Intervention Review

You have free access to this content

Topical silver for preventing wound infection

  1. Marja N Storm-Versloot1,*,
  2. Cornelis G Vos1,
  3. Dirk T Ubbink2,
  4. Hester Vermeulen2

Editorial Group: Cochrane Wounds Group

Published Online: 17 MAR 2010

Assessed as up-to-date: 1 NOV 2009

DOI: 10.1002/14651858.CD006478.pub2


How to Cite

Storm-Versloot MN, Vos CG, Ubbink DT, Vermeulen H. Topical silver for preventing wound infection. Cochrane Database of Systematic Reviews 2010, Issue 3. Art. No.: CD006478. DOI: 10.1002/14651858.CD006478.pub2.

Author Information

  1. 1

    Academic Medical Centre at the University of Amsterdam, Department of Surgery, Amsterdam, Netherlands

  2. 2

    Academic Medical Centre, University of Amsterdam, Quality Assurance & Process Innovation, Amsterdam, Netherlands

*Marja N Storm-Versloot, Department of Surgery, Academic Medical Centre at the University of Amsterdam, Meibergdreef 9, Amsterdam, 1105 AZ, Netherlands. m.n.storm@amc.uva.nl.

Publication History

  1. Publication Status: New
  2. Published Online: 17 MAR 2010

SEARCH

 
Characteristics of included studies [ordered by study ID]
Afilalo 1992

MethodsComputer-generated random numbers table


Participantsn = 48
Patients in Emergency Department with partial-thickness burns, <15% TBSA
Duration of wound in both groups: < 48 hours
Unit of allocation: patient
Period of follow-up: not reported


InterventionsGroup 1: SSD cream (1%) with chlorhexidine-impregnated gauze (Bactigras®) (n = 15)
Group 2: hydrocolloid dressing (Duoderm®) (n = 15)


OutcomesInfection rate; wound healing rate; pain; patient satisfaction


Miscellaneous quality issuesMedical Ethics Committee: approved the trial
Informed consent: yes


NotesCountry: Canada
Definition of infection: not reported
Concurrent illness: none


Risk of bias

BiasAuthors' judgementSupport for judgement

Adequate sequence generation?Low riskComputer-generated random numbers table

Allocation concealment?Unclear riskNot reported

Blinding?
All outcomes
High riskParticipants, caregivers and outcome assessors not blinded

Incomplete outcome data addressed?
Drop out rate described and acceptable (> 80%)
High risk18 dropouts were described: follow-up 62%
A reason for drop out was occurrence of infection (n = 3)

Incomplete outcome data addressed?
ITT analysis
High riskITT not reported, but patients were excluded from analysis when they would not comply with the protocol

Financial support for trial or trialists?High riskSupported, in part, by a grant from Convatec Division of Bristol-Meyers

Groups similar at baseline?Low riskAuthors stated no significant differences with respect to location, size and causative agent

Co interventions avoided or similar?Low riskSame cleaning of the wound and instructions given

Carneiro 2002

MethodsRandomised


Participantsn = 64
Patients with 2nd degree burns, < 30% TBSA
Duration of wound: not reported for either group
Unit of allocation: patient
Period of follow-up: until discharge


InterventionsGroup 1: SSD/chlorhexidine (Silverex) (n = 32)
Group 2: Diphenyldantoin (phenytoin) (n = 32)


OutcomesInfection rate; wound healing rate; pain; length of hospital stay


Miscellaneous quality issuesMedical Ethics Committee: approved the trial
Informed consent: yes


NotesCountry: Tanzania
Definition of infection: cultures
Concurrent illness: none


Risk of bias

BiasAuthors' judgementSupport for judgement

Adequate sequence generation?Unclear riskReported as "Randomised, controlled, prospective study"

Allocation concealment?Unclear riskNot reported

Blinding?
All outcomes
Unclear riskNot reported

Incomplete outcome data addressed?
Drop out rate described and acceptable (> 80%)
Low riskNo dropouts were reported: follow-up 100%

Incomplete outcome data addressed?
ITT analysis
Low riskNot stated, but evident from study assessment

Financial support for trial or trialists?High riskSupport for the trial from Dreyfus Health Foundation, New York

Groups similar at baseline?Low riskAuthors stated no significant differences with respect to age, sex and extent of burn injury. Also no significant difference in positive bacterial cultures obtained on admission

Co interventions avoided or similar?Low riskSame cleaning of the wound, same wound assessment procedure.

Caruso 2006

MethodsStratified randomisation schedule


Participantsn = 84
Patients with superficial, mid-dermal or mixed partial-thickness burns of 5-40% TBSA
Duration of wound in both groups: < 36 hours
Unit of allocation: patient
Period of follow-up: 3 weeks


InterventionsGroup 1: SSD cream (n = 40)
Group 2: hydrofibre dressing containing ionic silver (Aquacel® Ag) (n = 42)


OutcomesInfection rate; wound healing rate; adverse effects; use of systemic antibiotics; pain; costs


Miscellaneous quality issuesMedical Ethics Committee: approved the trial
Informed consent: yes


NotesCountry: USA
Definition of infection: not reported
Concurrent illness: not reported


Risk of bias

BiasAuthors' judgementSupport for judgement

Adequate sequence generation?Low riskStratified randomisation schedule

Allocation concealment?Unclear riskNot reported

Blinding?
All outcomes
High riskAuthors stated that the study treatment was not blinded

Incomplete outcome data addressed?
Drop out rate described and acceptable (> 80%)
Low risk2 dropouts described: follow-up 98%

Incomplete outcome data addressed?
ITT analysis
High risk2 participants did not receive study treatment and, therefore, were not analysed

Financial support for trial or trialists?High riskThe trial was supported by a grant from Convatec, a Bristol-Myers Squibb company. Convatec supervised the design, data-analysis and the development of the manuscript of this study

Groups similar at baseline?Low riskAuthors stated baseline characteristics were comparable

Co interventions avoided or similar?Unclear riskNot reported

De Gracia 2001

MethodsPatients assigned consecutively according to a pre-established randomised sequence


Participantsn = 60
Patients with moderate and severe burns, >15% TBSA
Duration of wound in both groups: < 24 hours
Unit of allocation: patient
Period of follow-up: not reported


InterventionsGroup 1: SSD cream (Flamazine®) (n = 30)
Group 2 SSD/cerium-nitrate (Flammacerium®) (n = 30)


OutcomesInfection rate; wound healing rate; adverse effects; use of systemic antibiotics; length of hospital stay


Miscellaneous quality issuesMedical Ethics Committee: not reported
Informed consent: yes


NotesCountry: Philippines
Definition of infection: cultures and clinical criteria
Concurrent illness: majority none


Risk of bias

BiasAuthors' judgementSupport for judgement

Adequate sequence generation?Unclear riskPatients were assigned consecutively according to a pre-established randomised sequence

