Intervention Review

You have free access to this content

Foam dressings for healing diabetic foot ulcers

  1. Jo C Dumville1,*,
  2. Sohan Deshpande2,
  3. Susan O'Meara3,
  4. Katharine Speak4

Editorial Group: Cochrane Wounds Group

Published Online: 6 JUN 2013

Assessed as up-to-date: 11 APR 2013

DOI: 10.1002/14651858.CD009111.pub3


How to Cite

Dumville JC, Deshpande S, O'Meara S, Speak K. Foam dressings for healing diabetic foot ulcers. Cochrane Database of Systematic Reviews 2013, Issue 6. Art. No.: CD009111. DOI: 10.1002/14651858.CD009111.pub3.

Author Information

  1. 1

    University of Manchester, Department of Nursing, Midwifery and Social Work, Manchester, UK

  2. 2

    Kleijnen Systematic Reviews, York, UK

  3. 3

    University of York, Department of Health Sciences, York, UK

  4. 4

    York, UK

*Jo C Dumville, Department of Nursing, Midwifery and Social Work, University of Manchester, Manchester, M13 9PL, UK. jo.dumville@manchester.ac.uk.

Publication History

  1. Publication Status: New search for studies and content updated (no change to conclusions)
  2. Published Online: 6 JUN 2013

SEARCH

 
Characteristics of included studies [ordered by study ID]
Baker 1993

MethodsSingle centre, two-arm RCT comparing a foam dressing (Allevyn, Smith & Nephew) with a calcium alginate dressing (Sorbsan, Aspen Medical) undertaken in the UK.
Duration of follow-up: Until the wound healed or for a maximum period of 12 weeks.


Participants20 participants
Inclusion criteria: Patients above 18 years of age with clean diabetic foot ulcers that were neuropathic in origin located on weight bearing areas of the foot. Patients who were able to give informed consent and willingly compliant to study protocol. Patients geographically able to comply with the study's demands.
Exclusion criteria: Patients with necrotic, sloughy ulcers or peripheral vascular disease, presence of infection in ulcerative foot, patients with history of poor compliance, patients unable to attend regularly, unable to follow simple instructions and those who could not comprehend to the nature of the trial.


InterventionsGroup A (n = 10): Foam dressing (Allevyn, Smith & Nephew). No secondary dressing was applied .
Group B (n = 10): Calcium alginate dressing (Sorbsan, Aspen Medical). A secondary low adherent absorbant dressing was used.
In both groups, dressings were cut to the required size but they did not overlap the wound margins by more than 3 cm or less than 0.5 cm. The dressings were secured by standard podiatric methods; e.g. padding and or strapping. Frequency of dressing changes depended on the quantity of exudate produced by the ulcers. If upon removal either dressing should appear to be stuck, they were to be irrigated as necessary with sterile saline.
Co-intervention: Ulcers were cleansed with warm sterile saline only and debridement was undertaken where required.


OutcomesPrimary outcome: Ulcer healing (number of ulcers healed at 12 weeks; median healing times).
Secondary outcome: Not reported.


NotesTrial data:  Analysis 4.1

Funding source not reported.


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskQuote: "Each subject will be randomly allocated to one of the two treatment groups either to the Allevyn or the Sorbsan group. This will be determined by the randomisation code which is computer generated." 
Comment: Method of generation of random schedule reported.   

Allocation concealment (selection bias)Unclear riskComment: The process of randomising participants, including who did this is not reported.

Blinding of participants and personnel (performance bias)
All outcomes
Unclear riskComment: No mention of blinding in study report.

Blinding of outcome assessment (detection bias)
All outcomes
Unclear riskComment: No mention of blinding in study report.

 

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskComment: One withdrawal (1/20 = 5%). Viewed as limited attrition.

Selective reporting (reporting bias)Low riskComment: Based on paper only, protocol not obtained.

Other biasUnclear riskComment: Funding source not reported.

