): 0.15, n = 27; Group 3 (polyurethane foam): 0.50, n = 30. Trialists reported that between‐group differences in pain prevalence and severity were not statistically significant at any point in the trial (P value not reported). Change in mean scores (calculated by review authors): Group 1 (hydrocolloid, Comfeel): ‐3.19; Group 2 (hydrocolloid, Granuflex): ‐3.66; Group 3 (polyurethane foam): ‐4.27. Pain ‐ proportion of patients reporting mid to lower‐range pain severity on VAS and McGill Pain Questionnaire (MPQ‐PPI) at week 12 (secondary referenceCharles 2002): Trialists reported that the number of participants reporting pain fell to < 10%. Unclear whether this was overall, or for a specific treatment group. No data by group presented. Dressing performance ‐ ease of use of dressings: The foam dressing was significantly easier to remove compared with the two hydrocolloid dressings at 6 weeks and at last dressing (P value = 0.021 and P value = 0.037 respectively for the 2 time points; unclear how these P values relate to the between‐group comparisons given there would be 2 comparisons at each time point). | |
| Number (%) participants experiencing adverse events: Group 1 (polyurethane foam): 23/75 (30.7); Group 2 (hydrocellular foam): 22/81 (27.2). Trialists did not report the P value for between‐group difference.
Category of dressing‐related adverse event and number of occurrences: Group 1 (polyurethane foam): not related to dressing = 21; possibly related to dressing = 4; definitely related to dressing = 11; Group 2 (hydrocellular foam): not related to dressing = 15; possibly related to dressing = 4; definitely related to dressing = 11.
Number (%) participants who died during trial: Group 1 (polyurethane foam): 0/75 (0); Group 2 (hydrocellular foam): 1/81 (1.2).
Pre‐dressing pain at week 4 (VAS) – median (interquartile range): Group 1 (polyurethane foam): 1.5 (3.6), n = 53; Group 2 (hydrocellular foam): 1.3 (3.7) n = 53. The trialists reported P value = 0.18 for between‐group difference.
Post‐dressing pain at week 4 (VAS) – median (interquartile range): Group 1 (polyurethane foam): 1.0 (2.0), n = 51 Group 2 (hydrocellular foam): 0.5 (1.8), n = 52 Trialists reported P value = 0.12 for between‐group difference.
McGill Pain Questionnaire sensory pain scores at week 4 – median (interquartile range): Group 1 (polyurethane foam): 2.0 (4.0), n = 47; Group 2 (hydrocellular foam): 1.0 (3.0), n = 31. Trialists reported P value = 0.40 for between‐group difference.
McGill Pain Questionnaire affective pain scores at week 4– median (interquartile range): Group 1 (polyurethane foam): 0.0 (0.0), n = 42; Group 2 (hydrocellular foam): 0.0 (0.0), n = 47. Trialists reported P value = 0.69 for between‐group difference
Trialists reported that the levels of pain reduced significantly for both the VAS and McGill pain scores in both groups, however, the between‐group differences at week 4 were not statistically significant. | |
| Number (%) participants experiencing adverse events (type of adverse event): Group 1 (hydrocapillary dressing): 8/49 (16%) (4 adverse events defined as related to dressing, of which: ulcer infection = 2, skin reactions = 2, description of other types of adverse events not provided). All 8 participants withdrew from the trial. Group 2 (polyurethane foam): 11/48 (23%) (7 defined as related to dressing, of which: ulcer infections = 2, eczema = 3, allergic reaction = 1, peri‐ulceration = 1, description of other types of adverse events not provided). All 11 participants withdrew from the trial.
Number (%) participants developing clinical signs of wound infection (all withdrew): Group 1 (hydrocapillary dressing): 5/49 (10%); Group 2 (polyurethane foam): 4/48 (8%). It is unclear whether these participants with wound infection were also counted among those described above (adverse events).
Maceration: trialists reported that: 1 in 3 participants experienced some degree of maceration during the study; severe maceration was only reported in the polyurethane foam group (5%), and there were no significant between‐group differences in maceration, however, no further data were provided, neither were P values.
Number of times that signs of allergy on peri‐ulcer skin were reported: Group 1 (hydrocapillary dressing): ;7 Group 2 (polyurethane foam): 16. The trialists reported that allergic reactions were confirmed in 1 participant in each group by a positive patch test, and that there were no significant between‐group differences in allergy.
Number of times that severe erythema was reported: Group 1 (hydrocapillary dressing): 24 (1.7%); Group 2 (polyurethane foam): 45 (3.3%), Percentages as reported by trialists – denominators not stated. Trialists reported no significant between‐group difference in erythema.
Quality of life – median WHO‐5 Well‐Being Index scores: Group 1 (hydrocapillary): 68; Group 2 (polyurethane foam): 68. Trialists reported that there were no significant differences in the initial or final values between groups (P value not reported).
Wound pain between dressing changes: Trialists reported that there was no significant difference in the incidence of wound pain between the 2 groups. Data by group and P value not reported.
Pain – number (%) assessing dressing comfort as good vs medium vs poor: Group 1 (hydrocapillary): 42/43 (97.6) vs 1/43 (2.3) vs 0/43 (0); Group 2 (polyurethane foam): 32/45 (71.1) vs 8/45 (17.8) vs 5/45(11.1). Trialists reported that the comfort of the dressing in Group 1 was rated as significantly better than Group 2 (P value < 0.001).
Dressing performance, exudate handling – number (%) of incidents of little to moderate vs severe leakage at dressing change: Group 1 (hydrocapillary): 139 (11) vs 1 (0.1); Group 2 (polyurethane foam): 144 (13) vs 22 (2). Trialists reported that no significant differences in leakage were found between the 2 treatment groups. Denominators not reported.
Trialists reported that the hydrocapillary dressing was rated by nurses as having a significantly better capacity to absorb exudate compared with the polyurethane foam (P value < 0.05 – no other data provided).
Dressing performance, adherence – number (%) dressing changes where dressing adherence to the wound was reported: Group 1 (hydrocapillary dressing): 95/542 (18); Group 2 (polyurethane foam): 113/388 (29). Trialists reported P value < 0.05 for between‐group difference.
Percentage of dressing changes (as reported by trialists) where adherence was rated as none vs little vs moderate vs severe: Group 1 (hydrocapillary dressing): 82.5% vs 10.5% vs 6.6% vs 0.4%; Group 2 (polyurethane foam): 70.9% vs 12.6% vs 13.1 % vs 3.4%.
Mean wear time of dressing, days: Group 1 (hydrocapillary): 3.2; Group 2 (polyurethane foam): 3.2. | |
| Number (%) all participants (venous leg ulcers and pressure ulcers) experiencing adverse events linked to the dressing (described as due to adhesive nature of dressings causing minor trauma/erythema on removal): Group 1 (hydrocolloid): 7/99 (7); Group 2 (polyurethane foam): 10/100 (10). Numbers by ulcer type not reported. Number of participants with additional minor dressing related adverse events (venous leg ulcers only): Group 1 (hydrocolloid): maceration = 6; bleeding from the wound bed = 1; excessive granulation = 3; Group 2 (polyurethane foam): maceration = 0; bleeding from the wound bed = 0; excessive granulation = 0. Pain ‐ dressing comfort: Trialists reported that Group 2 (polyurethane foam) was significantly more likely to be scored as comfortable (comfortable vs otherwise, P value = 0.023) at the first dressing change. No data by group were presented.
Number participants with wound dehydration and dressing adherence to wound (all wound types): Group 1 (hydrocolloid): 0; Group 2 (polyurethane foam): 2. Group denominators not stated.
Dressing performance (exudate handling) – number (%) participants who had dressings removed at the first dressing change due to leakage: Group 1 (hydrocolloid): 27/50 (54); Group 2 (polyurethane foam): 7/50 (14). Trialists reported that the between‐group difference was statistically significant (P value < 0.0001). Dressing performance ‐ ease of removal: Trialists reported that dressing removal in both treatment groups was rated by nurses as 'easy' in the majority of cases at the first dressing change. No data by group were presented.
Mean time the dressing was in place prior to the first change, days: Group 1 (hydrocolloid): 2.7; Group 2 (polyurethane foam): 2.8. | |
| Number (%) participants experiencing adverse events: Group 1 (hydrocolloid): 13/55 (24) experienced 1 or more possibly dressing‐related adverse event (19 events in total); Group 2 (hydrocellular foam): 15/52 (29) experienced 18 possibly dressing‐related adverse events.
Common adverse events and number of participants affected: Group 1 (hydrocolloid): new wound development in different location = 6; other non‐specific wound events (e.g. affecting participant wounds other than the study wound) = 5; reason not reported = 2; Group 2 (hydrocellular foam): maceration = 6; new wound development in different location = 4; other non‐specific wound events (e.g. affecting participant wounds other than the study wound) = 5. Review authors' comment: numbers/types of adverse events per group were presented twice in the main text of the primary reference. Numbers of patients with different types of adverse events were discrepant between the 2 reports for Group 2. Trialists reported no significant between‐group difference (P value = 1.00, not clear if this referred to the difference in number of patients with adverse events or difference between total number of adverse events in group).
Serious adverse events and number of participants affected: Group 1 (hydrocolloid): angina = 1; musculoskeletal = 1; died = 3; Group 2 (hydrocellular foam): unspecified blood disease = 1.
Number (%) dressing changes with investigator‐rated condition of peri‐ulcer skin classified as normal vs maceration vs erythema vs eczema vs lipodermatosclerosis vs cellulitis vs dermatitis vs other: Group 2 (hydrocolloid): 269/615 (44) vs 135/615 (22) vs 52/615 (8) vs 5/615 (1) vs 119/615 (19) vs 0/615 (0) vs 16/615 (3) vs 19/615 (3); Group 1 (hydrocellular foam): 190/529 (36) vs 143/529 (27) vs 48/529 (9) vs 10/529 (2) vs 112/529 (21) vs 1/529 (<1) vs 16/529 (3) vs 9/529 (2). Denominators are total number of dressing changes. Trialists reported that no between‐group differences were statistically significant, but no P value provided.
Pain at final assessment, number (%) with reduction in pain vs no effect on pain vs uncertain vs unable to respond/no response vs increased pain vs missing data: Group 1 (hydrocolloid): 36/55 (65) vs 7/55 (13) vs 5/55 (9) vs3/55 (5) vs 4/55 (7) vs 0/55 (0); Group 2 (hydrocellular foam): 21/52 (40) vs 15/52 (29) vs 4/52 (8) vs 0/52 (0) vs 9/52 (17) vs 3/52 (6). Percentages for Group 2 recalculated by review authors to take account of missing data. Trialists reported that there were no statistically significant between‐group differences (P value = 0.10). Dressing performance (exudate handling) – number (%) participants with minimal or moderate levels of exudate at dressing change: Group 1 (hydrocolloid): 43/55(79); Group 2 (hydrocellular foam): 43/52 (83). Trialists reported that the between‐group difference was not statistically significant but P value not provided.
Dressing performance (exudate handling) – number (%) dressing changes with no exudate vs minimal exudate (< 25% of dressing saturated) vs moderate exudate (25%‐75% of dressing saturated) vs considerable exudate (> 75% of dressing saturated): Group 1 (hydrocolloid): 51/615 (8) vs 204/615 (33) vs 234/615 (38) vs 126/615 (20); Group 2 (hydrocellular foam): 42/529 (8) vs 199/529 (38) vs 196/529 (37) vs 89/529 (17). Trialists reported that the between‐group difference was not statistically significant but P value not provided.
Mean (SD) number of dressing changes per week: Group 1 (hydrocolloid): 11.4 (6.5); Group 2 (hydrocellular foam): 10.1 (6.3). Trialists reported that the between‐group difference was not statistically significant but P value not provided. Mean (SD) dressing wear time days: Group 1 (hydrocolloid): 5.6 (1.3); Group 2 (hydrocellular foam): 5.6 (1.2). Trialists reported that the between‐group difference was not statistically significant but P value not provided.
Dressing performance ‐ percentage of participants who rated non‐traumatic dressing removal categories of excellent/very good vsgood vsfair or poor (read from graph): Group 1 (hydrocolloid): 84% vs 11% vs 5% vs 0% Group 2 (hydrocellular foam): 68% vs 22% vs 4% vs 6% Trialists reported P value = 0.07 for between‐group differences. Dressing performance ‐ percentage of participants who rated ease of removal categories of excellent/very good vs good vs fair vs poor (read from graph): Group 1 (hydrocolloid): 84% vs 10% vs 4% vs 2%; Group 2 (hydrocellular foam): 74% vs 18% vs 6% vs 2%. Trialists reported P value = 0.19 for between‐group differences. | |
| Number (%) patients experiencing adverse events (type of adverse event): Group 1 (polyurethane foam): 0/10 (0); Group 2 (hydrocellular foam): 1/8 (38) ‐ allergic reaction which necessitated withdrawal.
Pain reduction and dressing comfort: Trialists stated that all participants reported a great reduction in pain, and that participants in Group 1 (polyurethane foam) preferred the comfort of the dressing. No data by group were presented. | |
| Number (%) patients experiencing adverse events: Group 1 (hydrocolloid): 2/19 (11) showed an allergic reaction and an intolerance in the peri‐ulcer skin area in the form of blisters; Group 2 (hydrocellular foam): no unwanted side effects were reported.