Allocation concealment?Unclear riskNot reported

Blinding?
All outcomes
Unclear riskStated that double-blinding was not possible, but not explicitly reported for the outcome assessor

Incomplete outcome data addressed?
Drop out rate described and acceptable (> 80%)
Low risk11 dropouts were described, but only 1 patient was not included in the analysis

Incomplete outcome data addressed?
ITT analysis
High riskOne participant was not analysed due to poor compliance

Financial support for trial or trialists?Unclear riskNot reported

Groups similar at baseline?Low riskNo significant difference with respect to age. The extent of burn injury differed, but was compensated with multiple linear regression analysis

Co interventions avoided or similar?Low riskCleansing of wounds and other treatments stated

Dire 1995

MethodsRandomised, opaque numbered vials


Participantsn = 465
Patients with minor, uncomplicated soft-tissue wounds necessitating suturing
Duration of wound in both groups: < 12 hours
Unit of allocation: patient
Period of follow-up: not reported


InterventionsGroup 1: SSD cream (n = 99)
Group 2: BAC (n = 109)
Group 3: neomycin sulfate (n = 110)
Group 4: petrolatum (n = 108)


OutcomesInfection rate


Miscellaneous quality issuesMedical Ethics Committee: approved the trial
Informed consent: yes


NotesCountry: Texas
Definition of infection: clinical criteria
Concurrent illness: none


Risk of bias

BiasAuthors' judgementSupport for judgement

Adequate sequence generation?Unclear riskRandomised, opaque numbered vials

Allocation concealment?Unclear riskStudy agent in randomised opaque vials labelled with identification numbers, but not clear whether the person responsible for determining eligibility of participants had influence on the assignment sequence

Blinding?
All outcomes
Low riskReported as double-blinded: likely that the outcome assessor was blinded

Incomplete outcome data addressed?
Drop out rate described and acceptable (> 80%)
Low risk39 dropouts were described: follow-up 92%

Incomplete outcome data addressed?
ITT analysis
High risk1 participant received the wrong study treatment and was excluded from analysis

Financial support for trial or trialists?Unclear riskNot reported

Groups similar at baseline?Low riskNo significant differences with respect to age, or wounds' depth and location

Co interventions avoided or similar?Low riskSame wound treatment except for study agent

Fang 1987

MethodsSelected at random


Participantsn = 27 (54 wound sites)
Patients with similar size and injury-matched areas of 2nd degree burns
Duration of wound: not reported in either group
Unit of allocation: wounds
Period of follow-up: not reported


InterventionsGroup 1: SSD cream (1%) (n = 27)
Group 2: synthetic dressing containing silver (Hydron AgSD (1-3%)) (n = 27)


OutcomesInfection rate; wound healing rate


Miscellaneous quality issuesMedical Ethics Committee: not reported
Informed consent: not reported


NotesCountry: China
Definition of infection: cultures
Concurrent illness: not reported


Risk of bias

BiasAuthors' judgementSupport for judgement

Adequate sequence generation?Unclear riskSelected at random

Allocation concealment?Unclear riskNot reported

Blinding?
All outcomes
Unclear riskNot reported for participants and caregivers
Outcome assessors not blinded

Incomplete outcome data addressed?
Drop out rate described and acceptable (> 80%)
Unclear riskNot reported

Incomplete outcome data addressed?
ITT analysis
Unclear riskNot reported

Financial support for trial or trialists?Unclear riskNot reported

Groups similar at baseline?Low riskAuthor reported the wounds as injury and size matched.

Co interventions avoided or similar?Low riskParticipants acted as their own control

Gerding 1988

MethodsComputer-generated codes within sealed, numbered envelopes that were opened sequentially


Participantsn = 47 (50 wounds)
Inpatients with acute partial-thickness burns
Duration of wound in both groups: < 6 hours
Unit of allocation: wounds
Period of follow-up: not reported


InterventionsGroup 1: SSD cream (1%) (n = 23)
Group 2: biosynthetic dressing (Biobrane®) (n = 27)


OutcomesInfection rate; wound healing rate; pain; costs


Miscellaneous quality issuesMedical Ethics Committee: not reported
Informed consent: not reported


NotesCountry: USA
Definition of infection: cultures and clinical criteria
Concurrent illness: not reported


Risk of bias

BiasAuthors' judgementSupport for judgement

Adequate sequence generation?Low riskComputer-generated codes within sealed, numbered envelopes that were opened sequentially

Allocation concealment?Low riskSequentially-opened sealed, numbered envelopes

Blinding?
All outcomes
Unclear riskNot reported

Incomplete outcome data addressed?
Drop out rate described and acceptable (> 80%)
Low risk4 dropouts described: follow-up 91%

Incomplete outcome data addressed?
ITT analysis
High risk1 participant excluded from analysis due to protocol violation

Financial support for trial or trialists?Unclear riskNot reported

Groups similar at baseline?Low riskAuthor stated no differences for sex, race and burn agent

Co interventions avoided or similar?Low riskSame wound treatment except for study agent. 7 participants acted as their own control

Gerding 1990

MethodsComputer-generated codes within sealed, numbered envelopes that were opened sequentially


Participantsn = 64 (analysed 56 wounds)
Outpatients with acute partial-thickness thermal burns, < 10% TBSA
Duration of wound in both groups: < 24 hours
Unit of allocation: wounds
Period of follow-up: not reported


InterventionsGroup 1: SSD cream (1%) (n = 26)
Group 2: biosynthetic dressing (Biobrane®) (n = 30)


OutcomesInfection rate; wound healing rate; pain; costs


Miscellaneous quality issuesMedical Ethics Committee: approved the trial
Informed consent: not reported


NotesCountry: USA
Definition of infection: clinical criteria, but not described in detail
Concurrent illness: not reported


Risk of bias

BiasAuthors' judgementSupport for judgement

Adequate sequence generation?Low riskComputer-generated codes within sealed, numbered envelopes that were opened sequentially

Allocation concealment?Low riskSequentially-opened sealed, numbered envelopes

Blinding?
All outcomes
High riskParticipants, caregivers and outcome assessors not blinded

Incomplete outcome data addressed?
Drop out rate described and acceptable (> 80%)
Low risk12 dropouts described: follow-up 81%

Incomplete outcome data addressed?
ITT analysis
High risk5 participants excluded from analysis due to protocol violations

Financial support for trial or trialists?Unclear riskNot reported

Groups similar at baseline?Low riskAuthor stated no differences for sex, race and TBSA burned

Co interventions avoided or similar?Low riskSame cleaning of the wound and follow-up procedure

Hansbrough 1995

MethodsRandomly assigned


Participantsn = 79 (158 wounds)
Patients with 2 similar partial-thickness burns of 1-25% TBSA
Duration of wound in both groups: < 4 days
Unit of allocation: wounds
Period of follow-up: not reported