Blackman 1994

MethodsSingle centre, two-arm RCT comparing a foam dressing (PolyMem, Ferris) with wet-to-dry saline gauze dressing undertaken in the USA.
Duration of follow-up: Until healed or until 6 months after the treatment was started. Five participants initially treated with saline gauze (Group B) conventional therapy CROSSED OVER to the polymeric membrane after 2 months. It is difficult to interpret 6 month healing data because of this cross over thus it has not been presented here.


Participants18 participants
However, four participants (two in each arm) were excluded under the criterion: ulcers progressed to Wagner grade 3 or higher. It is not clear if these were post-randomisation exclusions and in fact 22 participants were randomised.
Inclusion criteria: Diabetic patients (Type 1 and Type 2) with foot ulcers free of hard eschar.
Exclusion criteria: Participants needing vascular surgical therapy, participants with ulcers from Charcot joints, participants with ulcers of non-diabetic origin.


InterventionsGroup A (n = 11): Foam dressing (PolyMem, Ferris). Instructed to change dressing once daily as a minimum or when dressing was saturated. In keeping with manufacturers directions - those using the polymeric dressing were instructed not to use topical antibiotics of disinfectants.
Group B (n = 7): Wet-to-dry saline gauze dressings. Instructed to change dressing once daily as a minimum or when dressing was saturated.
Co-intervention: All participants were encouraged to obtain orthotic footwear and minimise weight bearing as much as possible.


OutcomesPrimary outcome: Ulcer healing (number of ulcers healed at 2 months; average size of ulcer at 2 months compared to baseline; substantial improvement noted in ulcer size).
Secondary outcomes: Not reported.


NotesTrial data:  Analysis 4.1

Source of funding: Ferris manufacturing corporation (Burr Ridge, IL).


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskQuote: "After qualifying for the study subjects were randomly assigned to conventional or polymeric membrane treatment".
Comment: Method of generating the random schedule not reported.

Allocation concealment (selection bias)Unclear riskComment: The process of randomising participants, including who did this is not reported.

Blinding of participants and personnel (performance bias)
All outcomes
Unclear riskComment: No mention of blinding in study report.

Blinding of outcome assessment (detection bias)
All outcomes
Unclear riskComment: No mention of blinding in study report.

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskComment: Four participants (2 in each group) were excluded under the criterion "ulcers progressed to Wagner stage 3 or higher" We are not clear if these participants were post-randomisation exclusion meaning that 22 participants were randomised.

Selective reporting (reporting bias)Low riskComment: Based on paper only, protocol not obtained.

Other biasUnclear riskComment: Some differences in baseline characteristics between groups e.g. mean age 51 years in Group A and 59 years in group Group B. Small sample size means trial is at high risk of chance imbalance. Funded by commercial organisation.

Clever 1995

MethodsTwo-arm RCT (not clear if single centre or multi-centred) comparing a foam dressing (Allevyn, Smith & Nephew) with a hydrocolloid (polyurethane matrix) dressing (Cutinova Hydro, Smith & Nephew, previously Beiersdorf) undertaken in Germany.
Duration of follow-up: Until healing occurred or for a maximum of 16 weeks.


Participants40 participants
Inclusion criteria: Patients aged 18 to 80 years with a pure neuropathic superficial ulcer 1 to 5 cm in diameter and with no clinical and radiological signs of osteomyelitis or tendon involvement.
Exclusion criteria: Patients an ABPI < 0.8 (measured using doppler ultrasound ) and with clinical or radiological signs of osteomyelitis or tendon involvement. Ulcers requiring topical treatment were also excluded, as were patients with know allergies to any product being used.


InterventionsGroup A (n = 20): Foam dressing (Allevyn, Smith & Nephew).
Group B (n = 20): Hydrocolloid (polyurethane matrix) dressing (Cutinova Hydro, Smith & Nephew).
In both groups, dressing changes were performed as often as required but at least once a week.
Co-intervention: Pressure relief comprising a half-shoe or so-called 'heal sandal', therapeutic footwear with cushioned insoles, and crutches as required to meet individual needs, infection control with systemic antibiotics if required, wound cleansing with Ringer's solution and debridement with removal of callus if needed.