Mean (SD) number of dressing changes per week: Group 1 (hydrocolloid): 26.9; Group 2 (hydrocellular foam): 19.5 Trialists reported that the between‐group difference was not significant (P value = 0.14).
Mean pain scores over the course of treatment: Group 1 (hydrocolloid dressing): scores decreased during the first 2 weeks of treatment (P value = 0.005 for within group difference over time), increased slightly in the third week and then stabilised from then on. Group 2 (hydrocellular foam dressing): scores decreased during the first week of treatment (P value = 0.005 for within group difference over time) and remained stable from then on. Trialists reported no significant difference between groups.
Mean pain score at 12 weeks (read from graph): Group 1 (hydrocolloid): 1.5; Group 2 (hydrocellular foam): 1.7. Trialists reported that the between‐group difference was not significant (P value = 0.52). Change in mean scores (calculated by review authors from graph): Group 1 (hydrocolloid): ‐0.8; Group 2 (hydrocellular foam): ‐0.9. |
Abbreviation
VAS = visual analogue scale
Hydrocellular foam dressings compared with polyurethane foam dressings
Three RCTs recruiting a total of 295 participants were identified that compared a hydrocellular foam dressing with a polyurethane foam dressing (Weiss 1996; Andersen 2002; Franks 2007).
Primary outcomes
Time to complete ulcer healing
Weiss 1996 randomised eight participants to a hydrocellular foam dressing (Allevyn) and ten to a polyurethane foam dressing (Cutinova). All participants received compression therapy in the form of graduated compression stockings. The trialists reported a mean time to healing of 6.5 weeks in the hydrocellular foam group compared with 5.6 weeks in the polyurethane foam group. No variance data or P value for the between‐group difference was reported. This RCT was considered to be at overall high risk of bias as it was described as a non‐blinded evaluation. This RCT was a pilot study with a small sample size.
Andersen 2002 randomised 60 participants to receive a hydrocellular foam dressing (Allevyn) and 58 to receive a polyurethane foam dressing (Biatain). Short‐stretch compression bandaging was used. Participants were assessed every seven days until the ulcer was completely healed or the trial period (eight weeks) was completed. It should be noted that the median baseline ulcer duration was longer in the group receiving the hydrocellular foam dressing. The mean (SD) time to healing in the hydrocellular foam dressing group was 5.0 (1.7) weeks (n = 46) compared with 5.2 (1.9) weeks (n = 53) in the polyurethane foam dressing group. A P value for the between‐group difference was not reported by the trialists. This RCT was judged to be at overall high risk of bias as the RCT was described as not blinded and participants who withdrew were not included in the final analysis.
Franks 2007 randomised 81 participants to receive a hydrocellular foam dressing (Allevyn) and 75 to receive a silicone‐faced polyurethane foam dressing (Mepilex). The factorial design of the RCT also allowed evaluation of two compression systems (four‐layer bandage versus short‐stretch bandage). Approximately 50% of participants in each randomised dressing group were randomised to each of the two compression bandaging systems. The cumulative healing rates at 12 weeks derived from Kaplan‐Meier survival curves were 47.5% in the hydrocellular foam group compared with 50.7% in the polyurethane group. The hazard ratio (95% CI) estimation for healing derived from a Cox proportional hazards model adjusted for baseline covariates (ulcer size, ulcer duration and participant mobility) did not suggest a statistically significant difference between treatment groups: 1.50 (0.86 to 2.62), P value 0.16. The hazard ratio is an expression of the hazard (chance) of an event (e.g. ulcer healing) occurring in one treatment group as a ratio of the hazard of the event occurring in the other group. These results reported by Franks 2007 can be interpreted as an absence of a statistically significant difference between the hydrocellular foam dressing group and the polyurethane foam dressing group in the rate at which ulcers healed over the course of the trial, irrespective of variations in treatment delivery or participant characteristics. This RCT was considered to be at overall high risk of bias, as the outcome assessment was not blinded.
Proportion of ulcers completely healed
All three RCTs reported this outcome (Weiss 1996; Andersen 2002; Franks 2007).
In the RCT by Andersen 2002, the number of ulcers healed at eight weeks was 18/60 (30%) in the hydrocellular foam dressing group compared with 18/58 (31%) in the polyurethane foam dressing group. The between‐group difference was not statistically significant (RR 1.03, 95% CI 0.60 to 1.78) (Analysis 1.1).
Weiss 1996 reported that the number of ulcers completely healed at 16 weeks was 8/10 (80%) in the polyurethane foam group compared with 4/8 (50%) in the hydrocellular foam group. The between‐group difference was not statistically significant (RR 1.60, 95% CI 0.75 to 3.42) (Analysis 1.1).
Franks 2007 reported that the proportion of ulcers healed at 24 weeks was 50/81 (61.7%) in the hydrocellular foam dressing group compared with 50/75 (66.7%) in the polyurethane foam dressing group. The between‐group difference in was not statistically significant (RR 1.08, 95% CI 0.85 to 1.37) (Analysis 1.1).
A pooled estimate of effect across these RCTs was not undertaken because of variation in follow up periods.
Change in ulcer size
Weiss 1996 reported the percentage reduction in ulcer size at week three of the 16‐week RCT, but did not report the dimension (area, length/width, diameter or circumference). The percentage reduction was 27% in the hydrocellular foam group compared with 13% in the polyurethane foam group. No variance data were reported. The trialists did not report a P value for the between‐group difference.
Andersen 2002 reported that the mean reduction in ulcer area at eight weeks was 3.77 cm2 in the hydrocellular foam group compared with 3.18 cm2 in the polyurethane foam group. No variance data were reported. The percentage reduction in ulcer area (read from graph) was 62% in the hydrocellular foam group compared with 64% in the polyurethane foam group. The trialists did not report a P value for the between‐group differences for either the absolute or relative change.
Secondary outcomes
Adverse events
Weiss 1996 recorded the incidence of an allergic reaction and reported that 1/8 (38%) participants in the hydrocellular foam group experienced an allergic reaction (type not specified) that necessitated withdrawal from the RCT. The RCT report contained no statement regarding adverse events with respect to the polyurethane foam group.
Andersen 2002 assessed peri‐ulcer skin reactions using a verbal rating scale (VRS) and reported that 2/60 (3%) participants in the hydrocellular foam group and 3/58 (5%) participants in the polyurethane foam group experienced adverse events. These included maceration, erythema, secondary infected dermatitis, erysipelas, allergic reaction and pain at dressing change; some participants experienced more than one type of adverse event. Two out of the 60 participants in the hydrocellular foam group were hospitalised and one died; these were classified by the trialists as serious adverse events. No serious adverse events were reported in the polyurethane foam group.
Franks 2007 stated that 22/81 (27%) participants in the hydrocellular foam group reported 30 adverse events compared with 23/75 (31%) reporting 23 adverse events in the polyurethane foam group. Fifteen adverse events in each group were deemed as being possibly, or definitely, dressing‐related. The trialists also reported that one participant in the hydrocellular foam group died compared with none in the polyurethane foam group. The RR estimate for the number of participants experiencing adverse events did not indicate a statistically significant difference between groups: RR 1.13 (95% CI 0.69 to 1.85), Analysis 1.2.
Pooling of data was not undertaken because of the diversity of the nature of adverse events, methods of assessment, and reporting across the three RCTs.
Cost: material costs
Andersen 2002 reported that the average material costs per participant per week was 18.99 dollars in the hydrocellular foam group compared with 10.87 dollars in the polyurethane foam group (type of dollars and price year not reported). The estimation included the cost of dressings (though it was unclear whether bandages and other treatments were included), but excluded nursing time. No variance data were reported. The trialists did not report a P value for the between‐group difference.
Cost: number of dressing changes
Andersen 2002 reported that the mean number of dressing changes per week was 3.34 in the hydrocellular foam group compared with 2.14 in the polyurethane foam group. No variance data were reported. The trialists reported a statistically significant between‐group difference (P value < 0.0005).
Pain
Weiss 1996 reported that all participants experienced a reduction in pain and stated that the participants in the polyurethane foam group preferred the comfort of the dressing (but what they were asked to compare it to was not clear). The method of assessment was not reported, and no data by group were presented.
In the RCT by Andersen 2002, participants were asked to report ulcer pain during weekly assessments. No pain was recorded for the hydrocellular foam group during 169/319 (53%) weekly assessments compared with 226/379 (60%) weekly assessments in the polyurethane foam group. The P value for the between‐group difference was not reported by the trialists.
Franks 2007 reported median scores with inter‐quartile ranges at four weeks for pre‐ and post‐dressing pain derived from a visual analogue scale (VAS) and sensory pain and affective pain from the McGill pain questionnaire. The levels of pain reduced significantly for both the VAS and McGill pain scores in both groups, however, the between‐group differences at week four were not statistically significant. No variance data were reported.
Dressing performance: exudate management
Andersen 2002 reported that dressing absorbency, assessed by recording the incidence of leakage, was rated as excellent in 12/170 (7%) of dressing changes in the hydrocellular foam group compared with 124/163 (76%) of dressing changes in the polyurethane foam group. The trialists reported a statistically significant between‐group difference (P value < 0.0005). Leakage of exudate was observed in 198/309 (64%) of weekly assessments in the hydrocellular foam group compared with 172/355 (48%) of weekly assessments in the polyurethane foam group. The trialists reported a statistically significant between‐group difference (P value < 0.0005).
Hydrocellular foam dressings compared with polyurethane foam dressings: summary of results
Evidence from three RCTs indicated that there was no statistically significant difference between hydrocellular foam dressings and polyurethane foam dressings in the proportion of ulcers completely healed at 8, 16 or 24 weeks (Weiss 1996; Andersen 2002; Franks 2007). One of these RCTs was a pilot study with a small sample size (Weiss 1996) and all three were considered to be at overall high risk of bias.
One RCT reported an adjusted hazard ratio estimate suggesting that there was no statistically significant difference between treatment groups in terms of probability of healing over the 24‐week study period (Franks 2007). Data from the same RCT indicated no statistically significant difference between groups for the proportion of participants experiencing adverse events. Insufficient data were provided for the review authors to calculate measures of treatment effect for other outcomes, however, it is possible that polyurethane foam dressings perform better than hydrocellular foam dressings for exudate handling, and that pain scores are similar for the two dressings.
Paraffin gauze dressings compared with foam dressings
Two RCTs were identified that compared a paraffin‐impregnated gauze dressing with a hydrocellular foam dressing (Banerjee 1990; Andriessen 2009).
Primary outcomes
Proportion of ulcers completely healed
In the RCT by Banerjee 1990, 35 participants were randomised to receive a paraffin gauze dressing (Paratulle) and 36 were randomised to receive a polyurethane foam dressing (Synthaderm). The report provided no details about the use of compression therapy. The length of treatment was 17 weeks, or until the ulcer healed. More participants allocated to the paraffin gauze dressing group had mobility problems, and more participants allocated to the foam dressing group had infected ulcers. At 17 weeks 8/35 (23%) ulcers had healed in the paraffin gauze group compared with 11/36 (31%) in the polyurethane foam group. The between‐group difference was not statistically significant (RR 1.34, 95% CI 0.61 to 2.92) (Analysis 2.1). This RCT was considered to be at overall high risk of bias, as, in addition to participants reported as withdrawing, a number of other participants appeared to be missing from the final analysis.
Change in ulcer size
In one RCT, the change in median ulcer area from baseline to 17 weeks (estimated by the review authors from information provided in the RCT report) was ‐7.0 cm2 in the paraffin gauze group compared with ‐5.9 cm2 in the polyurethane foam group (Banerjee 1990). The other RCT was a three‐arm pilot study allocating 12 participants to receive one of three dressings: paraffin gauze (proprietary name not reported); polyurethane foam (Suprasorb P); and protease‐modulating matrix (Suprasorb C). All participants wore short‐stretch high compression bandages. The length of treatment was four weeks. The trialists reported that, as this was a pilot study, the sample size was not based on statistical considerations. The mean percentage change in ulcer area (range) at four weeks was ‐17.2% (‐31% to 4.6%) in the paraffin gauze group compared with ‐26.4% (‐17.3% to ‐32%) in the polyurethane foam group. No other variance data were reported. The trialists did not report a P value for the between‐group difference. The comparison between foam and protease‐modulating matrix dressings was reported in a later section (Andriessen 2009).
Secondary outcomes
Adverse events
Banerjee 1990 reported that 3/35 (9%) of participants in the paraffin gauze group and 7/36 (19%) of participants in the polyurethane foam group died during the course of the RCT. Andriessen 2009 reported that no adverse events were recorded during the RCT.
Cost: number of dressing changes
Banerjee 1990 reported that more dressings and pads were used in the polyurethane foam group than the paraffin gauze group, but did not report data by group, or a P value for between‐group differences.
Cost: nursing time
Banerjee 1990 reported that the nursing time spent per patient per week was five hours 40 minutes in the polyurethane foam group compared with four hours 10 minutes in the paraffin gauze group (not reported whether these values were means or medians). The trialists reported a P value of < 0.05 for the between‐group difference.
Pain
Banerjee 1990 assessed ulcer pain weekly on a four‐point scale (details of scale not reported), reporting that the score was similar across groups with no statistically significant difference, however, no data were presented. Andriessen 2009 recorded pain at each dressing removal on a 10 cm visual analogue scale (VAS). In the paraffin gauze group participants reported either moderate (mean VAS score of 4 to 6) or severe (mean VAS score of 7 to 9) pain at dressing removal, whereas in the polyurethane foam group participants reported little or no pain (mean VAS score of 1 to 2).