InterventionsGroup 1: SSD cream (1%) (Silvadene) (n = 79)
Group 2: collagenase ointment applied with polymyxin B sulfate/bacitracin powder (Santyl®) (n = 79)


OutcomesInfection rate; wound healing rate; pain


Miscellaneous quality issuesMedical Ethics Committee: not reported
Informed consent: not reported


NotesCountry: USA
Definition of infection: not reported
Concurrent illness: not reported


Risk of bias

BiasAuthors' judgementSupport for judgement

Adequate sequence generation?Unclear riskRandomly assigned

Allocation concealment?Unclear riskNot reported

Blinding?
All outcomes
Unclear riskAlthough a designated observer evaluated time to clean wound bed, not reported whether observer was blinded regarding treatment

Incomplete outcome data addressed?
Drop out rate described and acceptable (> 80%)
High risk34 participants discontinued early and described. Follow-up 56%

Incomplete outcome data addressed?
ITT analysis
Low riskParticipants with incomplete data assigned censored values for missing data points

Financial support for trial or trialists?High riskStudy sponsored by Knoll Pharmaceutical Company, Whyppany, NJ

Groups similar at baseline?Low riskNo significant differences for wound size and location

Co interventions avoided or similar?Low riskParticipants acted as their own control

Homann 2007

MethodsComputer-generated randomisation list


Participantsn = 47
Patients with 2 comparable partial-thickness burns (degree IIa) without wound infection, < 50% TBSA
Duration of wound in both groups: < 72 hours
Unit of allocation: wounds
Period of follow-up: until healing


InterventionsGroup 1: SSD cream (n = 43)
Group 2: liposome hydrogel with PVP-I (n = 43)


OutcomesInfection rate; wound healing rate; adverse effects; pain


Miscellaneous quality issuesMedical Ethics Committee: approved the trial
Informed consent: yes


NotesCountry: Germany
Definition of infection: clinical criteria
Concurrent illness: not sufficiently reported


Risk of bias

BiasAuthors' judgementSupport for judgement

Adequate sequence generation?Low riskComputer-generated randomisation list using the program Rancode

Allocation concealment?Unclear riskNot sufficiently reported

Blinding?
All outcomes
Low riskParticipants and caregivers not blinded; outcome assessors blinded

Incomplete outcome data addressed?
Drop out rate described and acceptable (> 80%)
Low risk4 participants excluded due to protocol violations for inclusion criteria, and 4 participants did not complete the study: follow-up 91%

Incomplete outcome data addressed?
ITT analysis
High risk4 participants excluded due to protocol violations

Financial support for trial or trialists?High riskThe study was sponsored by Mundipharma GmbH

Groups similar at baseline?Low riskNo significant difference for wound size

Co interventions avoided or similar?Low riskParticipants acted as their own control

Hutchinson 1993

MethodsProspective, controlled, randomised investigation


Participantsn = 292
Patients with venous leg ulcers, partial-thickness burns or partial-thickness donor sites.
Duration of wound in both groups: not reported for burns or donor sites.
Duration of wound for leg ulcers: Group 1: 318 weeks; Group 2: 102 weeks; Group 3: 162 weeks.
Unit of allocation: patient.
Period of follow-up: for burns and donor sites 3 weeks; for leg ulcers 10 weeks


InterventionsGroup 1: SSD cream/hydrocolloid (n = 58)
Group 2: hydrocolloid (n = 108)
Group 3: non-occlusive paraffin impregnated gauze (n = 126)


OutcomesInfection rate


Miscellaneous quality issuesMedical Ethics Committee: not reported
Informed consent: not reported


NotesCountry: USA, United Kingdom, Netherlands
Definition of infection: clinical criteria
Concurrent illness: not reported


Risk of bias

BiasAuthors' judgementSupport for judgement

Adequate sequence generation?Unclear riskProspective, controlled, randomised investigation

Allocation concealment?Unclear riskNot reported

Blinding?
All outcomes
Unclear riskNot reported

Incomplete outcome data addressed?
Drop out rate described and acceptable (> 80%)
High risk70 dropouts, reasons not described: follow-up 76%

Incomplete outcome data addressed?
ITT analysis
Unclear riskNot reported

Financial support for trial or trialists?High riskAuthor works for ConvaTec Ltd

Groups similar at baseline?Low riskGroups statistically homogeneous

Co interventions avoided or similar?Unclear riskNot reported

Inman 1984

MethodsRandomly assigned


Participantsn = 121
Patients with fresh full-thickness burns
Duration of wound in both groups: < 24 hours
Unit of allocation: patient
Period of follow-up: not reported


InterventionsGroup 1: SSD cream (Flamazine®) (n = 54)
Group 2: SSD/chlorhexidine digluconate cream (0.2%) (Silvazine®) (n = 67)


OutcomesInfection rate; use of systemic antibiotics; pain


Miscellaneous quality issuesMedical Ethics Committee: approved the trial
Informed consent: yes


NotesCountry: Canada
Definition of infection: clinical criteria accompanied with cultures with > 105 organisms per gram of tissue
Concurrent illness: not sufficiently reported


Risk of bias

BiasAuthors' judgementSupport for judgement

Adequate sequence generation?Unclear riskRandomly assigned

Allocation concealment?Unclear riskNot reported

Blinding?
All outcomes
Unclear riskNot reported

Incomplete outcome data addressed?
Drop out rate described and acceptable (> 80%)
Unclear riskReasons for drop out described, but number of excluded participants not reported. Unclear if follow-up was > 80%

Incomplete outcome data addressed?
ITT analysis
High riskNot all participants included in the analysis according to randomisation group to which allocated. Patients excluded from analysis if they did not survive for 7 days, if all eschar excised before day 7, and if they were discharged before day 7

Financial support for trial or trialists?High riskFunding support for microbiological studies and statistical analysis by British Columbia Professional Firefighters Association, and Smith and Nephew

Groups similar at baseline?High riskGroups comparable, except that scald burns more frequent in SSD group. No adjustment made in the analysis

Co interventions avoided or similar?Low riskSame wound cleaning and same procedure when cultures were taken

Innes 2001

MethodsRandomisation table assigned dressings to site A or site B


Participantsn = 17 (32 wound sites)
Patients with burns who required split-thickness skin graft
Duration of wound in both groups: not reported
Unit of allocation: wounds
Period of follow-up: > 3 months


InterventionsGroup 1: nanocrystalline silver-coated dressing (Acticoat®) (n = 16)
Group 2: hydrophilic polyurethane dressing (Allevyn®) (n = 16)


OutcomesInfection rate; wound healing rate; costs


Miscellaneous quality issuesMedical Ethics Committee: approved the trial
Informed consent: yes


NotesCountry: Canada
Definition of infection: cultures and clinical criteria
Concurrent illness: not sufficiently reported