OutcomesPrimary outcome: Ulcer healing (number of ulcers healed; mean time to healing; median time to healing; wound size at 4 weeks mm2).
Secondary outcomes: Adverse events; costs (mean number of dressing changes between clinical visits). Health-related quality of life; amputations, ulcer recurrence not reported.


NotesTrial data:  Analysis 4.1

Source of funding: Beiersdorf AG, Hamburg.


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskQuote: "Conducted an open, randomised, controlled study".
Comment: Method of generation of random schedule not reported.

Allocation concealment (selection bias)Unclear riskComment: The process of randomising participants, including who did this is not reported.

Blinding of participants and personnel (performance bias)
All outcomes
Unclear riskQuote: "Conducted an open, randomised, controlled study".
Comment: This was labelled an open trial not clear if blinded evaluation was conducted.

Blinding of outcome assessment (detection bias)
All outcomes
Unclear riskQuote: "Conducted an open, randomised, controlled study".
Comment: This was labelled an open trial not clear if blinded evaluation was conducted.

Incomplete outcome data (attrition bias)
All outcomes
High riskComment: In total six participants were withdrawn, or 15% of the total study population. The study report states that withdrawals were excluded from the analyses.

Selective reporting (reporting bias)Low riskComment: Based on paper only, protocol not obtained.

Other biasUnclear riskComment: Funded by commercial organisation.

Foster 1994

MethodsTwo-arm RCT (not clear if single centre or multi-centred) comparing a foam dressing (Allevyn, Smith & Nephew) with a calcium-alginate dressing (Kaltostat, ConvaTec ) undertaken in the UK.
Duration of follow-up: Until ulcer healed or for a maximum of 8 weeks.


Participants30 participants
Inclusion criteria: Patients above the age of 18 years with clean diabetic foot ulcers and who were willing and able to comply with study protocol.  
Exclusion criteria: Ulcer was sloughy, necrotic or infected.


Interventions30 participants
Group A (n = 15): Foam dressing (Allevyn, Smith & Nephew). Where surrounding skin was in good condition the dressing was secured with hypo-allergic tape. If the skin was atrophic or fragile then no tape was applied but a conforming bandage was used to secure the dressing. To apply the dressing to the patients' lesser toes, a strip of polyurethane foam dressing was doubled over and fastened  at the sides to form a sleeve that fitted over the toe.   
Group B (n = 15): Calcium-alginate dressing (Kaltostat, ConvaTec). The dressing was moistened with saline. A perforated film absorbent dressing was used as a secondary dressing, as before this was secured with hypo-allergic tape or a conforming bandage, depending on the state of the skin.         
Co-interventions: None reported.       


OutcomesPrimary outcome: Ulcer healing (number of ulcers healed; ulcers improved; median time to healing).
Secondary outcomes: Adverse events. Health-related quality of life; amputations; costs; amputations, ulcer recurrence not reported.  


NotesThe dressing performance parameters such as patient comfort and ease of removal were assessed using the three-point graded categorical scores. For each parameter the mean category score for each patient over the repeated dressing assessment was calculated. These data were not extracted as this approach has not been validated and does not facilitate comparison between studies.
Trial data:  Analysis 4.1

Funding source not reported.


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskQuote: "Thirty out patients entered study and 15 were randomised to each dressing".
Comment: Method of generation of random schedule not reported.

Allocation concealment (selection bias)Unclear riskComment: The process of randomising participants, including who did this is not reported.

Blinding of participants and personnel (performance bias)
All outcomes
Unclear riskComment: No mention of blinding in study report.

Blinding of outcome assessment (detection bias)
All outcomes
Unclear riskComment: No mention of blinding in study report.

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskComment: Four participants were withdrawn from the alginate group (4/30 = 13%). There were no withdrawals from the foam group. It is not clear how data from the withdrawn participants were used.

Selective reporting (reporting bias)Low riskComment: Based on paper only, protocol not obtained.

Other biasUnclear riskComment: Some differences in baseline characteristics between groups e.g. mean age 61 years in Group A and 70 years in group Group B. Small sample size means trial is at high risk of chance imbalance.
Funding source not reported.