Dressing performance: adherence to the wound bed
Andriessen 2009 reported that the paraffin gauze dressing stuck to the wound bed in 72% of the dressing changes. The number of dressing changes was not reported for the polyurethane foam group.
Paraffin gauze dressings compared with foam dressings: summary of results
Evidence from one RCT indicated no statistically significant between‐group difference in foam dressings compared with paraffin gauze dressings in the proportion of ulcers completely healed at 17 weeks (Banerjee 1990); this RCT was considered to be at overall high risk of bias. Otherwise, data were limited and measures of treatment effect could not be estimated for other outcomes.
Hydrocapillary dressings compared with foam dressings
We identified one RCT that compared a hydrocapillary dressing with a foam dressing (Norkus 2005). Hydrocapillary dressings are designed to act as a wick so that exudate is drawn away from the wound surface. The dressing evaluated by Norkus 2005 (Alione adhesive, Coloplast) comprises a hydrocapillary pad surrounded by a hydrocolloid adhesive, which exposes a non‐adherent ulcer contact layer, covered by a water‐resistant, bacteria‐proof, semipermeable top film (Morris 2003).
Primary outcomes
Time to complete ulcer healing
Norkus 2005 randomised 49 participants to receive a hydrocapillary dressing (Alione) and 48 participants to receive two polyurethane foam dressings sequentially: one at the start of the RCT (Tielle plus) followed by the other (Tielle) when dressing changes were required less frequently. Thirty‐six of the 49 participants receiving the hydrocapillary dressing (73%) and 32 of the 48 participants receiving the foam dressings (67%) received compression therapy during the RCT (type not reported). Ulcers were treated until they healed, or for a maximum of 12 months. The trialists reported that no significant differences were found between the two treatment groups for time to healing, but did not present data by group or a P value for the between‐group difference, and the analysis method was unclear.
Proportion of ulcers completely healed
Norkus 2005 reported that 25/49 (51%) ulcers had healed at 12 months in the hydrocapillary dressing group compared with 19/48 (40%) ulcers in the polyurethane foam group. The between‐group difference was not statistically significant (RR 0.78, 95% CI 0.50 to 1.21) (Analysis 3.1). The sample comprised participants with ulcers of venous, mixed or other aetiology and wounds with medium and high levels of wound exudate. Participants receiving the hydrocapillary dressing had larger and more chronic ulcers at baseline compared with those in the polyurethane foam group (median values provided).
Secondary outcomes
Adverse events
In the RCT by Norkus 2005, research nurses assessed adverse events (odour, maceration, leakage, erythema and eczema) on a four‐point scale (none, mild, moderate or severe). In the hydrocapillary dressing group 8/49 (16%) experienced adverse events (four were defined as related to the dressing: two ulcer infections and two skin reactions; however, the remaining four participants’ adverse events were not described), compared with 11/48 (23%) in the polyurethane foam group (seven were defined as related to the dressing: two ulcer infections, three eczema, one allergic reaction, and one peri‐ulceration). The between‐group difference was not statistically significant (RR 1.40, 95% CI 0.62 to 3.18) (Analysis 3.2). All participants experiencing these adverse events withdrew from the RCT.
Five out of 49 participants (10%) in the hydrocapillary dressing group and 4/48 (8%) participants in the polyurethane foam group developed clinical signs of wound infection (all withdrew), but it was unclear whether any of the participants developing wound infection were amongst those withdrawing due to other adverse events. The trialists reported that one in three participants experienced some degree of maceration during the RCT, that severe maceration was only reported in the polyurethane foam group (5%), but that there were no significant between‐group differences in maceration, but did not report further data or P values. The trialists reported that there were 24 (1.7%) occurrences of severe erythema in the hydrocapillary dressing group compared with 45 (3.3%) in the polyurethane foam group, however, it was unclear whether these occurrences referred to dressing changes or something else, as the denominator was not reported. The trialists reported no significant between‐group difference.
Health related quality of life
Norkus 2005 assessed health related quality of life at the start and the end of the RCT using the WHO‐5 Well‐being Index. Responses to five questions were scaled from 0 to 5 and the total multiplied by four to give a maximum 100 value (the higher the score, the better the participant’s well being). At baseline, the median scores were 60 for the hydrocapillary dressing group and 48 in the polyurethane foam group; both groups achieved a median score of 68 at 12 months. The trialists reported that there were no significant differences in the initial or final values between‐groups (P value not reported).
Cost: wear time
Norkus 2005 reported that the estimated mean wear time was 3.2 days for both dressing groups, however, no other data were presented for this outcome.
Pain
In the RCT by Norkus 2005 wound pain at dressing removal, and at assessment visits between dressing changes, was reported by participants on a four‐point scale (none, mild, moderate or severe). There was no significant difference in the incidence of wound pain between the two groups, however, data by group and P‐values were not reported
Dressing performance: exudate management and adherence to the wound bed
In the RCT by Norkus 2005 dressing absorption was measured on a three‐point scale (good, moderate or poor). The RCT personnel were asked to estimate additional wear time based on saturation of the dressing. Adherence of the dressing to the wound bed was also assessed. The trialists reported that no significant differences in the incidence of little to moderate and severe leakage at dressing change were found between the two treatment groups, but that there was a significant difference in favour of the hydrocapillary dressing in the number dressing changes in which the dressing adhered to the wound. There was a discrepancy in the RCT report, however, regarding the denominator (the number of dressing changes) for these two outcomes, and it was unclear what the total number of dressing changes should have been in the RCT.
Hydrocapillary dressings compared with foam dressings: summary of results
Evidence from one RCT indicated that there was no statistically significant difference between foam dressings and hydrocapillary dressings in the proportion of ulcers completely healed at 12 months (Norkus 2005). This RCT was considered to have an overall unclear risk of bias.
Evidence from the same RCT indicated that there was no statistically significant difference between foam dressings and hydrocapillary dressings in the number of people who experienced adverse events (Norkus 2005). Insufficient data were available to estimate treatment effects for other outcomes.
Hydrocolloid dressings compared with foam dressings
We identified five RCTs that compared hydrocolloid dressings with foam dressings (Zuccarelli 1992; Bowszyc 1995; Thomas 1997; Charles 2002; Vanscheidt 2004).
Primary outcomes
Time to complete ulcer healing
Time to complete ulcer healing was reported in three RCTs that compared a hydrocolloid dressing with a foam dressing (Bowszyc 1995; Charles 2002; Vanscheidt 2004).
Bowszyc 1995 randomised 40 participants to a hydrocolloid dressing and 40 participants to a polyurethane foam dressing. A high‐compression elastic bandage was used with both dressings. The length of treatment was 16 weeks, or until the ulcer had healed. The trialists reported that the time to healing was 10.50 weeks in the foam group, compared with 9.34 weeks in the hydrocolloid group. The trial report did not state whether these values were means or medians. The trialists reported a P value of 0.35 for the between‐group difference. This RCT was considered to be at overall unclear risk of bias.
The RCT by Charles 2002 was a three‐armed RCT with 31 participants randomised to receive one hydrocolloid dressing (Granuflex), 29 participants randomised to receive another hydrocolloid dressing (Comfeel), and 31 who were randomised to a polyurethane foam dressing. All participants received short‐stretch compression bandaging. Length of treatment was 12 weeks, or until the ulcer healed. The trialists reported that the mean time to healing was 7.5 weeks and 6.0 weeks respectively in the two hydrocolloid groups, compared with 7.0 weeks in the polyurethane foam group. The trialists did not report the P value for between‐group difference, nor variance data. This RCT was considered to be at an overall high risk of bias.
Vanscheidt 2004 randomised 55 participants to receive a hydrocolloid dressing and 52 participants to receive a hydrocellular foam dressing. All participants received a high‐compression elastic bandage. Length of treatment was 12 weeks, or until healing of the ulcer occurred. The mean (SD) time to healing was reported as 66.0 (3.4) days in the hydrocolloid group compared with 72.6 (3.1) days in the foam dressing group; trialists reported a P value of 0.47 for the between‐group difference. This RCT was considered as being at unclear risk of bias overall.
Proportion of ulcers completely healed
Three of the RCTs evaluating hydrocolloid dressings compared with foam dressings reported number of ulcers healed at 12 weeks (Zuccarelli 1992; Charles 2002; Vanscheidt 2004).
Zuccarelli 1992 randomised 19 participants to receive a hydrocolloid dressing and 19 to receive a hydrocellular foam dressing. Compression therapy was in the form of a two‐layer system. This first layer was kept in place, whereas the second layer was removed by the participant before going to bed and re‐applied before getting up in the morning. The length of treatment was 12 weeks. At the end of the 12 weeks 9/19 (47%) ulcers had healed in both the hydrocolloid dressing group and the hydrocellular foam dressing group.
Charles 2002 reported that 17/29 (59%) ulcers in the Comfeel hydrocolloid dressing group and 17/31 (55%) in the Granuflex hydrocolloid dressing group had healed at 12 weeks, compared with 18/31 (58%) ulcers in the polyurethane foam group.
The number of ulcers that healed at 12 weeks in the hydrocolloid group in the RCT by Vanscheidt 2004 was 20/55 (36%) compared with 20/52 (38%) in the hydrocellular foam group.
Thomas 1997 randomised 50 participants to receive a hydrocolloid dressing and 50 to receive a polyurethane foam dressing. All participants received compression consisting of orthopaedic padding plus a high compression elastic bandage. The number of ulcers healed at 13 weeks in the hydrocolloid dressing group was 19/50 (38%) compared with 17/50 (34%) in the polyurethane foam group.
One RCT comparing a hydrocolloid dressing with a foam dressing reported the number of legs healed at 16 weeks (Bowszyc 1995). Bowszyc 1995 reported that 24/41 (59%) legs had healed at 16 weeks in the group receiving the hydrocolloid dressing compared with 24/41 (59%) legs in the group receiving the polyurethane foam dressing. The trialists reported that two of the 80 participants who were randomised had bilateral leg ulcers (one in each group), however, they did not report how many participants in each of the two groups had healed at 16 weeks.
We compared each of the hydrocolloid dressings evaluated by Charles 2002 in separate pooled analyses. The pooled estimates for the between‐group difference in complete ulcer healing from 12 to 16 weeks across the five RCTs were not statistically significant (including the Charles 2002 Comfeel group, RR 0.99, 95% CI 0.80 to 1.22; including the Charles 2002 Granuflex group, RR 1.00, 95% CI 0.81 to 1.24) (Analysis 4.1). Clinical advice was sought as to whether the two hydrocolloid dressings evaluated by Charles 2002 were similar enough to combine into a single treatment group for analysis (the method is detailed in the Cochrane Handbook Chapter 16.5 Higgins 2011b); this was confirmed as being appropriate. The pooled estimate for the between‐group difference across the five RCTs was not statistically significant (RR 1.00, 95% CI 0.81 to 1.22) (Analysis 4.1) (Figure 4). There was no evidence of statistically significant heterogeneity, the Chi2 P value being 0.99 and I2 estimation 0% for all three meta‐analyses.
Forest plot of comparison: 4 Hydrocolloid compared with foam, outcome: 4.1 Proportion of ulcers completely healed at 12 to 16 weeks.
Change in ulcer size
Two RCTs reported the mean percentage change in ulcer area at final follow up (Thomas 1997; Charles 2002). Thomas 1997 presented mean change estimates at 13 weeks of ‐78% for the group receiving the hydrocolloid dressing and ‐88% for those allocated foam (read from figure). Reviewer‐extrapolated estimates from Charles 2002 at 12 weeks were as follows: hydrocolloid (Comfeel) ‐82.1%; hydrocolloid (Granuflex) ‐83.6%; and foam dressing ‐67.3%. Both RCT reports indicated that differences between groups were not statistically significant, but no P values were provided.
Healing rate
Two RCTs mentioned undertaking assessment of healing rates (Zuccarelli 1992; Vanscheidt 2004). One presented only minimal information (Zuccarelli 1992), and reported that the mean ulcer size in both groups reduced steadily throughout the RCT, but no statistically significant between‐group difference was observed (P value 0.86). No rate of healing data by group were reported. Data from the second RCT indicated similar median rates of healing for hydrocolloid and foam dressings, respective values being ‐0.41 cm2/week and ‐0.43 cm2/week. The respective values for median percentage healing rates per week were ‐7.3% and ‐6.1%. The trial authors reported P values of 0.13 and 0.27 respectively, for the between‐group differences (Vanscheidt 2004).
Secondary outcomes
Adverse events
Zuccarelli 1992 did not report how adverse events were assessed, but reported that 2/19 (11%) participants in the hydrocolloid group experienced an allergic reaction and an intolerance in the peri‐ulcer skin area in the form of blisters, whereas no unwanted side effects were reported in the hydrocellular foam group.
Bowszyc 1995 reported that complications of secondary infection and maceration occurred, and that there were eight cases of streptococcal cellulitis that required antibiotic treatment (two participants withdrew because of this). Whilst the trialists reported that the same number of participants were affected in each group, they did not report the number of participants experiencing adverse events by treatment group. Furthermore, they did not describe the method of assessment of adverse events and it was unclear whether adverse events other than maceration or cellulitis had occurred.