Risk of bias

BiasAuthors' judgementSupport for judgement

Adequate sequence generation?Unclear riskRandomisation table assigned dressings to either site A or B

Allocation concealment?Unclear riskNot reported

Blinding?
All outcomes
High riskNot reported for participants; caregivers not blinded. Although authors stated that 4 independent observers viewed standard images of wounds for re-epitheliailisation and that scar was rated by a blinded observer, they stated that the daily wound observer was an experienced burn surgeon. It is likely that he was not blinded to the treatment

Incomplete outcome data addressed?
Drop out rate described and acceptable (> 80%)
Low risk2 dropouts described: follow-up 88%

Incomplete outcome data addressed?
ITT analysis
Low riskNot stated, but evident from study assessment

Financial support for trial or trialists?Unclear riskNot reported

Groups similar at baseline?Low riskNo differences in wound size

Co interventions avoided or similar?Low riskParticipants acted as their own controls

Jacobs 2008

MethodsRandomly assigned


Participantsn = 40
Diabetic patients with Wagner grade 1 or 2 ulcers
Duration of wound in both groups: not reported
Unit of allocation: patients
Period of follow-up: 6 weeks


InterventionsGroup 1: SSD cream (n = 20)
Group 2: benzoic acid-6%, salicylic acid-3% and Quercus rubra extract-3% (Bensal HP) (n = 20)


OutcomesInfection rate; wound healing rate


Miscellaneous quality issuesMedical Ethics Committee: not reported
Informed consent: not reported


NotesCountry: Canada
Definition of infection: cultures and clinical criteria
Concurrent illness: all patients had diabetes, patients with peripheral vascular disease were excluded. Additional co-morbid conditions were not evaluated


Risk of bias

BiasAuthors' judgementSupport for judgement

Adequate sequence generation?Low riskDrawing marbles of different colours out of a sock; marbles were not replaced (information retrieved from author)

Allocation concealment?High riskParticipants blindly drew a marble out of a sock. This technique has a high risk of subversion since there is no audit trail

Blinding?
All outcomes
Unclear riskNot sufficiently reported. The authors stated that the study was blinded, but did not report who was blinded

Incomplete outcome data addressed?
Drop out rate described and acceptable (> 80%)
Low riskStated, that there were no dropouts; follow-up 100%

Incomplete outcome data addressed?
ITT analysis
Low riskNot stated, but evident from study assessment

Financial support for trial or trialists?Unclear riskNot reported

Groups similar at baseline?Low riskGroups comparable for wound size, and authors stated that there were no differences between the two groups

Co interventions avoided or similar?Low riskAll patients treated with off-loading, debridement, instructions for application, wound coverage

Jude 2007

MethodsRandomly assigned to instructions in a sealed envelope


Participantsn = 134
Patients with diabetic foot ulcers of Wagner grade 1 or 2 of non-ischaemic aetiology with area ≥1 cm2
Duration of wound in both groups: not reported
Unit of allocation: patient
Period of follow-up: 8 weeks


InterventionsGroup 1: hydrofibre dressing containing ionic silver (Aquacel® Ag) (n = 67)
Group 2: calcium alginate dressing (Algosteril®) (n = 67)


OutcomesInfection rate; wound healing rate; adverse effects


Miscellaneous quality issuesMedical Ethics Committee: approved the trial
Informed consent: yes


NotesCountry: United Kingdom, Germany, Sweden, France
Definition of infection: clinical criteria
Concurrent illness: DM, types 1 and 2


Risk of bias

BiasAuthors' judgementSupport for judgement

Adequate sequence generation?Low riskRandomly assigned to instructions in a sealed envelope

Allocation concealment?Unclear riskInstructions in a sealed envelope, but not clear if the envelopes were sequentially numbered and opaque

Blinding?
All outcomes
High riskParticipants, caregivers and outcome assessors not blinded: open label

Incomplete outcome data addressed?
Drop out rate described and acceptable (> 80%)
Low risk21 participants discontinued the study, but were included in the ITT analysis

Incomplete outcome data addressed?
ITT analysis
Low riskStated as undertaken

Financial support for trial or trialists?High riskStudy supported by clinical grant from ConvaTec, a Bristol-Myers Squibb Company, Princeton, NJ, USA

Groups similar at baseline?Unclear riskIt seemed that at baseline there were larger ulcers in the control group and more frequent use of antibiotics in the hydrofibre-silver group

Co interventions avoided or similar?Low riskSame wound cleansing and treatment, except for study agent

Jurczak 2007

MethodsComputer-generated randomisation scheme with sealed envelopes opened sequentially


Participantsn = 67
Patients with open surgical wounds or open traumatic wounds left to heal by secondary intent and requiring an antimicrobial dressing
Duration of wound in both groups: < 12 hours
Unit of allocation: patient
Period of follow-up: 2 weeks


InterventionsGroup 1: Hydrofibre dressing containing ionic silver (Aquacel® Ag) (n = 35)
Group 2: povidone iodine gauze (n = 32)


OutcomesInfection rate; wound healing rate; adverse effects; pain


Miscellaneous quality issuesMedical Ethics Committee: approved the trial
Informed consent: yes


NotesCountry: Great Britain, Germany, France
Definition of infection: clinical criteria
Concurrent illness: not sufficiently reported


Risk of bias

BiasAuthors' judgementSupport for judgement

Adequate sequence generation?Low riskComputer-generated randomisation scheme with sealed envelopes opened sequentially

Allocation concealment?Low riskSequentially-opened, sealed, numbered envelopes

Blinding?
All outcomes
High riskParticipants, caregivers and outcome assessors not blinded: open label study

Incomplete outcome data addressed?
Drop out rate described and acceptable (> 80%)
Low risk5 dropouts described

Incomplete outcome data addressed?
ITT analysis
Low riskDropouts included in the ITT analysis for primary endpoints

Financial support for trial or trialists?High riskStudy supported by grant from Convatec, a Bristol-Myers Squibb Company. Convatec monitored study design, study conduct, and data collection, and supervised data analysis and preparation of the manuscript

Groups similar at baseline?Low riskOnly mean ulcer area was larger in the povidone iodine gauze group due to an outlier (976 mm2 vs 1463 mm2), but median ulcer area was comparable (both 600 mm2)

Co interventions avoided or similar?Low risk

Livingston 1990

MethodsRandomisation by labelling cards, shuffling and drawing in blinded fashion; because of resulting imbalance in group size, last 7 consecutive patients all placed in silver nitrate group


Participantsn = 52
Patients with thermal injury who required skin grafting
Duration of wound in both groups:
Group 1: mean 3-4 days to first graft;
Group 2: mean 4 days;
Group 3: 4-9 days
Unit of allocation: patient
Period of follow-up: not reported