Mazzone 1993

MethodsTwo-arm RCT (not clear if single or multi-centred) comparing a foam dressing (PolyMem, Ferris) with wet-to-dry saline gauze dressing undertaken in the USA.     
Duration of follow-up: 8 weeks.                                      


Participants19 participants
Inclusion criteria: Diabetic foot ulcer. 
Exclusion criteria: Not reported.


InterventionsGroup A (n = 11): Foam membrane dressing (PolyMem, Ferris). No other details.
Group B (n = 8): Wet-to-dry saline gauze dressings. No other details.
Co-interventions: None reported.


OutcomesPrimary outcome: Ulcer healing (number of ulcers healed; % reduction in wound size).
Secondary outcomes: Not reported.


NotesTrial data:  Analysis 4.1

Conference abstract.
Funding source: Ferris manufacturing corporation (Burr Ridge,IL).


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskQuote: "..were randomly assigned…"
Comment: Method of generation of random schedule not reported.

Allocation concealment (selection bias)Unclear riskComment: The process of randomising participants, including who did this is not reported.

Blinding of participants and personnel (performance bias)
All outcomes
Unclear riskComment: No mention of blinding in study report.

Blinding of outcome assessment (detection bias)
All outcomes
Unclear riskComment: No mention of blinding in study report.

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskComment: Not clear (conference abstract).

Selective reporting (reporting bias)Unclear riskComment: Based on conference abstract, protocol not obtained.

Other biasUnclear riskComment: Funded by commercial organisation.

Roberts 2001

MethodsTwo-arm  RCT (not clear if single or multi-centred) comparing a foam dressing (Allevyn, Smith & Nephew) with saline-soaked low adherent wound contact dressings (Tricotex, Smith & Nephew) undertaken in the UK.
Duration of follow-up: 13 weeks.     


Participants30 participants
Inclusion criteria: Type 1 diabetes with neuropathic ulcer.              
Exclusion criteria: ABPI < 0.8.


InterventionsGroup A (n = 14): Foam dressing (Allevyn, Smith & Nephew).
Group B (n = 16): Saline-soaked low adherent wound contact dressings (Tricotex, Smith & Nephew).
Dressings were changed weekly.
Co-interventions: Standard podiatric care (no other details given).


OutcomesPrimary outcome: Ulcer healing (number of ulcers healed; % ulcers reduced by 50% in size)
Secondary outcomes: Not reported.


NotesTrial data:  Analysis 4.1

Conference abstract. 
Time to healing reported as not significantly different but values not reported.
Funding source: Smith and Nephew.    


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskComment: Described in title as randomised controlled trial but method of generation of random schedule not reported.

Allocation concealment (selection bias)Unclear riskComment: The process of randomising participants, including who did this is not reported.

Blinding of participants and personnel (performance bias)
All outcomes
Unclear riskComment: No mention of blinding in conference abstract.

Blinding of outcome assessment (detection bias)
All outcomes
Unclear riskComment: No mention of blinding in study report.

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskComment: Not clear (conference abstract).

Selective reporting (reporting bias)Unclear riskComment: Based on conference abstract, protocol not obtained.

Other biasUnclear riskComment: Funded by commercial organisation.

 
Characteristics of excluded studies [ordered by study ID]

StudyReason for exclusion

Agas 2006Study did not randomise participants.

Ahroni 1993The dressing groups evaluated in this study were not foam dressings.

Altman 1993No single, identifiable dressing type evaluated.

Alvarez 2003The dressing groups evaluated in this study were not foam dressings.

Apelqvist 1990The dressing groups evaluated in this study were not foam dressings.

Apelqvist 1996No single, identifiable dressing type evaluated.

Apelqvist 2004No single, identifiable dressing type evaluated.

Armstrong 2004No single, identifiable dressing type evaluated.

Belcaro 2010The dressing groups evaluated in this study were not foam dressings.

Bogaert 2004Study did not randomise participants.

Bradshaw 1989Trial stopped after recruiting six participants. No data presented. Authors not contacted for healing data.

Caravaggi 2003Other intervention, not dressings, differ between trial arms.