Thomas 1997 reported that 7/99 (7%) participants in the hydrocolloid group and 10/100 (10%) in the foam group experienced adverse events related to the adhesive nature of the dressing such as minor trauma or erythema (skin redness); this report related to the overall sample of participants with leg ulcers and pressure ulcers and was not broken down by wound type. Additional adverse events noted by the trialists included maceration, bleeding from the wound area and excessive granulation; of those with leg ulcers, these affected six, one and three participants respectively in the hydrocolloid group, and none in the foam group.
The RCT by Charles 2002 did not describe how, or which, adverse events were assessed, but reported that no serious dressing‐related adverse events were observed during the RCT.
Vanscheidt 2004 recorded dressing‐related adverse events and severe adverse events deemed unrelated to the dressing, together with descriptions. In the hydrocolloid group 13/55 (24%) participants experienced one or more possible dressing‐related adverse event (19 events in total) compared with 15/52 (29%) participants in the hydrocellular foam group who experienced 18 possible dressing‐related events. The between‐group difference in participants experiencing adverse events was not statistically significant (RR 1.22, 95% CI 0.64 to 2.31) (Analysis 4.2). The trialists also reported on the occurrence of common adverse events (e.g. maceration, new wound development in different location, and non‐specific wound events), however, there were discrepancies between the numbers reported in the primary and secondary references. The trialists reported no statistically significant between‐group difference (P value 1.00), but it was not clear from the report if the difference referred to the number of participants with adverse events or the difference between the total number of adverse events in each group.
The adverse events assessment and reporting across these RCTs was considered too diverse for statistical pooling of data.
Pain
Zuccarelli 1992 evaluated pain using a numeric scale with lower scores representing less pain. Methods for collecting these data were not described. Data were presented graphically, using a range from 1.3 to 2.7 (full range of scores not stated). Reading from the graph, the mean pain score at week 12 was 1.5 in the group receiving the hydrocolloid dressing compared with 1.7 in the group receiving the hydrocellular foam dressing. No variance data were reported. The respective changes in mean scores from baseline (calculated from graph by review authors) were ‐0.8 and ‐0.9. The trialists reported that during the RCT there was a significant pain reduction in both dressing groups (P value 0.005) but that the between‐group difference was not statistically significant (P value 0.52).
Bowszyc 1995 assessed ulcer pain at dressing removal on a scale of 1 to 4 (1 = very painful and 4 = no pain) at the end of the RCT. The trialists reported a mean pain score (SD) of 3.63 (0.83) in the group receiving the hydrocolloid dressing compared with a mean score of 3.72 (0.55) in the group receiving the polyurethane foam dressing. The trialists reported that the between‐group difference was not statistically significant (P value not reported). The mean change in scores from baseline with variance estimate was not reported.
Charles 2002 reported ulcer‐associated pain using a 10‐point visual analogue scale with lower scores indicating less pain. The mean score at the week 12 follow up was 0.64 in the hydrocolloid Comfeel dressing group (n = 28), 0.15 in the hydrocolloid Granuflex dressing group (n = 27), and 0.50 in the polyurethane foam group (n = 30). No variance data were reported. The changes in mean scores from baseline (calculated by review authors) were ‐3.19, ‐3.66 and ‐4.27 respectively. The trialists reported that between‐group differences in pain prevalence and severity were not statistically significant at any point in the RCT (P value not reported).
In the RCT by Vanscheidt 2004, the number of participants with a reduction in pain, no effect on pain, uncertain effect, unable to respond/no response, increased pain, and those with missing data were reported for the final assessment (data were collected using the Johns Hopkins Pain Rating Instrument). The trialists reported that there were no statistically significant between‐group differences (P value 0.10 for overall effect of dressing on participant‐reported ulcer pain).
Due to the varied assessment methods and outcome reporting of pain assessment in these RCTs, a between‐group difference in pain with 95% confidence intervals within or across these RCTs was not estimated.
Cost: dressing changes
Zuccarelli 1992 reported the mean number of dressing changes over the 12 weeks of treatment as being 26.9 in the hydrocolloid group compared with 19.5 in the hydrocellular foam dressing group. The trialists did not report variance data, but did report that the between‐group difference was not statistically significant (P value 0.14).
Bowszyc 1995 reported the UK Drug Tariff price per dressing (at October 1993) as GBP 2.08 for a 10 cm x 10 cm piece of the hydrocolloid dressing compared with GBP 0.92 for a 10 cm x 10 cm piece of the polyurethane foam dressing. The reported mean number of dressing changes per week in the first month was 1.5 in the hydrocolloid group compared with 1.6 in the foam group. The trialists reported no variance data nor a statistically significant between‐group difference (P value 0.49).
The mean (SD) number of dressing changes per week in the RCT by Vanscheidt 2004 was 11.4 (6.5) in the group receiving the hydrocolloid dressing compared with 10.1 (6.3) in the group receiving the hydrocellular foam dressing. The trialists reported that there was no statistically significant between‐group difference in this dressing performance outcome, however, the number of participants in each treatment group at follow up was not reported.
We could not estimate between‐group differences in the number of dressing changes with 95% confidence intervals for foam dressings compared with hydrocolloid dressings from these RCTs.
Cost: wear time
Thomas 1997 reported that the mean time that the original dressing was in place prior to the first change was 2.7 days in the group receiving the hydrocolloid dressing compared with 2.8 days in the group receiving the polyurethane foam dressing. The trialists reported no variance data, nor the P value for between‐group difference.
Vanscheidt 2004 reported that the mean (SD) wear time was 5.6 (1.3) days in the hydrocolloid group compared with 5.6 (1.2) days in the hydrocellular foam group. The trialists reported that there was no statistically significant between‐group difference in this outcome, however, they did not report the number of participants in each treatment group at follow up.
Dressing performance: exudate management
Thomas 1997 reported that 27/50 (54%) of participants in the hydrocolloid group had dressings removed at the first dressing change due to leakage compared with 7/50 (14%) in the polyurethane foam group. The trialists reported that the between‐group difference was statistically significant (P value < 0.0001).
In the RCT by Vanscheidt 2004, the exudate‐handling property of the dressing was reported in terms of the number of dressing changes where no exudate, minimal exudate (i.e. less than 25% of dressing saturated), moderate exudate (25% to 75% of dressing saturated), or considerable exudate (more than 75% of dressing saturated) was observed. The proportion of dressing changes in each category appeared similar between groups (approximately 8%, 35%, 38% and 19% respectively), however, the trialists did not report a P value for the between‐group differences.
Ease of dressing removal/adherence to the wound bed
The RCT by Bowszyc 1995, assessed nurse‐rated ease of dressing removal using an 11‐point scale (0 = easy to remove, 11 = difficult to remove). The mean (SD) score in the hydrocolloid group was 1.42 (0.54) compared with 1.15 (0.42) in the polyurethane foam group. The trialists reported a P value of 0.016 for the between‐group difference.
Thomas 1997 reported that in both treatment groups, dressing removal was rated by nurses as 'easy' in the majority of cases at the first dressing change, however, no data by group were presented.
In the three‐arm RCT by Charles 2002, the foam dressing was reported to be significantly easier to remove compared with the two hydrocolloid dressings at six weeks and at the last dressing change (P value 0.021 and P value 0.037 respectively for the two time points), however, it was unclear how these P values related to the between‐group comparisons, given that the trialists would have been considering three comparisons at each time point. No between‐group difference with 95% confidence intervals could be estimated for this RCT.
Vanscheidt 2004 assessed participant‐rated ease of dressing removal and non‐traumatic dressing removal on a five‐point scale (1 = excellent, 2 = very good, 3 = good, 4 = fair, 5 = poor) at the final evaluation. The proportions of participants reporting dressing changes in each of the five categories for each outcome by treatment group were presented graphically, however, the denominator (number of participants in each treatment group) was not presented. The trialists reported a between‐group difference of P value 0.19 for ease of removal and P value 0.07 for non‐traumatic dressing removal. No between‐group differences with 95% confidence intervals could be estimated from this RCT.
Hydrocolloid dressings compared with foam dressings: summary of results
Pooled evidence from five RCTs indicated no statistically significant difference between hydrocolloid dressings and foam dressings in the proportion of ulcers completely healed from 12 to 16 weeks (Zuccarelli 1992; Bowszyc 1995; Thomas 1997; Charles 2002; Vanscheidt 2004). One of these RCTs was considered to be at overall high risk of bias (Charles 2002), whilst the others were judged as unclear.
Evidence from one RCT indicated no statistically significant between‐group difference in the number of participants experiencing dressing‐related adverse events. Insufficient data were provided to estimate measures of treatment effect for other outcomes, however, it is possible that the foam dressing was better than hydrocolloid for exudate handling and ease of removal.
Knitted viscose dressings compared with foam dressings
We identified one RCT that compared a knitted viscose dressing with a hydrocellular polyurethane foam dressing (Callam 1992). This RCT had a factorial design in which participants were randomised to one of two compression bandage systems as well as to one of two dressings. One of the compression systems consisted of orthopaedic padding, a high‐compression elastic bandage and a graduated compression tubular bandage; the other comprised orthopaedic padding, a short‐stretch crepe bandage, and a short‐stretch cohesive bandage. Sixty‐six participants were randomised to receive the knitted viscose dressing and 66 to receive the foam dressing. There was no statistically significant interaction between dressings and bandages (P value 0.87). The length of treatment was 12 weeks, or until the ulcer healed, whichever occurred sooner.
Primary outcomes
Time to healing
Callam 1992 reported no significant difference between groups from a Cox regression model adjusted for baseline ulcer size (P value 0.08 but no hazard ratio estimate provided).
Proportion of ulcers completely healed
At the end of 12 weeks of treatment Callam 1992 reported that 23/66 (35%) of ulcers had healed in the knitted viscose group compared with 31/66 (47%) in the hydrocellular foam group. The between‐group difference was not statistically significant (RR 1.35, 95% CI 0.89 to 2.05) (Analysis 5.1). The trialists considered participants who withdrew from treatment as failures of treatment rather than lost to follow up. This RCT was at overall high risk of bias.
Change in ulcer size
Callam 1992 reported a categorical analysis for percentage change in ulcer area from baseline to 12 weeks or participant withdrawal. Seven categories were used: more than 100% increase, 50% to 100% increase, 0% to 50% increase, 0% to 50% decrease, 50% to 100% decrease, healed and not known; data were presented in a figure. Initial or final ulcer tracings were not available for three participants, and these were excluded from the analysis. A P value of 0.051 was reported by the trialists for the overall between‐group difference in ulcer area reduction.
Secondary outcomes
Adverse events
Callam 1992 reported that there were no cases of adverse reactions in the knitted viscose dressing group and that 2/66 (2%) of the hydrocellular foam group experienced local skin reactions. The between‐group difference was not statistically significant (RR 5.00, 95% CI 0.24 to 102.19) (Analysis 5.2).
Pain
Callam 1992 assessed the number of participants experiencing pain at 0%, 1% to 25%, 26% to 50%, 51% to 75%, 76% to 99%, and 100% of clinic visits. The trialists reported P value 0.01 in favour of the foam dressing for the between‐group difference in patients experiencing pain at all (100%) clinic visits.
Knitted viscose dressings compared with foam dressings: summary of results
Evidence from one RCT suggested no statistically significant difference between knitted viscose dressings and foam dressings in the proportion of ulcers completely healed at 12 weeks (Callam 1992). This RCT was considered to be at overall high risk of bias. Data from the same RCT indicated no statistically significant between‐group difference in the number of participants who experienced adverse events during treatment. Insufficient data were available to estimate measures of treatment effect for other outcomes, however, it is possible that the foam dressing is associated with less pain than the knitted viscose dressing.
Protease‐modulating matrix dressing compared with foam dressing
We identified one three‐armed RCT that compared a protease‐modulating dressing, a paraffin gauze dressing and a foam dressing (Andriessen 2009). The comparison between the paraffin gauze and foam dressings has been described earlier.
Primary outcomes
Change in ulcer size
Andriessen 2009 reported that the mean percentage change in ulcer area (range) at four weeks was ‐31.8% (‐28 to ‐34%) in the protease‐modulating matrix group compared with ‐26.4% (‐17.3 to ‐32%) in the polyurethane foam group. The trialists did not report a P value for the between‐group differences. No other variance data were reported. A mean between‐group difference with 95% CI could not be estimated for this outcome.
Secondary outcomes
Adverse events
Andriessen 2009 recorded the severity and timing of any adverse event and reported that no adverse events were observed during the RCT.
Pain
Using data generated from a 10‐point VAS (with lower scores representing less pain), Andriessen 2009 reported that participants experienced little or no pain (classified from a mean VAS score of 1 to 2) in both groups.
Dressing performance: adherence to the wound bed
Andriessen 2009 reported that the paraffin gauze dressing stuck to the wound bed in 72% of the dressing changes (total number of dressing changes not provided), but there was no report of this outcome for the foam or protease‐modulating dressing groups.
Protease‐modulating matrix dressings compared with foam dressings: summary of results
A single RCT reported this comparison (Andriessen 2009); available data were very limited and therefore inconclusive.
Summary of Findings tables
We have included a Summary of Findings table for the following treatment comparisons; these provide a concise overview and synthesis of the volume and quality of available evidence.