InterventionsGroup 1: silver nitrate 0.5% (n = 19)
Group 2: Ringer's lactate (n = 15)
Group 3: neomycin (1 g/L) + bacitracin (50,000 Units/L) (n = 18)


OutcomesInfection rate; length of hospital stay


Miscellaneous quality issuesMedical Ethics Committee: approved the trial
Informed consent: yes


NotesCountry: USA
Definition of infection: > 105 organisms per gram of tissue
Concurrent illness: not reported


Risk of bias

BiasAuthors' judgementSupport for judgement

Adequate sequence generation?Low riskRandomisation by labelling cards, shuffling and drawing in blinded fashion

Allocation concealment?Unclear riskNot reported

Blinding?
All outcomes
Unclear riskNot reported

Incomplete outcome data addressed?
Drop out rate described and acceptable (> 80%)
Low riskNo dropouts reported: follow-up 100%

Incomplete outcome data addressed?
ITT analysis
Low riskEvident from study assessment

Financial support for trial or trialists?Unclear riskNot reported

Groups similar at baseline?Unclear riskAverage age and percentage TBSA similar, however, operative procedures and hospital stay not similar

Co interventions avoided or similar?Low riskSame nutritional support and antibiotic prophylaxis

Mashhood 2006

MethodsRandomly divided into two groups


Participantsn = 50
Patients with superficial and partial-thickness burns, < 15% TBSA
Duration of wound in both groups: not reported
Unit of allocation: patient
Period of follow-up: 6 months


InterventionsGroup 1: SSD cream (n = 25)
Group 2: honey (pure, unprocessed, and undiluted) (n = 25)


OutcomesWound healing rate; pain; costs


Miscellaneous quality issuesMedical Ethics Committee: not reported
Informed consent: not reported


NotesCountry: Pakistan
Definition of infection: cultures and clinical criteria
Concurrent illness: none


Risk of bias

BiasAuthors' judgementSupport for judgement

Adequate sequence generation?Unclear riskRandomly divided into two groups

Allocation concealment?Unclear riskNot reported

Blinding?
All outcomes
Unclear riskNot reported

Incomplete outcome data addressed?
Drop out rate described and acceptable (> 80%)
Low riskNo dropouts reported: follow-up 100%

Incomplete outcome data addressed?
ITT analysis
Low riskEvident from study assessment

Financial support for trial or trialists?Unclear riskNot reported

Groups similar at baseline?Unclear riskNot reported

Co interventions avoided or similar?Low riskGeneral management in wound care was the same

Miller 1990

MethodsComputer-generated randomisation schedule


Participantsn = 59
Patients with full-thickness burns, < 40% TBSA
Duration of wound in both groups: < 48 hours
Unit of allocation: wounds
Period of follow-up: 2 weeks


InterventionsGroup 1: SSD cream (Silvadene) (n = 51)
Group 2: Polyethylene glycol 400, poly-2-hydroxyethyl methacrylate, and dimethyl sulfoxide: Dimac-containing SSD (Sildimac®) (n = 51)


OutcomesInfection rate; adverse effects


Miscellaneous quality issuesMedical Ethics Committee: not reported
Informed consent: yes


NotesCountry: USA
Definition of infection: > 105 organisms per gram of tissue
Concurrent illness: none


Risk of bias

BiasAuthors' judgementSupport for judgement

Adequate sequence generation?Low riskComputer-generated randomisation schedule

Allocation concealment?Unclear riskNot reported

Blinding?
All outcomes
Unclear riskNot reported for outcome assessors. Blinding was not possible for participants

Incomplete outcome data addressed?
Drop out rate described and acceptable (> 80%)
Low risk8 dropouts described; follow-up 86%

Incomplete outcome data addressed?
ITT analysis
Unclear riskITT used for adverse effects, but 8 participants not analysed for infection rate

Financial support for trial or trialists?High riskPartly supported by grant from Marion Laboratories, Inc

Groups similar at baseline?Low riskParticipants with two comparable wound sites. No significant difference before treatment in number of positive biopsies

Co interventions avoided or similar?Low riskParticipants acted as their own controls

Muangman 2006

MethodsRandomised into two groups


Participantsn = 50
Patients with partial-thickness burns, < 25% TBSA
Duration of wound in both groups: not reported
Unit of allocation: patient
Period of follow-up: not reported


InterventionsGroup 1: SSD cream (n = 25)
Group 2: nanocrystalline silver-coated dressing (Acticoat®) (n = 25)


OutcomesInfection rate; pain; length of hospital stay


Miscellaneous quality issuesMedical Ethics Committee: not reported
Informed consent: not reported


NotesCountry: Thailand
Definition of infection: swabs and clinical criteria
Concurrent illness: not reported


Risk of bias

BiasAuthors' judgementSupport for judgement

Adequate sequence generation?Unclear riskRandomised into two groups

Allocation concealment?Unclear riskNot reported

Blinding?
All outcomes
High riskNot explicitly stated, but materials different, so participants and caregivers not blinded. Not reported for outcome assessor

Incomplete outcome data addressed?
Drop out rate described and acceptable (> 80%)
Low riskNo dropouts reported: follow-up 100%

Incomplete outcome data addressed?
ITT analysis
Low riskAll participants included in analysis, not likely that ITT was violated

Financial support for trial or trialists?Unclear riskNot reported

Groups similar at baseline?Low riskNo differences in baseline characteristics

Co interventions avoided or similar?Unclear riskNot reported

Noordenbos 1999

MethodsRandomised, chosen wound sites


Participantsn = 14
Patients with 2 comparable-sized, moderate to deep, partial-thickness burns of 2-30% TBSA
Duration of wound in both groups: < 24 hours
Unit of allocation: wounds
Period of follow-up: 12 months


InterventionsGroup 1: SSD cream (n = 14)
Group 2: biosynthetic dressing with skin substitute (Transcyte on Biobrane mesh) (n = 14)


OutcomesInfection rate; wound healing rate


Miscellaneous quality issuesMedical Ethics Committee: approved the trial
Informed consent: not reported


NotesCountry: USA
Definition of infection: not reported
Concurrent illness: not reported


Risk of bias

BiasAuthors' judgementSupport for judgement

Adequate sequence generation?Unclear riskRandomised, chosen wound sites

Allocation concealment?Unclear riskNot reported

Blinding?
All outcomes
High riskNon-blinded study: participants, caregivers and outcome assessors not blinded

Incomplete outcome data addressed?
Drop out rate described and acceptable (> 80%)
Low riskNo dropouts reported: follow-up 100%

Incomplete outcome data addressed?
ITT analysis
Low riskEvident from study assessment

Financial support for trial or trialists?Low riskNon-sponsored, investigator-initiated investigational device exemption

Groups similar at baseline?Low riskParticipants with two comparable-sized wound sites

Co interventions avoided or similar?Low riskParticipants acted as their own controls