Chang 2000Study did not include diabetic foot ulcers.

Chauhan 2003Other intervention, not dressings, differ between trial arms.

Chirwa 2010Study did not randomise participants.

Cuevas 2007No single, identifiable dressing type evaluated.

D'Hemecourt 1998The dressing groups evaluated in this study were not foam dressings.

Dash 2009Other intervention, not dressings, differ between trial arms

Diehm 2005Study did not randomise participants

Donaghue 1998The dressing groups evaluated in this study were not foam dressings

Driver 2006Other intervention, not dressings, differ between trial arms

Edmonds 2009Other intervention, not dressings, differ between trial arms

Eginton 2003No single, identifiable dressing type evaluated.

Etoz 2004Study did not randomise participants.

Farac 1999Author contacted: study not suitable for inclusion due to data quality issues.

Foo 2004The dressing groups evaluated in this study were not foam dressings.

Foster 1999Other intervention, not dressings, differ between trial arms.

Gao 2007Other intervention, not dressings, differ between trial arms.

Gentzkow 1996Other intervention, not dressings, differ between trial arms.

Gottrup 2011The dressing groups evaluated in this study were not foam dressings.

Hanft 2002Other intervention, not dressings, differ between trial arms.

Jeffcoate 2009The dressing groups evaluated in this study were not foam dressings.

Jeffery 2008Study did not randomise participants.

Jensen 1998The dressing groups evaluated in this study were not foam dressings.

Jude 2007The dressing groups evaluated in this study were not foam dressings.

Kordestani 2008The dressing groups evaluated in this study were not foam dressings.

Lalau 2002The dressing groups evaluated in this study were not foam dressings.

Landsman 2010Other intervention, not dressings, differ between trial arms.

Lazaro-Martinez 2007Other intervention, not dressings, differ between trial arms.

Lipkin 2003Other intervention, not dressings, differ between trial arms.

Markevich 2000The dressing groups evaluated in this study were not foam dressings.

Marston 2001Other intervention, not dressings, differ between trial arms.

McCallon 2000Study did not randomise participants.

Mody 2008Study did not include diabetic foot ulcers.

Moretti 2009Other intervention, not dressings, differ between trial arms.

Mueller 1989Other intervention, not dressings, differ between trial arms.

Mulder 1994The dressing groups evaluated in this study were not foam dressings.

Munter 2006Study included a range of wound types, with data not reported separately for diabetic foot ulcers. We were unsuccessful in contacting the authors to query the availability of relevant data.

Novinscak 2010No single, identifiable dressing type evaluated.

Ogce 2007The dressing groups evaluated in this study were not foam dressings.

Palao i Domenech 2008Study included a range of wound types, with data not reported separately for diabetic foot ulcers. The control arm of the study received 'local best practice' with no further information provided. The primary ou come was ulcer pain and the follow-up period was seven days only. We were unsuccessful in contacting the authors to query the availability of relevant healing data.

Parish 2009Other intervention, not dressings, differ between trial arms.

Pham 1999Other intervention, not dressings, differ between trial arms.

Piaggesi 1997Study did not randomise participants.

Piaggesi 2001The dressing groups evaluated in this study were not foam dressings.

Reyzelman 2009No single, identifiable dressing type evaluated.

Robson 2005Other intervention, not dressings, differ between trial arms.

Robson 2009Study did not include diabetic foot ulcers.

Sabolinski 2000Other intervention, not dressings, differ between trial arms.

Sabolinski 2001Other intervention, not dressings, differ between trial arms.

Shaw 2010Other intervention, not dressings, differ between trial arms.

Shukrimi 2008Other intervention, not dressings, differ between trial arms.

Sibbald 2011Study included a range of wound types, with data not reported separately for diabetic foot ulcers. We were unsuccessful in contacting the authors to query the availability of further information.

Solway 2011Study did not randomise participants.

Steed 1992Other intervention, not dressings, differ between trial arms.

Steed 1995Other intervention, not dressings, differ between trial arms.

Steed 1996Other intervention, not dressings, differ between trial arms.

Subrahmanyam 1993The dressing groups evaluated in this study were not foam dressings.