-
polyurethane foam dressings compared with hydrocellular foam dressings (summary of findings Table for the main comparison);
-
foam dressings compared with paraffin gauze dressings (summary of findings Table 2);
-
foam dressings compared with hydrocapillary dressings (summary of findings Table 3);
-
foam dressings compared with hydrocolloid dressings (summary of findings Table 4), and;
-
foam dressings compared with knitted viscose dressings (summary of findings Table 5).
There is no Summary of Findings table for foam compared with protease‐modulating matrix dressings because only limited data were reported in the single RCT evaluating this comparison (Andriessen 2009).
The Summary of Findings tables indicate that, where evidence is available, it is generally of low quality. The exception to this was for the outcome of complete healing (at 12 to 16 weeks) for the comparison of foam and hydrocolloid dressings, where the evidence was of moderate quality. Overall, the Summary of Findings tables indicated that there is no evidence of a benefit of using foam dressings compared with alternative dressings for treating venous leg ulcers.
Discussion
The purpose of this review was to determine the effects of foam dressings compared with alternatives on the healing of venous leg ulcers. In addition, we considered the effects of these dressing on: adverse events, health‐related quality of life, costs, pain, and dressing performance (management of wound exudate and ease of removal).
Summary of main results
Twelve RCTs were included in this review. Three compared two types of foam dressings (Weiss 1996; Andersen 2002; Franks 2007), whilst the others involved comparison between foam and other types of dressings. The comparator dressings were: paraffin gauze (Banerjee 1990; Andriessen 2009); protease‐modulating matrix (Andriessen 2009); hydrocapillary (Norkus 2005); hydrocolloid (Zuccarelli 1992; Bowszyc 1995; Thomas 1997; Charles 2002; Vanscheidt 2004); and knitted viscose (Callam 1992).
Primary outcomes
No statistically significant differences were detected between treatment groups for any healing outcome for comparisons of different variants of foam dressings, nor for foam compared with other types of dressings (paraffin gauze, hydrocapillary, hydrocolloid, knitted viscose and protease‐modulating matrix). The length of follow up varied across RCTs and ranged between four weeks and one year. Meta‐analysis was feasible for one comparison (foam compared with hydrocolloid) where five RCTs were pooled for the outcome of complete healing at 12 to 16 weeks: RR 1.00 (95% CI 0.81 to 1.22).
Secondary outcomes
Secondary outcomes included adverse events, health‐related quality of life, costs, pain and dressing performance; outcome definition varied across RCTs. Data were generally sparse for the outcomes of health‐related quality of life and costs/resource use.
Where the review authors were able to estimate measures of treatment effect from the available data, no statistically significant between‐group differences in the proportion of participants experiencing adverse events were detected when hydrocellular foam dressings were compared with polyurethane foam dressings (Franks 2007), or when foam dressings were compared with hydrocapillary (Norkus 2005), hydrocolloid (Vanscheidt 2004), or knitted viscose dressings (Callam 1992). Regarding dressing performance, data from one RCT suggested that the polyurethane foam dressing may be superior to hydrocellular foam for handling exudate (Andersen 2002), and other data indicated that foam dressings could be better than hydrocolloid for the same outcome plus ease of removal (Bowszyc 1995; Thomas 1997). One RCT described fewer reports of pain for the foam dressing when compared with knitted viscose dressings (Callam 1992), whilst another found no difference in pain scores between groups receiving hydrocellular foam and polyurethane foam dressings (Franks 2007). Otherwise, RCTs either reported no statistically significant differences between groups for these outcomes, or, more commonly, presented only limited information. The secondary outcome data should generally be viewed with caution because of the subjective (and non‐blind) nature of the assessments.
Overall completeness and applicability of evidence
Participant and intervention characteristics
The majority of participants were ambulatory and had venous leg ulceration confirmed with ABPI assessment; these characteristics are likely to approximate to patients seen in clinical practice. Several RCTs imposed limits to the baseline wound size, sometimes with the idea of recruiting participants with wound sizes which would not exceed the study dressing dimensions (Weiss 1996; Thomas 1997; Andersen 2002; Charles 2002; Vanscheidt 2004). Also, most RCTs listed clinical infection of the ulcer as an exclusion criterion. This may limit applicability to clinical practice, where people with a wide range of wound sizes and/or infected ulcers are likely to be encountered. A further factor which could limit external validity is that nine of the 12 RCTs included were published ten, or more, years ago. If there are differences in earlier versions of foam dressings relative to their more recent counterparts, differential responses to treatment may be observed in clinical practice relative to this body of evidence.
With one exception (Banerjee 1990), all of the included RCTs reported the use of compression therapy as part of the intervention, so we were unable to undertake our planned subgroup analysis for the concurrent presence versus absence of compression. The types of compression therapy differed across the RCTs and included both elastic and short‐stretch bandage systems with varying components, and compression stockings.
Primary outcomes
Time to healing was assessed in seven of the 12 included RCTs, however, the analysis method was unclear in one RCT and no data by group were reported (Norkus 2005). Five RCTs evaluated this outcome as a mean time to healing (Bowszyc 1995; Weiss 1996; Andersen 2002; Charles 2002; Vanscheidt 2004). This analysis approach would only account for those participants whose ulcers healed. Participants whose ulcers did not heal during the RCT (censored data) would not have been accounted for by this analysis method which could have resulted in a biased effect estimate (Deeks 2011). Only one RCT, comparing a hydrocellular foam dressing with a polyurethane foam dressing, reported survival analysis as hazard ratios for time to healing, reporting that there was no statistically significant between‐group difference over 24 weeks (Franks 2007). The limited way in which this outcome was analysed and reported across the other RCTs restricts any inference regarding the efficacy of foam dressings in terms of time to healing across this body of evidence.
With one exception (Andriessen 2009), all included RCTs reported the proportion of ulcers healed and we were able to estimate between‐group differences from these RCTs for this outcome. For three comparisons, estimates were based on a single RCT: foam versus paraffin gauze (Banerjee 1990); foam versus hydrocapillary dressing (Norkus 2005); and foam versus knitted viscose dressing (Callam 1992).
Due to limited reporting in four RCTs that evaluated change in ulcer size (Callam 1992; Weiss 1996; Andersen 2002; Andriessen 2009), specifically in terms of variance estimates, no estimation of between‐group differences could be undertaken. None of these RCTs adjusted for baseline ulcer size.
Secondary outcomes
All secondary outcomes varied greatly across the included RCTs in terms of outcome definition, methods of assessment and reporting quality. Lack of available data meant that the review authors were unable to estimate between‐group differences in many instances.
In terms of adverse events, only four RCTs provided sufficient data to enable estimation of between‐group differences (Callam 1992; Vanscheidt 2004; Norkus 2005; Franks 2007). Cost and resource use data (e.g. mean number of dressing changes) were infrequently reported and where mean values were provided, often lacked variance data, or P values for tests of between‐group differences, or both. The method and timing of assessments of pain was varied, and no estimates for between‐group differences could be undertaken by the review authors. Outcomes relating to dressing performance were assessed subjectively, using non‐validated methods. One RCT reported that dressings were weighed to measure the amount of exudate; whilst this is potentially a more objective method of assessment, no results were presented (Norkus 2005). A quality of life assessment was reported by only one RCT (Norkus 2005), with only limited data provided.
Quality of the evidence
Six of the 12 included RCTs were considered to be at overall high risk of bias (Banerjee 1990; Callam 1992; Weiss 1996; Andersen 2002; Charles 2002; Franks 2007), and the remaining six were classified as having unclear risk of bias overall (Zuccarelli 1992; Bowszyc 1995; Thomas 1997; Vanscheidt 2004; Norkus 2005; Andriessen 2009). Of those at high risk of bias, three compared two types of foam dressings (Weiss 1996; Andersen 2002; Franks 2007), one compared a paraffin‐impregnated gauze dressing with foam (Banerjee 1990), one compared hydrocolloid dressing with foam (Charles 2002), and one compared knitted viscose dressing with foam dressing (Callam 1992). In three of these, the outcome assessment was not blinded (Weiss 1996; Andersen 2002; Franks 2007), and in four not all participants who were randomised were accounted for in the analysis (Banerjee 1990; Callam 1992; Andersen 2002; Charles 2002). Missing outcome data, due to attrition (drop‐out) during the trial or exclusion of participants from the analysis can bias the effect estimate, and lack of blinding of participants or healthcare providers could bias the results by affecting the actual outcomes of the participants in the trial (Higgins 2011a). Two RCTs were described as pilot studies, and had very small sample sizes (Weiss 1996; Andriessen 2009). One RCT presented a unit of analysis issue as, whilst the participants were the unit of randomisation, the unit of analysis was legs, with some participants reported as having bilateral leg ulcers (Bowszyc 1995). An important principle of any RCT is that the analysis must take into account the level at which randomisation occurred and the number of observations in the analysis should match the number of ‘units’ that were randomised (Deeks 2011).
All of the RCTs identified for inclusion in this review were published prior to the current CONSORT (Consolidated Standards of Reporting Trials) guidelines for the reporting of RCTs (Schulz 2010). Key aspects of best practice in RCT design to minimise bias include a robust randomisation method, concealment of treatment group allocation, and, where possible, blinding of participants and trial personnel, and blinded outcome assessment; all of which should be clearly stated in the RCT report. The majority of the included RCTs in this review did not report the method of the random sequence generation or an adequate method of allocation concealment.
The Summary of Findings tables indicate that the current evidence base for the effects of foam dressings compared with alternatives on the healing of venous leg ulcers is mainly of low quality. The quality of available evidence for the outcome of complete healing at 12 to 16 weeks for the comparison of foam and hydrocolloid dressings was moderate. A Summary of Findings table was not generated for the comparison between foam and protease‐modulating matrix dressings because only limited data were reported (Andriessen 2009). This highlights a potential gap in the evidence base as there is currently much interest in the role of proteases in chronic wound management (Ovington 2007; Snyder 2012).
Potential biases in the review process
In addition to the electronic searches of bibliographic databases, the search for evidence for this review included handsearching, contact with trialists and manufacturers, and the searching of ongoing trials registers. Although this search strategy was comprehensive, the possibility of publication bias cannot be discounted. Nonetheless, given the high or unclear risk of bias of the RCTs identified for inclusion, coupled with the absence of any significant treatment effects on ulcer healing, it is considered unlikely that any additional unpublished data would contribute significantly to the overall findings of this review.
Agreements and disagreements with other studies or reviews
The evidence base to guide dressing choice suggests that there is no evidence to support foam dressings as being better or worse than other dressing treatments for the healing of venous leg ulcers. This observation is in agreement with the systematic review by Palfreyman 2007, who concluded that there were insufficient data available to draw strong conclusions about any one dressing type being more effective in healing venous leg ulcers; the review did not comment on foam dressings specifically. Only eight of the 12 RCTs of foam dressings included in this review were included in the review by Palfreyman 2007.
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included trials.
Risk of bias summary: review authors' judgements about each risk of bias item for each included trial.
Flow diagram of the study selection process
Forest plot of comparison: 4 Hydrocolloid compared with foam, outcome: 4.1 Proportion of ulcers completely healed at 12 to 16 weeks.
Comparison 1 Hydrocellular foam compared with polyurethane foam, Outcome 1 Proportion of ulcers completely healed.
Comparison 1 Hydrocellular foam compared with polyurethane foam, Outcome 2 Participants experiencing adverse events.
Comparison 2 Paraffin gauze compared with foam, Outcome 1 Proportion of ulcers healed at 17 weeks.
Comparison 3 Hydrocapillary compared with foam, Outcome 1 Proportion of ulcers healed at 12 months.
Comparison 3 Hydrocapillary compared with foam, Outcome 2 Participants experiencing adverse events.
Comparison 4 Hydrocolloid compared with foam, Outcome 1 Proportion of ulcers completely healed at 12 to 16 weeks.
Comparison 4 Hydrocolloid compared with foam, Outcome 2 Participants experiencing dressing‐related adverse events.
Comparison 5 Knitted viscose compared with foam, Outcome 1 Proportion of ulcers healed at 12 weeks.