Soroff 1994

MethodsRandomisation schedule


Participantsn = 15
Patients with partial-thickness burns, < 25% TBSA
Duration of wound in both groups: 1-10 days
Unit of allocation: wounds
Period of follow-up: till healing


InterventionsGroup 1: SSD cream (n = 15)
Group 2: collagenase ointment applied with polymyxin B sulfate/Bacitracin spray (n = 15)


OutcomesWound healing rate


Miscellaneous quality issuesMedical Ethics Committee: approved the trial
Informed consent: yes


NotesCountry: USA
Definition of infection: not reported
Concurrent illness: not reported


Risk of bias

BiasAuthors' judgementSupport for judgement

Adequate sequence generation?Unclear riskRandomisation schedule

Allocation concealment?Unclear riskNot reported

Blinding?
All outcomes
High riskNot explicitly stated, but materials were different, so participants and caregivers not blinded. Not reported for outcome assessor

Incomplete outcome data addressed?
Drop out rate described and acceptable (> 80%)
Low risk2 dropouts described: follow-up 87%

Incomplete outcome data addressed?
ITT analysis
High risk1 cross-over not treated as randomised and, therefore, there was no ITT analysis

Financial support for trial or trialists?High riskSupported by Knoll Pharmaceutical Company

Groups similar at baseline?Unclear riskAlthough the author stated that groups were not statistically different, the mean burn size was larger in the collagenase group (182.7 cm2 vs 163.7 cm2)

Co interventions avoided or similar?Low riskParticipants acted as their own controls

Subrahmanyam 1998

MethodsRandomly allocated to two groups


Participantsn = 50
Patients with superficial thermal burns, < 40% TBSA
Duration of wound in both groups: < 6 hours
Unit of allocation: patient
Period of follow-up: > 3 weeks


InterventionsGroup 1: SSD cream (n = 25)
Group 2: honey (pure, unprocessed, undiluted from the hive) (n = 25)


OutcomesInfection rate; wound healing rate


Miscellaneous quality issuesMedical Ethics Committee: not reported
Informed consent: not reported


NotesCountry: India
Definition of infection: clinical criteria
Concurrent illness: not reported


Risk of bias

BiasAuthors' judgementSupport for judgement

Adequate sequence generation?Unclear riskRandomly allocated to two groups

Allocation concealment?Unclear riskNot reported

Blinding?
All outcomes
High riskNot explicitly stated, but materials were different, so participants and caregivers not blinded. Not reported for outcome assessor

Incomplete outcome data addressed?
Drop out rate described and acceptable (> 80%)
Low riskNo dropouts reported: follow-up 100%

Incomplete outcome data addressed?
ITT analysis
Low riskEvident from study assessment

Financial support for trial or trialists?Unclear riskNot reported

Groups similar at baseline?Low riskNo differences in baseline characteristics

Co interventions avoided or similar?Low riskSame wound assessment and same procedure when biopsies were taken

Tredget 1998

MethodsComputer-generated randomisation schedule


Participantsn = 30 (60 wounds)
Patients with deep partial- and full-thickness burns
Duration of wound in both groups: < 72 hours
Unit of allocation: wounds
Period of follow-up: mean 4 days (until first operative procedure)


InterventionsGroup 1: nanocrystalline silver-coated dressing (Acticoat®) (n = 30)
Group 2: fine-mesh gauze moistened with a 0.5% solution of silver nitrate (n = 30)


OutcomesInfection rate; wound healing rate; pain


Miscellaneous quality issuesMedical Ethics Committee: not reported
Informed consent: yes


NotesCountry: Canada
Definition of infection: > 105 organisms per gram of tissue
Concurrent illness: none


Risk of bias

BiasAuthors' judgementSupport for judgement

Adequate sequence generation?Low riskComputer-generated randomisation schedule

Allocation concealment?Unclear riskNot reported

Blinding?
All outcomes
Unclear riskNot reported

Incomplete outcome data addressed?
Drop out rate described and acceptable (> 80%)
Low risk1 dropout described: follow-up 97%

Incomplete outcome data addressed?
ITT analysis
Low riskEvident from study assessment

Financial support for trial or trialists?High riskSupported by funding from the Westaim Corporation, Fort Saskatchewan, Alberta, Canada

Groups similar at baseline?Low riskTwo comparable wounds

Co interventions avoided or similar?Low riskParticipants acted as their own controls

Wunderlich 1991

MethodsRandomised


Participantsn = 40
Patients with venous leg ulcers
Duration of wound in both groups: Group 1: 7.6 years; Group 2: 7.9 years
Unit of allocation: patient
Period of follow-up: 6 weeks


InterventionsGroup 1: activated charcoal xerodressing silver-impregnated (SIAX) (Actisorb plus) (n = 19)
Group 2: conventional phase-adapted therapy using diverse topical modalities (n = 19)


OutcomesInfection rate; wound healing rate


Miscellaneous quality issuesMedical Ethics Committee: not reported
Informed consent: not reported


NotesCountry: Germany
Definition of infection: swabs
Concurrent illness: DM excluded, other ulcers excluded


Risk of bias

BiasAuthors' judgementSupport for judgement

Adequate sequence generation?Unclear riskRandomised

Allocation concealment?Unclear riskNot reported

Blinding?
All outcomes
High riskOpen RCT: participants, caregivers and outcome assessors not blinded

Incomplete outcome data addressed?
Drop out rate described and acceptable (> 80%)
Low risk2 dropouts described: follow-up 95%

Incomplete outcome data addressed?
ITT analysis
Low riskEvident from study assessment

Financial support for trial or trialists?Unclear riskNot reported

Groups similar at baseline?Low riskBaseline characteristics similar

Co interventions avoided or similar?Low riskSame wound care, except for study agent

Wyatt 1990

MethodsRandomly assigned


Participantsn = 50
Patients with minor second degree burns
Duration of wound in both groups: < 48 hours
Unit of allocation: patient
Period of follow-up: until healing


InterventionsGroup 1: SSD cream (n = 20)
Group 2: hydrocolloid (Duoderm® Hydroactive) (n = 22)


OutcomesInfection rate; wound healing rate; pain


Miscellaneous quality issuesMedical Ethics Committee: not reported
Informed consent: yes


NotesCountry: USA
Definition of infection: clinical criteria, but not described in detail
Concurrent illness: none


Risk of bias

BiasAuthors' judgementSupport for judgement

Adequate sequence generation?Unclear riskRandomly assigned

Allocation concealment?Unclear riskNot reported

Blinding?
All outcomes
Low riskParticipants, caregivers not blinded
Outcome assessor blinded

Incomplete outcome data addressed?
Drop out rate described and acceptable (> 80%)
Low risk8 dropouts described: follow-up 84%

Incomplete outcome data addressed?
ITT analysis
High riskFour participants were wrongfully randomised and therefore excluded from analysis. Four other participants were lost to follow-up and not analysed