Trial 2010The dressing groups evaluated in this study were not foam dressings.

Turns 2012Study did not randomise participants.

Urbaneie-Rovan 1999No single, identifiable dressing type evaluated.

Vandeputte 1997The dressing groups evaluated in this study were not foam dressings.

Varma 2006No single, identifiable dressing type evaluated.

Veves 2001Other intervention, not dressings, differ between trial arms.

Veves 2002The dressing groups evaluated in this study were not foam dressings.

Whalley 2001The dressing groups evaluated in this study were not foam dressings.

Woo 2010Study included a range of wound types, with data not reported separately for diabetic foot ulcers. We were unsuccessful in contacting the authors to query the availability of further information.

Yao 2007Other intervention, not dressings, differ between trial arms.

Zimny 2003Other intervention, not dressings, differ between trial arms.

 
Comparison 1. Foam dressing compared with basic wound contact dressing

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Number of ulcers healed3Risk Ratio (M-H, Fixed, 95% CI)Subtotals only

 2 Number of ulcers healed - pooled data249Risk Ratio (M-H, Fixed, 95% CI)2.03 [0.91, 4.55]

 
Comparison 2. Foam dressing compared with alginate dressing

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Number of ulcers healed250Risk Ratio (M-H, Fixed, 95% CI)1.5 [0.92, 2.44]

 
Comparison 3. Foam dressing compared with hydrocolloid (matrix) dressing

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Number of ulcers healed1Risk Ratio (M-H, Fixed, 95% CI)Subtotals only

 
Comparison 4. Trial data

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Trial dataOther dataNo numeric data

 
Analysis 4.1 Comparison 4 Trial data, Outcome 1 Trial data.
Trial data

StudyGroupsPrimary outcome: ulcer healingSecondary: health-related quality of lifeNumber and level of amputationsAdverse events, including painCostUlcer recurrence

Baker 1993Group A (n = 10): Foam dressing
Group B (n = 10): Calcium alginate dressing
Number of ulcers healed at 12 weeks
Group A: 9
Group B: 4
Median healing time (days)
Group A: 28
Group B: Median time to healing was not reached by 84 days
Cox's proportional hazards model adjusted for initial ulcer size and duration of ulcer at baseline and treatment effect gave a hazard ratio of 4.04 in favour of foam dressing (95% CI 1.18 to 13.84).
n/rn/rn/rn/rn/r

Blackman 1994Group A (n = 11): Foam dressing
Group B (n = 7): Wet-to-dry saline gauze dressings.
Number of ulcers healed at 2 months
Group A: 3
Group B: 0
Average size of ulcer at 2 months compared to baseline (unclear if bracketed figures are sd)
Group A: 35% (16%)
Group B: 105% (26%
Substantial improvement noted in ulcer size (not defined)
Group A: 10
Group B: 2
n/rn/rn/rn/rn/r

Clever 1995Group A (n = 20): Foam dressing
Group B (n = 20): Hydrocolloid (polyurethane matrix) dressing
Number of ulcers healed:
Group A:14
Group B:16
Mean time to healing in days (sd)
Group A: 20.43 (14.74)
Group B: 25.19 (23.52)
Median time to healing in days
Group A: 16.5 (range 4 to 52)
Group B: 15.5 (range 4 to 76)
Wound size at 4 weeks mm2 (sd):
Group A: 33.46 (75.22)
Group B: 32.37 (54.12)
Number of adverse events (Reasons not reported separately for two groups):
Group A: 5
Group B: 1
Mean number of dressing changes between clinical visits (sd):
Group A: 2.37 (2.18)
Group B: 2.23 (2.19)

Foster 1994Group A (n = 15): Foam dressing  
Group B (n = 15): Calcium-alginate dressing
Ulcer healing:
Number of ulcers healed:
Group A: 9
Group B: 8
Ulcers improved (not defined):
Group A: 6
Group B: 3
Median time to healing (days) (K-M plot is presented but no median time to healing or HR from a Cox's Proportional Hazards analysis has been presented. The median time to healing was estimated by the reviewer author from the graph).
Group A: 40
Group B: 42
n/rn/rGroup A: 0
Group B: Severe pain = 1; Dressings plugged a plantar lesion, preventing free drainage of exudate = 3  (1 developed cellulitis) 
n/rn/r