Comparison 5 Knitted viscose compared with foam, Outcome 2 Participants experiencing adverse events.
| polyurethane foam dressing compared to hydrocellular foam dressing for venous leg ulceration | ||||||
| Patient or population: patients with venous leg ulceration | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No of Participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| Hydrocellular foam dressing | Polyurethane foam dressing | |||||
| Time to healing | Study population1 | RR 1.24 | 156 | ⊕⊕⊝⊝ | ||
| 617 per 1000 | 765 per 1000 | |||||
| Low1 | ||||||
| 465 per 1000 | 577 per 1000 | |||||
| High1 | ||||||
| 757 per 1000 | 939 per 1000 | |||||
| Proportion of participants with healed ulcers | Study population1 | RR 1.03 | 118 | ⊕⊕⊝⊝ | ||
| 300 per 1000 | 309 per 1000 | |||||
| Low1 | ||||||
| 198 per 1000 | 204 per 1000 | |||||
| High1 | ||||||
| 507 per 1000 | 522 per 1000 | |||||
| Proportion of participants with healed ulcers | Study population1 | RR 1.60 | 18 | ⊕⊕⊝⊝ | ||
| 500 per 1000 | 800 per 1000 | |||||
| Low1 | ||||||
| 465 per 1000 | 744 per 1000 | |||||
| High1 | ||||||
| 757 per 1000 | 1000 per 1000 | |||||
| Proportion of participants with healed ulcers | Study population1 | RR 1.08 | 156 | ⊕⊕⊝⊝ | ||
| 617 per 1000 | 667 per 1000 | |||||
| Low1 | ||||||
| 465 per 1000 | 502 per 1000 | |||||
| High1 | ||||||
| 757 per 1000 | 818 per 1000 | |||||
| Mean change in wound size, with adjustment for baseline size | See comment | See comment | Not estimable | 0 | See comment | Outcome not reported (2 RCTs reported mean change in ulcer area, with no variance estimate, and no adjustment for baseline area).<BR/> |
| Proportion of participants experiencing adverse events | 272 per 10008 | 307 per 1000 | RR 1.13 | 156 | ⊕⊕⊝⊝ | |
| Health‐related quality of life | See comment | See comment | Not estimable | 0 | See comment | Outcome not reported. |
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (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). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1 Note: lower risk of the outcome is less favourable (i.e. lower risk of healing) than higher risk. Estimates for baseline low and high risks of healing at 60 days (8 weeks) and 120 days (17 weeks) have been taken from a meta‐analysis of RCTs evaluating different types of compression. The low risk estimate is based on a subset of participants with larger baseline ulcer area (greater than 5 cm squared). The high risk estimate is based on a subset of participants with smaller baseline ulcer surface area (5 cm squared or smaller). Most participants received a simple, low‐adherent dressing plus four‐layer bandage (O'Meara 2007). | ||||||
| foam dressing compared to paraffin gauze dressing for venous leg ulceration | ||||||
| Patient or population: patients with venous leg ulceration | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No of Participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| Paraffin gauze dressing | Foam dressing | |||||
| Time to healing | See comment | See comment | Not estimable | 0 | See comment | Outcome not reported. |
| Proportion of participants with healed ulcers | Study population1 | RR 1.34 | 71 | ⊕⊕⊝⊝ | ||
| 229 per 1000 | 306 per 1000 | |||||
| Low1 | ||||||
| 465 per 1000 | 623 per 1000 | |||||
| High1 | ||||||
| 757 per 1000 | 1000 per 1000 | |||||
| Mean change in wound size, with adjustment for baseline size | See comment | See comment | Not estimable | 0 | See comment | Outcome not reported (2 RCTs reported change in mean or median ulcer area, with no adjustment for baseline area).<BR/> |
| Adverse events | See comment | See comment | Not estimable | 0 | See comment | Limited data reported. |
| Health‐related quality of life | See comment | See comment | Not estimable | 0 | See comment | Outcome not reported. |
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (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). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1 Note: lower risk of the outcome is less favourable (i.e. lower risk of healing) than higher risk. Estimates for baseline low and high risks of healing at 120 days (17 weeks) have been taken from a meta‐analysis of RCTs evaluating different types of compression. The low risk estimate is based on a subset of participants with larger baseline ulcer area (greater than 5 cm squared). The high risk estimate is based on a subset of participants with smaller baseline ulcer surface area (5 cm squared or smaller). Most participants received a simple, low‐adherent dressing plus four‐layer bandage (O'Meara 2007). | ||||||
| foam dressing compared to hydrocapillary dressing for venous leg ulceration | ||||||
| Patient or population: patients with venous leg ulceration | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No of Participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| Hydrocapillary dressing | Foam dressing | |||||
| Time to healing | See comment | See comment | Not estimable | 0 | See comment | Limited information reported. |
| Proportion of participants with healed ulcers | Study population1 | RR 0.78 | 97 | ⊕⊕⊝⊝ | ||
| 510 per 1000 | 398 per 1000 | |||||
| Low1 | ||||||
| 465 per 1000 | 363 per 1000 | |||||
| High1 | ||||||
| 757 per 1000 | 590 per 1000 | |||||
| Mean change in wound size, with adjustment for baseline size | See comment | See comment | Not estimable | 0 | See comment | Outcome not reported. |
| Proportion of participants experiencing adverse events | 163 per 10006 | 229 per 1000 | RR 1.40 | 97 | ⊕⊕⊝⊝ | |
| Health‐related quality of life | See comment | See comment | Not estimable | 0 | See comment | Limited information provided. |
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (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). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1 Note: lower risk of the outcome is less favourable (i.e. lower risk of healing) than higher risk. Estimates for baseline low and high risks of healing at 120 days (17 weeks) have been taken from a meta‐analysis of RCTs evaluating different types of compression. The low risk estimate is based on a subset of participants with larger baseline ulcer area (greater than 5 cm squared). The high risk estimate is based on a subset of participants with smaller baseline ulcer surface area (5 cm squared or smaller). Most participants received a simple, low‐adherent dressing plus four‐layer bandage (O'Meara 2007). | ||||||
| foam dressing compared to hydrocolloid dressing for venous leg ulceration | ||||||
| Patient or population: patients with venous leg ulceration | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No of Participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| Hydrocolloid dressing | Foam dressing | |||||
| Time to healing | See comment | See comment | Not estimable | 0 | See comment | Limited data reported. |
| Proportion of participants with healed ulcers | Study population1 | RR 1.0 | 418 | ⊕⊕⊕⊝ | ||
| 471 per 1000 | 471 per 1000 | |||||
| Low1 | ||||||
| 465 per 1000 | 465 per 1000 | |||||
| High1 | ||||||
| 757 per 1000 | 757 per 1000 | |||||
| Mean change in wound size, with adjustment for baseline size | See comment | See comment | Not estimable | 0 | See comment | Two RCTs reported change in wound area, but not with baseline adjustment.<BR/> |
| Proportion of participants experiencing adverse events | 236 per 10004 | 288 per 1000 | RR 1.22 | 107 | ⊕⊕⊝⊝ | |
| Health‐related quality of life | See comment | See comment | Not estimable | 0 | See comment | Outcome not reported. |
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (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). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1 Note: lower risk of the outcome is less favourable (i.e. lower risk of healing) than higher risk. Estimates for baseline low and high risks of healing at 120 days (17 weeks) have been taken from a meta‐analysis of RCTs evaluating different types of compression. The low risk estimate is based on a subset of participants with larger baseline ulcer area (greater than 5 cm squared). The high risk estimate is based on a subset of participants with smaller baseline ulcer surface area (5 cm squared or smaller). Most participants received a simple, low‐adherent dressing plus four‐layer bandage (O'Meara 2007). | ||||||
| foam dressing compared to knitted viscose dressing for venous leg ulceration | ||||||
| Patient or population: patients with venous leg ulceration | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No of Participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| Knitted viscose dressing | Foam dressing | |||||
| Time to healing | See comment | See comment | Not estimable | 0 | See comment | Limited information provided. |
| Proportion of participants with healed ulcers | Study population1 | RR 1.35 | 132 | ⊕⊕⊝⊝ | ||
| 348 per 1000 | 470 per 1000 | |||||
| Low1 | ||||||
| 311 per 1000 | 420 per 1000 | |||||
| High1 | ||||||
| 696 per 1000 | 940 per 1000 | |||||
| Mean change in wound size, with adjustment for baseline size | See comment | See comment | Not estimable | 0 | See comment | Limited information provided. |
| Proportion of participants experiencing adverse events | RR 5.00 | 132 | ⊕⊕⊝⊝ | Assumed and corresponding risks not estimated because of zero event rate in knitted viscose group (3% in foam dressing group). | ||
| Health‐related quality of life | See comment | See comment | Not estimable | 0 | See comment | Outcome not reported. |
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (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). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1 Note: lower risk of the outcome is less favourable (i.e. lower risk of healing) than higher risk. Estimates for baseline low and high risks of healing at 90 days (12‐13 weeks) have been taken from a meta‐analysis of RCTs evaluating different types of compression. The low risk estimate is based on a subset of participants with larger baseline ulcer area (greater than 5 cm squared). The high risk estimate is based on a subset of participants with smaller baseline ulcer surface area (5 cm squared or smaller). Most participants received a simple, low‐adherent dressing plus four‐layer bandage (O'Meara 2007). | ||||||
| Time to healing weeks ‐ mean (SD): Group 1 (hydrocellular foam): 5.0 (1.7); Group 2 (polyurethane foam): 5.2 (1.9).
Number (%) of ulcers healed at 8 weeks: Group 1 (hydrocellular foam): 18/60 (30); Group 2 (polyurethane foam): 18/58 (31).
Change in ulcer size – mean change in area cm2: Group 1 (hydrocellular foam): ‐3.77; Group 2 (polyurethane foam): ‐3.18. P value for between‐group difference not reported by trialists.
Change in ulcer size – mean percentage change in area (read from graph): Group 1 (hydrocellular foam): ‐62%; Group 2 (polyurethane foam): ‐64%. P value for between‐group difference not reported by trialists. | |
| Change in ulcer size – mean (range) percentage change in area at 4 weeks: Group 1 (paraffin gauze): ‐17.2% (‐31% to 4.6%); Group 2: (protease‐modulating matrix ): ‐31.8% (‐28% to ‐34%); Group 3 (polyurethane foam): ‐26.4% (‐17.3% to ‐32%). Comments: trialists reported that all participants were compliant with treatment. | |
| Number (%) of ulcers healed at 17 weeks: Group 1 (paraffin gauze): 8/35 (23); Group 2 (polyurethane foam): 11/36 (31).
Ulcer area cm2 at 17 weeks median (range): Group 1 (paraffin gauze): 4.4 (0.2 to 22), n = 12; Group 2 (polyurethane foam): 6.3 (0.2 to 108.2), n = 14. Trialists did not report the change in ulcer size.
Change in ulcer area at 17 weeks cm2‐ difference in medians (calculated by review authors): Group 1 (paraffin gauze): ‐7.0; Group 2 (polyurethane foam): ‐5.9. | |
| Time to healing weeks: Group 1 (hydrocolloid): 9.34; Group 2 (polyurethane foam): 10.50. Not reported whether values are mean or median. Trialists reported P value = 0.35 for between‐group difference.
Number (%) of legs healed at 16 weeks: Group 1 (hydrocolloid): 24/41 legs (59); Group 2 (polyurethane foam): 24/41 legs (59).
Comments: 7 participants were excluded due to protocol violations, 5 due to heavily‐exuding wounds and 2 due to streptococcal cellulitis. Breakdown by group not reported, and not reported whether exclusion was pre‐ or post‐randomisation. There were 8 cases of streptococcal cellulitis (4 in each group) that required antibiotic treatment; 2 participants were withdrawn from the trial as a result. | |
| Time to healing: The reported P value for between‐group difference in time to healing from a Cox survival model including the covariate of initial ulcer size was 0.08. No other data were reported.
Number (%) of ulcers healed at 12 weeks: Group 1 (Knitted viscose): 23/66 (35); Group 2 (hydrocellular foam): 31/66 (47).
Change in ulcer size: Trialists reported a greater reduction in percentage ulcer area in Group 2 compared with Group 1, but the difference was not statistically significant at P value = 0.051. No average data were presented by group.
Percentage of ulcers with change in ulcer area ‐ > 100% increase vs 50%‐100% increase vs 0‐50% increase vs 0‐50% decrease vs 50%‐100% decrease vs healed vs not known (read from graph): Group 1 (Knitted viscose): 8% vs 8% vs 14% vs 15% vs 19% vs 35% vs 1%; Group 2 (hydrocellular foam): 4% vs 4% vs 10% vs 12% vs 21% vs 47% vs 2%. P value = 0.051 reported by trialists for between‐group difference in ulcer area reduction. Comments: trialists reported that initial or final ulcer tracing was not available for 3 patients. | |
| Mean time to healing weeks: Group 1 (hydrocolloid, Comfeel): 6.0; Group 2 (hydrocolloid, Granuflex): 7.5; Group 3 (polyurethane foam): 7.0.
Number (%) of ulcers healed at 12 weeks: Group 1 (hydrocolloid, Comfeel): 17/29 (59); Group 2 (hydrocolloid, Granuflex): 17/31 (55); Group 3 (polyurethane foam): 18/31 (58).
Change in ulcer size – mean percentage change in ulcer area: Group 1 (hydrocolloid Comfeel): ‐82.1%; Group 2 (hydrocolloid Granuflex): ‐83.6%; Group 3 (polyurethane foam): ‐67.3%. Trialists reported the percentage of ulcer area remaining, stating that the between‐group difference was not statistically significant (P value not provided). The percentage change values presented here were extrapolated by the review authors. | |
| Number (%) of ulcers healed at 24 weeks: Group 1 (polyurethane foam): 50/75 (66.7); Group 2 (hydrocellular foam): 50/81 (61.7).
Cumulative healing rates at 12 weeks derived from Kaplan‐Meier survival curves: Group 1 (polyurethane foam): 50.7%; Group 2 (hydrocellular foam): 47.5%.
Hazard ratio (95% CI) estimation from Cox proportional hazards model: Initial model adjusted for dressings, bandages and study centres HR 1.48 (0.87 to 2.54), P value = 0.15. Further model with additional adjustment for baseline covariates HR 1.50 (0.86 to 2.62), P value = 0.16. Comment: in cases where the original ulcer closed but a new area developed on the same limb while the original ulcer was still present, the limb was considered to be open until this new area of ulceration had also closed (Franks 2004, see references listed under Franks 2007). | |
| Time to healing: Trialists reported that no statistically significant difference was found between Group 1 and Group 2 for time to healing, but no data by group, or P value for difference were presented. Unclear what analysis method was used (described as Kaplan‐Meier principle).