Financial support for trial or trialists?Unclear riskNot reported

Groups similar at baseline?Low riskNo significant differences for baseline characteristics

Co interventions avoided or similar?Low riskSame wound assessment and follow-up

 
Characteristics of excluded studies [ordered by study ID]

StudyReason for exclusion

De Boer 1981Not an RCT

Ganai 2002No comparison of dressings

Guilbaud 1993Almost no silver used, and no separate figures reported on the effect of silver

Hadjiiski 1999Not an RCT

Huang 2007Wounds already infected at inclusion

Jorgensen 2006Wounds already infected at inclusion

Lanzara 2008Only abstract available; no response to attempts to contact investigator

Molnar 2004Only abstract available; no response to attempts to contact investigator

Munster 1980Not an RCT

Münter 2006Wounds already infected at inclusion

Planinsek 2007Only abstract available; no response to attempts to contact investigator

Riesinger 2006Only abstract available; not able to retrieve contact information

Silver 2007Not an RCT

Stair 1986Cross-over study

Subrahmanyam 1991The aim of the study was not prevention of infection

Terrill 1991Compared the bacteriological properties and clinical performance of polythene and polytetrafluoroethylene fabric bags, both containing SSD. The

silver-containing dressings were not compared.

Verdú 2004Not an RCT

Yue Seng 2005Only abstract available; no response to attempts to contact investigator

 
Characteristics of studies awaiting assessment [ordered by study ID]
Chen 2006

MethodsRandomly divided

ParticipantsPatients with superficial and deep burns (n = 191)

InterventionsGroup 1: SSD cream
Group 2: silver nanoparticle dressing
Group 3: Vaseline gauze

OutcomesInfection rate; wound healing rate

NotesArticle in Chinese

Hirsch 2008

MethodsSingle-centred randomised clinical trial

ParticipantsPatients with second degree burns (n = 40)

InterventionsGroup 1: SSD cream (Flamazine)
Group 2: moist exposed burn ointment (MEBO)

OutcomesInfection rate; wound healing rate; pain

Notes

Li 2006

MethodsMulti-centred randomised clinical trial

ParticipantsPatients with residual burns (n = 98; 166 wounds)

InterventionsGroup 1: SSD cream
Group 2: nanocrystalline silver dressing (Acticoat)

OutcomesWound healing time

NotesArticle in Chinese

Wang 2008

MethodsRandomised control study

ParticipantsPatients with wounds from dog bites (n = 40)

InterventionsGroup 1: ionic silver dressing (Aquacel)
Group 2: Duoderm Hydroactive gel

OutcomesWound healing

NotesArticle in Chinese

 
Characteristics of ongoing studies [ordered by study ID]
Serena 2008

Trial name or titleThe lack of reliability of clinical examination in the diagnosis of wound infection: preliminary communication

MethodsMulticentred randomised clinical trial

ParticipantsPatients with chronic venous leg ulcers (n = 49)

InterventionsCollagen-ORC antimicrobial matrix compared with moist wound dressings

OutcomesReduction in wound area, number of wounds healed in 12 weeks, infection, healing rate, pain, quality of life, ease of use

Starting dateAugust 2004 to October 2005

Contact information

NotesConference proceedings and preliminary results

 
Comparison 1. Silver sulfadiazine (SSD) cream (1%) vs biosynthetic dressing (Biobrane®)

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

 1 Number of patients that developed wound infection2106Risk Difference (M-H, Fixed, 95% CI)-.00 [-0.12, 0.12]

 2 Mean pain scores2106Mean Difference (IV, Fixed, 95% CI)1.41 [0.99, 1.83]

 3 Costs based on hospital charges (US dollars)156Mean Difference (IV, Fixed, 95% CI)70.0 [15.54, 124.46]

 
Comparison 2. SSD cream (1%) vs biosynthetic dressing with skin substitute (Transcyte® on Biobrane® mesh)

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

 1 Number of patients that developed wound infection128Risk Difference (M-H, Fixed, 95% CI)0.43 [0.16, 0.70]

 
Comparison 3. SSD cream (1%) with chlorhexidine-impregnated gauze (Bactigras®) vs hydrocolloid (Duoderm® Hydroactive)

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

 1 Number of patients that developed wound infection148Risk Difference (M-H, Fixed, 95% CI)-0.04 [-0.18, 0.09]

 
Comparison 4. SSD cream (1%) vs hydrocolloid (Duoderm® Hydroactive)

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

 1 Number of patients that developed wound infection142Risk Difference (M-H, Fixed, 95% CI)0.0 [-0.09, 0.09]

 2 Mean pain scores142Mean Difference (IV, Fixed, 95% CI)1.19 [0.56, 1.82]

 
Comparison 5. SSD cream (1%) vs honey

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

 1 Number of patients with clinical evidence of wound infection1Risk Difference (M-H, Fixed, 95% CI)Totals not selected

    1.1 Day 7
1Risk Difference (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]

    1.2 Day 21
1Risk Difference (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]

 2 Number of wounds completely healed1Risk Difference (M-H, Fixed, 95% CI)Totals not selected

    2.1 week 2
1Risk Difference (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]

    2.2 week 4
1Risk Difference (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]

    2.3 week 6
1Risk Difference (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]

 3 Number of patients with clinical evidence of wound healing (day 21)150Risk Difference (M-H, Fixed, 95% CI)-0.16 [-0.31, -0.01]

 4 Number of patients reporting free of pain1Risk Difference (M-H, Fixed, 95% CI)Totals not selected

    4.1 week 1
1Risk Difference (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]

    4.2 week 2
1Risk Difference (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]

    4.3 week 3
1Risk Difference (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]

    4.4 week 4
1Risk Difference (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]

 
Comparison 6. SSD cream (1%) vs liposome hydrogel with polyvinyl-pyrrolidone iodine

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

 1 Number of patients that developed wound infection186Risk Difference (M-H, Fixed, 95% CI)0.0 [-0.04, 0.04]

 2 Number of patients with adverse effects186Risk Difference (M-H, Fixed, 95% CI)0.02 [-0.05, 0.10]

 3 Number of patients reporting wound pain186Risk Difference (M-H, Fixed, 95% CI)-0.02 [-0.16, 0.12]

 
Comparison 7. SSD cream (1%) vs collagenase ointment applied with polymyxin B sulfate/bacitrin (Santyl®)

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

 1 Number of patients that developed wound infection1158Risk Difference (M-H, Fixed, 95% CI)-0.01 [-0.12, 0.10]

 2 Number of patients reporting pain1158Risk Difference (M-H, Fixed, 95% CI)-0.19 [-0.31, -0.07]

 
Comparison 8. SSD cream (1%)/chlorhexidine (0.2%) (Silverex) vs diphenyldantoin (Phenytoin)