Mazzone 1993Group A (n = 11): Foam membrane dressing (PolyMem, Ferris). No other details
Group B (n = 8): Wet-to-dry saline gauze dressings. No other details.
Ulcer healing
Number of ulcers healed:
Group A: 7
Group B: 2
% reduction in wound size:
Group A: 71
Group B: 29
n/rn/rn/rn/rn/r

Roberts 2001Group A (n = 14): Foam dressing
Group B (n = 16): Saline-soaked low adherent wound contact dressings .
Number of ulcers healed:
Group A: 6
Group B: 4
% ulcers reduced in area by 50% :
Group A: 93
Group B: 75
n/rn/rn/rn/rn/r

 
Summary of findings for the main comparison. Foam dressings compared to basic wound contact dressings for foot ulcers in people with diabetes

Foam dressings compared to basic wound contact dressings for foot ulcers in people with diabetes

Patient or population: patients with foot ulcers in people with diabetes
Settings:
Intervention: Foam dressings
Comparison: Basic wound contact dressings

OutcomesIllustrative comparative risks* (95% CI)Relative effect
(95% CI)
No of Participants
(studies)
Quality of the evidence
(GRADE)
Comments

Assumed riskCorresponding risk

Basic wound contact dressingsFoam dressings

Nunber of ulcers healed
Follow-up: mean 10 weeks
Low risk of healing1 RR 2.03
(0.91 to 4.55)
49
(2 studies)
⊕⊕⊝⊝
low2,3

340 per 1000690 per 1000
(309 to 1000)

Moderate risk of healing1

530 per 10001000 per 1000
(482 to 1000)

High risk of healing1

650 per 10001000 per 1000
(592 to 1000)

*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).
CI: Confidence interval; RR: Risk ratio.

GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

 1 Baseline risk of healing obtained from external source in which data from 27,630 patients with a diabetic neuropathic foot ulcer was used to develop a simple prognostic model to predict likelihood of ulcer healing (Margolis DJ, Allen-Taylor L, Hoffstad O, Berlin JA. Diabetic neuropathic foot ulcers: predicting which ones will not heal. Am J Med. 2003;115:627-31). It is important to note that given an outcome of ulcer healing, low risk refers to a low risk of healing and thus reflects the most severe patient populations. Conversely high risk refers to a high risk of healing.
2 The two studies reported insufficient information to make any judgement regarding quality of study design and associated risk of bias.
3 Very small samples sizes and short follow up times which limited the number of healing events resulted in large imprecision. 19 participants achieved the endpoint of healing in the two studies, this is an underpowered comparison. The confidence interval around the estimate of relative risk is consistent with a 9% relative reduction in healing with foam and a 450% relative increase in healing with foam.
 
Summary of findings 2. Foam dressings compared to alginate dressings for foot ulcers in people with diabetes

Foam dressings compared to alginate dressings for foot ulcers in people with diabetes

Patient or population: patients with foot ulcers in people with diabetes
Settings:
Intervention: Foam dressings
Comparison: Alginate dressings

OutcomesIllustrative comparative risks* (95% CI)Relative effect
(95% CI)
No of Participants
(studies)
Quality of the evidence
(GRADE)
Comments

Assumed riskCorresponding risk

Algiante dressingsFoam dressings

Number of ulcers healed
Follow-up: mean 10 weeks
Low risk of healing1 RR 1.50
(0.92 to 2.44)
50
(2 studies)
⊕⊕⊝⊝
low2

340 per 1000510 per 1000
(313 to 830)

Moderate risk of healing1

530 per 1000795 per 1000
(488 to 1000)

High risk of healing1

650 per 1000975 per 1000
(598 to 1000)

Adverse events
Follow-up: 8 weeks
See commentSee commentNot estimable0
(1 study)
See commentOne study reported limited adverse event data: no events in the foam-dressed group and four events in the alginate-dressed group.