Number (%) of ulcers healed at 12 months: Group 1 (hydrocapillary): 25/49 (51); Group 2 (polyurethane foam): 19/48 (40). | |
| Number (%) of ulcers healed at 13 weeks: Group 1 (hydrocolloid): 19/50 (38); Group 2 (polyurethane foam): 17/50 (34).
Change in ulcer area ‐ mean percentage change in ulcer area (read from graph): Group 1 (hydrocolloid): ‐78%; Group 2 (polyurethane foam): ‐88%. Trial authors reported no statistically significant difference between groups, but no P value presented.
Number (%) participants with ulcers that reduced in area during trial period: Group 1 (hydrocolloid): 34/50 (68); Group 2 (polyurethane foam): 44/50 (88). | |
| Time to healing days ‐ mean (SD): Group 1 (hydrocolloid): 66 (3.4); Group 2 (hydrocellular foam): 72.6 (3.1). Trialists reported that there were no statistically significant between‐group differences (P value = 0.47).
Number (%) of ulcers healed at 12 weeks: Group 1 (hydrocolloid): 20/55 (36); Group 2 (hydrocellular foam): 20/52 (38). Slight discrepancy between text and table in trial report (values above taken from table). Text of report states proportion of ulcers healed in Group 1 38.2%, and Group 2 38.5%.
Healing rate ‐ median change per week in cm2: Group 1 (hydrocolloid): ‐0.41; Group 2 (hydrocellular foam): ‐0.43. Trialists reported that there were no statistically significant between‐group differences (P value = 0.13).
Healing rate ‐ median percentage change per week: Group 1 (hydrocolloid): ‐7.3%; Group 2 (hydrocellular foam): ‐6.1%, Trialists reported that there were no statistically significant between‐group differences (P value = 0.27). | |
| Mean time to healing ‐ weeks: Group 1 (polyurethane foam): 5.6; Group 2 (hydrocellular foam): 6.5.
Number (%) of ulcers healed at 16 weeks: Group 1 (polyurethane foam): 8/10 (80); Group 2 (hydrocellular foam): 4/8 (50).
Change in ulcer size ‐percentage change (dimension not stated) at week 3: Group 1 (polyurethane foam): ‐27%; Group 2 (hydrocellular foam): ‐13%. P value for between‐group difference not reported by trialists. | |
| Number (%) of ulcers healed at 12 weeks: Group 1 (hydrocolloid): 9/19 (47); Group 2 (hydrocellular foam): 9/19 (47). Comment: trialists reported that the mean ulcer size in both groups reduced steadily throughout the trial, but no statistically significant between‐group difference was observed (P value = 0.86). |
| Number (%) participants experiencing adverse events: Group 1 (hydrocellular foam): Maceration/erythema/secondary infected dermatitis, 1 (2). Erysipelas, 1 (2). Serious adverse event: hospitalisation (pain), 1 (2); hospitalisation (cardiovascular accident), 1 (2); death (acute myocardial infarction), 1 (2). Group 2 (polyurethane foam): Maceration/erythema/secondary infected dermatitis, 1 (2). Pain after dressing change, 1 (2). Allergic reaction (dermatitis), 1 (2). Serious adverse event: no serious events reported. The trialists reported that they could not verify a relationship between the dressings and the adverse events (resulting in withdrawal from the study).
Average material costs per participant per week: Group 1 (hydrocellular foam): $ 18.99 ; Group 2 (polyurethane foam): $ 10.87. Included costs of dressings, unclear whether bandages and other treatment materials included, did not include nursing time, price year not stated, presumed currency is USD. Mean number of dressing changes per week: Group 1 (hydrocellular foam): 3.34; Group 2 (polyurethane foam): 2.14. The trialists reported a statistically significant difference (P value < 0.0005).
Pain at weekly assessments: Group 1 (hydrocellular foam): no pain was recorded at 169/319 (53%) weekly assessments; Group 2 (polyurethane foam): no pain was recorded at 226/379 (60%) weekly assessments. P value for between‐group differences not reported by trialists.
Pain ‐ comfort of the dressing: Participant ratings of comfort during treatment were reported as not significantly different between groups, but no data were presented.
Dressing performance ‐ absorbency: Group 1 (hydrocellular foam): dressing absorbency was rated as excellent in 12/170 (7%) of dressing changes; Group 2 (polyurethane foam): dressing absorbency was rated as excellent in 124/163 (76%) of dressing changes. The trialists reported a statistically significant difference (P value < 0.0005).
Dressing performance ‐ leakage: Group 1 (hydrocellular foam): leakage of exudate was observed in 198/309 (64%) of weekly assessments; Group 2 (polyurethane foam): leakage of exudate was observed in 172/355 (48%) of weekly assessments. The trialists reported a statistically significant difference (P value < 0.0005). | |
| Adverse events: The trialists stated that no adverse events were reported.
Pain at dressing removal: Group 1 (paraffin gauze): participants reported either moderate pain (mean VAS score of 4 to 6), or severe (mean VAS score of 7 to 9); Group 2: (protease‐modulating matrix dressing): participants reported little or no pain (mean VAS score of 1 to 2); Group 3 (polyurethane foam): participants reported little or no pain (mean VAS score of 1 to 2). Trialists report that the paraffin gauze dressing stuck to the wound bed in 72% of the dressing changes (no report of this outcome for other treatment groups). | |
| Adverse events ‐ number (%) participants died: Group 1 (paraffin gauze): 3/35 (9); Group 2 (polyurethane foam): 7/36 (19). Adverse events ‐ ulcer infection: Trialists reported that there were no statistically significant differences between groups in the proportion of patients with infected ulcers at weeks 1, 4, 8, 12 or 17, but further details were not provided. Pain: Trialists reported that the ulcer pain score was similar across groups with no statistically significant difference, but no data were presented.
Number of dressing changes per week: Trialists reported that more dressings and pads were used in Group 2 (polyurethane foam), but no data by group or P value for between‐group differences were reported.
Number of visits for healed/unhealed ulcers ‐ median: Group 1: (paraffin gauze) 6.5/25.0; Group 2: (polyurethane foam) 16.0/40.0. The trialists reported the following P values for between‐group differences: P < 0.05 for healed ulcers; and P < 0.002 for unhealed ulcers.
Nursing time spent per patient per week ‐ median: Group 1 (paraffin gauze): 4 hours 10 minutes Group 2 (polyurethane foam): 5 hours 40 minutes Trialists reported P value < 0.05 for between‐group difference. | |
| Number (%) patients experiencing adverse events: Not reported in full. Trialists reported that complications of secondary infection and maceration occurred: there were 8 cases of streptococcal cellulitis which required antibiotic treatment (2 patients withdrew because of this). The same number of patients were affected in each group.
Costs: UK Drug Tariff price per dressing (at October 1993) was reported as GBP 0.92 for 10 cm x 10 cm piece of foam dressing (Lyofoam) and GBP 2.08 for 10 cm x 10 cm piece of hydrocolloid dressing (Granuflex).
Pain ‐ dressing comfort score, mean (SD): Group 1 (hydrocolloid): 8.2 (2.5); Group 2 (polyurethane foam): 8.7 (2.2). Trialists reported no statistically significant between‐group difference (P value not provided).
Pain score at dressing removal, mean (SD): Group 1 (hydrocolloid): 3.63 (0.83); Group 2 (polyurethane foam): 3.72 (0.55). Trialists reported no statistical between‐group difference (P value not reported).
Mean number of dressing changes per week in first month: Group 1 (hydrocolloid): 1.5; Group 2 (polyurethane foam): 1.6. Trialists reported no statistically significant between‐group difference (P value = 0.49).
Nurse‐rated ease of removal – mean (SD) score from 11 point scale: Group 1: (hydrocolloid) 1.42 (0.54); Group 2 (polyurethane foam): 1.15 (0.42). Trialists reported P value = 0.016 for between‐group difference. | |
| Number (%) participants experiencing adverse events: Group 1 (knitted viscose): no adverse reactions; Group 2 (hydrocellular foam): local skin reaction, 2 (4). Number (%) participants experiencing skin complications ‐ sensitivity reaction to the dressing vs dry eczema vs wet eczema vs maceration vs other dressing complications (from supplied RCT report): Group 1 (knitted viscose): 3/66 (5) vs 47/66 (71) vs 16/66 (24) vs 3/66 (5) vs 5/66 (8); Group 2 (hydrocellular foam): 9/66 (14) vs 37/66 (56) vs 23/66 (35) vs 11/66 (17) vs 12/66 (18).
Pain at clinic visits: Number (%) participants experiencing pain at 0% vs 1%‐25% vs 26%‐50% vs 51%‐75% vs 76%‐99% and 100% of clinic visits: Group 1 (knitted viscose, n = 66 participants): 7 (11%) vs 8 (12%) vs 5 (8%) vs 5 (8%) vs 7 (11%) vs 34 (52%); Group 2 (hydrocellular foam, n = 66 participants): 13 (20%) vs 6 (9%) vs 9 (14%) vs 12 (18%) vs 9 (14%) vs 17 (26%). P value = 0.01 reported by trialists for between‐group difference in patients experiencing pain during 100% of clinic visits. Pain disturbing sleep: Number (%) participants reporting pain disturbing sleep at 0% vs 1%‐25% vs 26%‐50% vs 51%‐75% vs 76%‐99% and 100% of clinic visits: Group 1 (knitted viscose, n = 66 participants): 35 (53%) vs 16 (24%) vs 7 (11%) vs 1 (2%) vs 5 (8%) vs 2 (3%); Group 2 (hydrocellular foam, n = 66 participants): 34 (52%) vs 17 (26%) vs 6 (9%) vs 4 (6%) vs 0 (0%) vs 5 (8%). | |
| Adverse events: Trialists reported that no serious dressing‐related adverse events were observed during the trial.
Pain – percentage of participants reporting ulcer pain at 12 weeks (read from graph): Group 1 (hydrocolloid Comfeel): 8%; Group 2 (hydrocolloid Granuflex): 4%; Group 3 (polyurethane foam): 10%.
Pain ‐ mean ulcer‐associated VAS pain score at week 12: Group 1 (hydrocolloid, Comfeel): 0.64, n = 28; Group 2 (hydrocolloid, Granuflex): 0.15, n = 27; Group 3 (polyurethane foam): 0.50, n = 30. Trialists reported that between‐group differences in pain prevalence and severity were not statistically significant at any point in the trial (P value not reported). Change in mean scores (calculated by review authors): Group 1 (hydrocolloid, Comfeel): ‐3.19; Group 2 (hydrocolloid, Granuflex): ‐3.66; Group 3 (polyurethane foam): ‐4.27. Pain ‐ proportion of patients reporting mid to lower‐range pain severity on VAS and McGill Pain Questionnaire (MPQ‐PPI) at week 12 (secondary referenceCharles 2002): Trialists reported that the number of participants reporting pain fell to < 10%. Unclear whether this was overall, or for a specific treatment group. No data by group presented. Dressing performance ‐ ease of use of dressings: The foam dressing was significantly easier to remove compared with the two hydrocolloid dressings at 6 weeks and at last dressing (P value = 0.021 and P value = 0.037 respectively for the 2 time points; unclear how these P values relate to the between‐group comparisons given there would be 2 comparisons at each time point). | |
| Number (%) participants experiencing adverse events: Group 1 (polyurethane foam): 23/75 (30.7); Group 2 (hydrocellular foam): 22/81 (27.2). Trialists did not report the P value for between‐group difference.
Category of dressing‐related adverse event and number of occurrences: Group 1 (polyurethane foam): not related to dressing = 21; possibly related to dressing = 4; definitely related to dressing = 11; Group 2 (hydrocellular foam): not related to dressing = 15; possibly related to dressing = 4; definitely related to dressing = 11.
Number (%) participants who died during trial: Group 1 (polyurethane foam): 0/75 (0); Group 2 (hydrocellular foam): 1/81 (1.2).
Pre‐dressing pain at week 4 (VAS) – median (interquartile range): Group 1 (polyurethane foam): 1.5 (3.6), n = 53; Group 2 (hydrocellular foam): 1.3 (3.7) n = 53. The trialists reported P value = 0.18 for between‐group difference.
Post‐dressing pain at week 4 (VAS) – median (interquartile range): Group 1 (polyurethane foam): 1.0 (2.0), n = 51 Group 2 (hydrocellular foam): 0.5 (1.8), n = 52 Trialists reported P value = 0.12 for between‐group difference.
McGill Pain Questionnaire sensory pain scores at week 4 – median (interquartile range): Group 1 (polyurethane foam): 2.0 (4.0), n = 47; Group 2 (hydrocellular foam): 1.0 (3.0), n = 31. Trialists reported P value = 0.40 for between‐group difference.