*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).
CI: Confidence interval; RR: Risk ratio.

GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

 1 Baseline risk of healing obtained from external source in which data from 27,630 patients with a diabetic neuropathic foot ulcer was used to develop a simple prognostic model to predict likelihood of ulcer healing (Margolis DJ, Allen-Taylor L, Hoffstad O, Berlin JA. Diabetic neuropathic foot ulcers: predicting which ones will not heal. Am J Med. 2003;115:627-31). It is important to note that given an outcome of ulcer healing, low risk refers to a low risk of healing and thus reflects the most severe patient populations. Conversely high risk refers to a high risk of healing.
2 30 participants achieved the endpoint of healing in the two studies, this is an underpowered comparison. The confidence interval around the estimate of relative risk is consistent with a 8% relative reduction in healing with foam and a 244% relative increase in healing with foam.
 
Summary of findings 3. Foam dressings compared to hydrocolloid matrix dressings for foot ulcers in people with diabetes

Foam dressings compared to hydrocolloid matrix dressings for foot ulcers in people with diabetes

Patient or population: patients with foot ulcers in people with diabetes
Settings:
Intervention: Foam dressings
Comparison: Hydrocolloid matrix dressings

OutcomesIllustrative comparative risks* (95% CI)Relative effect
(95% CI)
No of Participants
(studies)
Quality of the evidence
(GRADE)
Comments

Assumed riskCorresponding risk

Hydrocolloid matrix dressingsFoam dressings

Number of ulcers healed
Follow-up: 16 weeks
Low risk of healing1 RR 0.88
(0.61 to 1.26)
40
(1 study)
⊕⊝⊝⊝
very low2,3

340 per 1000299 per 1000
(207 to 428)

Moderate risk of healing1

530 per 1000466 per 1000
(323 to 668)

High risk of healing1

650 per 1000572 per 1000
(397 to 819)

Adverse events
Follow-up: 16 weeks
Study populationNot estimable0
(1)
See commentLimited data from one study. Five events reported in the foam dressed group compared with one event in the hydrocolloid matrix group.

See commentSee comment

Moderate


*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).
CI: Confidence interval; RR: Risk ratio.

GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

 1 Baseline risk of healing obtained from external source in which data from 27,630 patients with a diabetic neuropathic foot ulcer was used to develop a simple prognostic model to predict likelihood of ulcer healing (Margolis DJ, Allen-Taylor L, Hoffstad O, Berlin JA. Diabetic neuropathic foot ulcers: predicting which ones will not heal. Am J Med. 2003;115:627-31). It is important to note that given an outcome of ulcer healing, low risk refers to a low risk of healing and thus reflects the most severe patient populations. Conversely high risk refers to a high risk of healing.
2 In total six participants were withdrawn, or 15% of the study population. The study states that withdrawals were excluded from the analysis.
3 30 participants achieved the endpoint of healing in the study, this is an underpowered comparison. The confidence interval around the estimate of relative risk is consistent with a 39% relative reduction in healing with foam and a 26% relative increase in healing with foam.
 
Table 1. Summary of studies

First AuthorGroup A Group B Duration of follow-up% healed data

Baker 1993Foam dressing (Allevyn,Smith & Nephew)Calcium-alginate dressing (Sorbsan, Aspen Medical)12 weeksyes

Blackman 1994Foam dressing (PolyMem, Ferris)Wet-to-dry saline gauze dressing24 weeksyes

Clever 1995Foam dressing (Allevyn, Smith & Nephew)Hydrocolloid (polyurethane matrix) dressing (Cutinova Hydro, Smith & Nephew)16 weeksyes

Foster 1994Foam dressing (Allevyn, Smith & Nephew).Calcium- alginate dressing (Kaltostat, ConvaTec)8 weeksyes

Mazzone 1993Foam dressing (PolyMem, Ferris)Wet to dry saline gauze8 weeksyes

Roberts 2001Foam dressing (Allevyn, Smith & Nephew).Saline soaked low adherence dressing (Tricotex, Smith & Nephew).13 weeksyes