McGill Pain Questionnaire affective pain scores at week 4– median (interquartile range): Group 1 (polyurethane foam): 0.0 (0.0), n = 42; Group 2 (hydrocellular foam): 0.0 (0.0), n = 47. Trialists reported P value = 0.69 for between‐group difference
Trialists reported that the levels of pain reduced significantly for both the VAS and McGill pain scores in both groups, however, the between‐group differences at week 4 were not statistically significant. | |
| Number (%) participants experiencing adverse events (type of adverse event): Group 1 (hydrocapillary dressing): 8/49 (16%) (4 adverse events defined as related to dressing, of which: ulcer infection = 2, skin reactions = 2, description of other types of adverse events not provided). All 8 participants withdrew from the trial. Group 2 (polyurethane foam): 11/48 (23%) (7 defined as related to dressing, of which: ulcer infections = 2, eczema = 3, allergic reaction = 1, peri‐ulceration = 1, description of other types of adverse events not provided). All 11 participants withdrew from the trial.
Number (%) participants developing clinical signs of wound infection (all withdrew): Group 1 (hydrocapillary dressing): 5/49 (10%); Group 2 (polyurethane foam): 4/48 (8%). It is unclear whether these participants with wound infection were also counted among those described above (adverse events).
Maceration: trialists reported that: 1 in 3 participants experienced some degree of maceration during the study; severe maceration was only reported in the polyurethane foam group (5%), and there were no significant between‐group differences in maceration, however, no further data were provided, neither were P values.
Number of times that signs of allergy on peri‐ulcer skin were reported: Group 1 (hydrocapillary dressing): ;7 Group 2 (polyurethane foam): 16. The trialists reported that allergic reactions were confirmed in 1 participant in each group by a positive patch test, and that there were no significant between‐group differences in allergy.
Number of times that severe erythema was reported: Group 1 (hydrocapillary dressing): 24 (1.7%); Group 2 (polyurethane foam): 45 (3.3%), Percentages as reported by trialists – denominators not stated. Trialists reported no significant between‐group difference in erythema.
Quality of life – median WHO‐5 Well‐Being Index scores: Group 1 (hydrocapillary): 68; Group 2 (polyurethane foam): 68. Trialists reported that there were no significant differences in the initial or final values between groups (P value not reported).
Wound pain between dressing changes: Trialists reported that there was no significant difference in the incidence of wound pain between the 2 groups. Data by group and P value not reported.
Pain – number (%) assessing dressing comfort as good vs medium vs poor: Group 1 (hydrocapillary): 42/43 (97.6) vs 1/43 (2.3) vs 0/43 (0); Group 2 (polyurethane foam): 32/45 (71.1) vs 8/45 (17.8) vs 5/45(11.1). Trialists reported that the comfort of the dressing in Group 1 was rated as significantly better than Group 2 (P value < 0.001).
Dressing performance, exudate handling – number (%) of incidents of little to moderate vs severe leakage at dressing change: Group 1 (hydrocapillary): 139 (11) vs 1 (0.1); Group 2 (polyurethane foam): 144 (13) vs 22 (2). Trialists reported that no significant differences in leakage were found between the 2 treatment groups. Denominators not reported.
Trialists reported that the hydrocapillary dressing was rated by nurses as having a significantly better capacity to absorb exudate compared with the polyurethane foam (P value < 0.05 – no other data provided).
Dressing performance, adherence – number (%) dressing changes where dressing adherence to the wound was reported: Group 1 (hydrocapillary dressing): 95/542 (18); Group 2 (polyurethane foam): 113/388 (29). Trialists reported P value < 0.05 for between‐group difference.
Percentage of dressing changes (as reported by trialists) where adherence was rated as none vs little vs moderate vs severe: Group 1 (hydrocapillary dressing): 82.5% vs 10.5% vs 6.6% vs 0.4%; Group 2 (polyurethane foam): 70.9% vs 12.6% vs 13.1 % vs 3.4%.
Mean wear time of dressing, days: Group 1 (hydrocapillary): 3.2; Group 2 (polyurethane foam): 3.2. | |
| Number (%) all participants (venous leg ulcers and pressure ulcers) experiencing adverse events linked to the dressing (described as due to adhesive nature of dressings causing minor trauma/erythema on removal): Group 1 (hydrocolloid): 7/99 (7); Group 2 (polyurethane foam): 10/100 (10). Numbers by ulcer type not reported. Number of participants with additional minor dressing related adverse events (venous leg ulcers only): Group 1 (hydrocolloid): maceration = 6; bleeding from the wound bed = 1; excessive granulation = 3; Group 2 (polyurethane foam): maceration = 0; bleeding from the wound bed = 0; excessive granulation = 0. Pain ‐ dressing comfort: Trialists reported that Group 2 (polyurethane foam) was significantly more likely to be scored as comfortable (comfortable vs otherwise, P value = 0.023) at the first dressing change. No data by group were presented.
Number participants with wound dehydration and dressing adherence to wound (all wound types): Group 1 (hydrocolloid): 0; Group 2 (polyurethane foam): 2. Group denominators not stated.
Dressing performance (exudate handling) – number (%) participants who had dressings removed at the first dressing change due to leakage: Group 1 (hydrocolloid): 27/50 (54); Group 2 (polyurethane foam): 7/50 (14). Trialists reported that the between‐group difference was statistically significant (P value < 0.0001). Dressing performance ‐ ease of removal: Trialists reported that dressing removal in both treatment groups was rated by nurses as 'easy' in the majority of cases at the first dressing change. No data by group were presented.
Mean time the dressing was in place prior to the first change, days: Group 1 (hydrocolloid): 2.7; Group 2 (polyurethane foam): 2.8. | |
| Number (%) participants experiencing adverse events: Group 1 (hydrocolloid): 13/55 (24) experienced 1 or more possibly dressing‐related adverse event (19 events in total); Group 2 (hydrocellular foam): 15/52 (29) experienced 18 possibly dressing‐related adverse events.
Common adverse events and number of participants affected: Group 1 (hydrocolloid): new wound development in different location = 6; other non‐specific wound events (e.g. affecting participant wounds other than the study wound) = 5; reason not reported = 2; Group 2 (hydrocellular foam): maceration = 6; new wound development in different location = 4; other non‐specific wound events (e.g. affecting participant wounds other than the study wound) = 5. Review authors' comment: numbers/types of adverse events per group were presented twice in the main text of the primary reference. Numbers of patients with different types of adverse events were discrepant between the 2 reports for Group 2. Trialists reported no significant between‐group difference (P value = 1.00, not clear if this referred to the difference in number of patients with adverse events or difference between total number of adverse events in group).
Serious adverse events and number of participants affected: Group 1 (hydrocolloid): angina = 1; musculoskeletal = 1; died = 3; Group 2 (hydrocellular foam): unspecified blood disease = 1.
Number (%) dressing changes with investigator‐rated condition of peri‐ulcer skin classified as normal vs maceration vs erythema vs eczema vs lipodermatosclerosis vs cellulitis vs dermatitis vs other: Group 2 (hydrocolloid): 269/615 (44) vs 135/615 (22) vs 52/615 (8) vs 5/615 (1) vs 119/615 (19) vs 0/615 (0) vs 16/615 (3) vs 19/615 (3); Group 1 (hydrocellular foam): 190/529 (36) vs 143/529 (27) vs 48/529 (9) vs 10/529 (2) vs 112/529 (21) vs 1/529 (<1) vs 16/529 (3) vs 9/529 (2). Denominators are total number of dressing changes. Trialists reported that no between‐group differences were statistically significant, but no P value provided.
Pain at final assessment, number (%) with reduction in pain vs no effect on pain vs uncertain vs unable to respond/no response vs increased pain vs missing data: Group 1 (hydrocolloid): 36/55 (65) vs 7/55 (13) vs 5/55 (9) vs3/55 (5) vs 4/55 (7) vs 0/55 (0); Group 2 (hydrocellular foam): 21/52 (40) vs 15/52 (29) vs 4/52 (8) vs 0/52 (0) vs 9/52 (17) vs 3/52 (6). Percentages for Group 2 recalculated by review authors to take account of missing data. Trialists reported that there were no statistically significant between‐group differences (P value = 0.10). Dressing performance (exudate handling) – number (%) participants with minimal or moderate levels of exudate at dressing change: Group 1 (hydrocolloid): 43/55(79); Group 2 (hydrocellular foam): 43/52 (83). Trialists reported that the between‐group difference was not statistically significant but P value not provided.
Dressing performance (exudate handling) – number (%) dressing changes with no exudate vs minimal exudate (< 25% of dressing saturated) vs moderate exudate (25%‐75% of dressing saturated) vs considerable exudate (> 75% of dressing saturated): Group 1 (hydrocolloid): 51/615 (8) vs 204/615 (33) vs 234/615 (38) vs 126/615 (20); Group 2 (hydrocellular foam): 42/529 (8) vs 199/529 (38) vs 196/529 (37) vs 89/529 (17). Trialists reported that the between‐group difference was not statistically significant but P value not provided.
Mean (SD) number of dressing changes per week: Group 1 (hydrocolloid): 11.4 (6.5); Group 2 (hydrocellular foam): 10.1 (6.3). Trialists reported that the between‐group difference was not statistically significant but P value not provided. Mean (SD) dressing wear time days: Group 1 (hydrocolloid): 5.6 (1.3); Group 2 (hydrocellular foam): 5.6 (1.2). Trialists reported that the between‐group difference was not statistically significant but P value not provided.
Dressing performance ‐ percentage of participants who rated non‐traumatic dressing removal categories of excellent/very good vsgood vsfair or poor (read from graph): Group 1 (hydrocolloid): 84% vs 11% vs 5% vs 0% Group 2 (hydrocellular foam): 68% vs 22% vs 4% vs 6% Trialists reported P value = 0.07 for between‐group differences. Dressing performance ‐ percentage of participants who rated ease of removal categories of excellent/very good vs good vs fair vs poor (read from graph): Group 1 (hydrocolloid): 84% vs 10% vs 4% vs 2%; Group 2 (hydrocellular foam): 74% vs 18% vs 6% vs 2%. Trialists reported P value = 0.19 for between‐group differences. | |
| Number (%) patients experiencing adverse events (type of adverse event): Group 1 (polyurethane foam): 0/10 (0); Group 2 (hydrocellular foam): 1/8 (38) ‐ allergic reaction which necessitated withdrawal.
Pain reduction and dressing comfort: Trialists stated that all participants reported a great reduction in pain, and that participants in Group 1 (polyurethane foam) preferred the comfort of the dressing. No data by group were presented. | |
| Number (%) patients experiencing adverse events: Group 1 (hydrocolloid): 2/19 (11) showed an allergic reaction and an intolerance in the peri‐ulcer skin area in the form of blisters; Group 2 (hydrocellular foam): no unwanted side effects were reported.
Mean (SD) number of dressing changes per week: Group 1 (hydrocolloid): 26.9; Group 2 (hydrocellular foam): 19.5 Trialists reported that the between‐group difference was not significant (P value = 0.14).
Mean pain scores over the course of treatment: Group 1 (hydrocolloid dressing): scores decreased during the first 2 weeks of treatment (P value = 0.005 for within group difference over time), increased slightly in the third week and then stabilised from then on. Group 2 (hydrocellular foam dressing): scores decreased during the first week of treatment (P value = 0.005 for within group difference over time) and remained stable from then on. Trialists reported no significant difference between groups.
Mean pain score at 12 weeks (read from graph): Group 1 (hydrocolloid): 1.5; Group 2 (hydrocellular foam): 1.7. Trialists reported that the between‐group difference was not significant (P value = 0.52). Change in mean scores (calculated by review authors from graph): Group 1 (hydrocolloid): ‐0.8; Group 2 (hydrocellular foam): ‐0.9. | |
| Abbreviation VAS = visual analogue scale | |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Proportion of ulcers completely healed Show forest plot | 3 | Risk Ratio (M‐H, Fixed, 95% CI) | Subtotals only | |
| 1.1 Ulcers healed at 8 weeks | 1 | 118 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.03 [0.60, 1.78] |
| 1.2 Ulcers healed at 16 weeks | 1 | 18 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.6 [0.75, 3.42] |
| 1.3 Ulcers healed at 24 weeks | 1 | 156 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.08 [0.85, 1.37] |
| 2 Participants experiencing adverse events Show forest plot | 1 | 156 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.13 [0.69, 1.85] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Proportion of ulcers healed at 17 weeks Show forest plot | 1 | 71 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.34 [0.61, 2.92] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Proportion of ulcers healed at 12 months Show forest plot | 1 | 97 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.78 [0.50, 1.21] |
| 2 Participants experiencing adverse events Show forest plot | 1 | 97 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.40 [0.62, 3.18] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Proportion of ulcers completely healed at 12 to 16 weeks Show forest plot | 5 | Risk Ratio (M‐H, Fixed, 95% CI) | Subtotals only | |
| 1.1 Meta‐analysis including Comfeel study arm from Charles 2002 | 5 | 387 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.99 [0.80, 1.22] |
| 1.2 Meta‐analysis including Granuflex study arm from Charles 2002 | 5 | 389 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.00 [0.81, 1.24] |
| 1.3 Meta‐analysis including Comfeel and Granuflex study arms combined from Charles 2002 | 5 | 418 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.00 [0.81, 1.22] |
| 2 Participants experiencing dressing‐related adverse events Show forest plot | 1 | 107 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.22 [0.64, 2.31] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Proportion of ulcers healed at 12 weeks Show forest plot | 1 | 132 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.35 [0.89, 2.05] |
| 2 Participants experiencing adverse events Show forest plot | 1 | 132 | Risk Ratio (M‐H, Fixed, 95% CI) | 5.0 [0.24, 102.19] |