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

 1 Number of patients with positive cultures (day 10)164Risk Difference (M-H, Fixed, 95% CI)0.38 [0.17, 0.58]

 2 Number of wounds completely healed164Risk Difference (M-H, Fixed, 95% CI)-0.16 [-0.34, 0.02]

 3 Number of patients reporting moderate to severe pain164Risk Difference (M-H, Fixed, 95% CI)0.31 [0.09, 0.54]

 
Comparison 9. Nanocrystalline silver-coated dressing (Acticoat®) vs hydrophilic polyurethane dressing (Allevyn®)

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

 1 Number of patients that developed wound infection132Risk Difference (M-H, Fixed, 95% CI)0.0 [-0.11, 0.11]

 2 Number of wounds healed by day of discharge132Risk Difference (M-H, Fixed, 95% CI)-0.69 [-0.92, -0.45]

 
Comparison 10. Silver nitrate (0.5%) vs Ringer's lactate

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

 1 Number of patients that developed wound infection134Risk Difference (M-H, Fixed, 95% CI)-0.43 [-0.72, -0.14]

 
Comparison 11. Silver nitrate (0.5%) vs neomycin with bacitracin

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

 1 Number of patients that developed wound infection137Risk Difference (M-H, Fixed, 95% CI)-0.23 [-0.49, 0.03]

 
Comparison 12. SSD/SILVER vs NO SILVER

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

 1 Number of patients that developed wound infection10Risk Difference (M-H, Random, 95% CI)Subtotals only

 
Comparison 13. SSD cream (1%) vs nanocrystalline silver-coated dressing (Acticoat®)

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

 1 Number of patients that developed wound infection150Risk Difference (M-H, Fixed, 95% CI)0.04 [-0.15, 0.23]

 2 Mean background pain scores150Mean Difference (IV, Fixed, 95% CI)1.0 [0.64, 1.36]

 3 Mean length of hospital stay150Mean Difference (IV, Fixed, 95% CI)0.0 [-6.43, 6.43]

 
Comparison 14. SSD cream (1%) vs hydrofibre dressing containing ionic silver (Aquacel® Ag)

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

 1 Number of patients that developed wound infection182Risk Difference (M-H, Fixed, 95% CI)-0.04 [-0.20, 0.12]

 2 Number of patients with re-epithelialisation within 21 days182Risk Difference (M-H, Fixed, 95% CI)-0.14 [-0.34, 0.06]

 3 Number of patients reporting adverse effects182Risk Difference (M-H, Fixed, 95% CI)-0.03 [-0.24, 0.19]

 4 Number of patients using systemic antibiotics182Risk Difference (M-H, Fixed, 95% CI)-0.04 [-0.20, 0.12]

 5 Total costs of clinical care (USD)182Mean Difference (IV, Fixed, 95% CI)140.80 [-216.12, 497.72]

 
Comparison 15. SSD cream (1%) vs synthetic dressing containing silver (Hydron AgSD (1-3%))

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

 1 Number of patients with positive cultures1196Risk Difference (M-H, Fixed, 95% CI)0.14 [0.01, 0.28]

 
Comparison 16. SSD cream (1%) (Flamazine®) vs SSD (1%) with 0.2% chlorhexidine digluconate cream (Silvazine®)

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

 1 Number of patients that developed wound infection1121Risk Difference (M-H, Fixed, 95% CI)-0.01 [-0.14, 0.13]

 2 Number of patients that received antibiotics1121Risk Difference (M-H, Fixed, 95% CI)0.10 [-0.03, 0.24]

 3 Number of patients reporting extreme pain at application1121Risk Difference (M-H, Fixed, 95% CI)-0.02 [-0.07, 0.03]

 
Comparison 17. SSD cream (1%) (Flamazine®) vs SSD (1%) cerium nitrate (2.2%) (SSD-CN) (Flammacerium®)

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

 1 Sepsis after 10 days160Risk Difference (M-H, Fixed, 95% CI)0.1 [-0.02, 0.22]

 2 Number of patients reporting subjective stinging effect160Risk Difference (M-H, Fixed, 95% CI)-0.37 [-0.58, -0.15]

 3 Number of patients receiving systemic antibiotics for at least 7 days160Risk Difference (M-H, Fixed, 95% CI)-0.03 [-0.20, 0.13]

 4 Mean length of hospital stay160Mean Difference (IV, Fixed, 95% CI)7.40 [-1.69, 16.49]

 
Comparison 18. SSD cream (1%) (Silvadene®) vs Dimac containing SSD (Sildimac®)

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

 1 Number of patients that developed clinical wound sepsis1102Risk Difference (M-H, Fixed, 95% CI)0.02 [-0.05, 0.09]

 2 Number of patients reporting local adverse effects1118Risk Difference (M-H, Fixed, 95% CI)0.03 [-0.10, 0.16]

 
Comparison 19. Nanocrystalline silver-coated dressing (Acticoat®) vs fine-mesh gauze with silver nitrate (0.5%)

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

 1 Number of patients that developed wound infection134Risk Difference (M-H, Fixed, 95% CI)-0.65 [-0.89, -0.40]

 2 Number of patients that developed bacteraemia134Risk Difference (M-H, Fixed, 95% CI)-0.24 [-0.48, 0.01]

 3 Mean overall painscore160Mean Difference (IV, Fixed, 95% CI)-0.28 [-0.93, 0.37]

 
Comparison 20. SSD cream (1%) vs bacitracin zinc ointment

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

 1 Number of patients that developed wound infection1208Risk Difference (M-H, Fixed, 95% CI)0.07 [-0.01, 0.14]

 
Comparison 21. SSD cream (1%) vs neomycin sulfate

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

 1 Number of patients that developed wound infection1209Risk Difference (M-H, Fixed, 95% CI)0.08 [0.00, 0.15]

 
Comparison 22. SSD cream (1%) vs petrolatum

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

 1 Number of patients that developed wound infection1207Risk Difference (M-H, Fixed, 95% CI)-0.05 [-0.15, 0.04]

 
Comparison 23. Hydrofibre dressing containing ionic silver (Aquacel® Ag) vs povidone iodine gauze

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

 1 Number of patients that developed wound infection167Risk Difference (M-H, Fixed, 95% CI)-0.01 [-0.17, 0.14]

 2 Number of wounds completely healed at end of treatment167Risk Difference (M-H, Fixed, 95% CI)0.13 [-0.04, 0.31]

 3 Number of patients that reported adverse effects167Risk Difference (M-H, Fixed, 95% CI)-0.09 [-0.21, 0.02]

 
Comparison 24. SSD cream (1%) vs benzoic acid, salicylic acid and Quercus rubra extract (Bensal HP)

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

 1 Number of wounds healed (6 weeks)140Risk Difference (M-H, Fixed, 95% CI)-0.10 [-0.39, 0.19]

 
Comparison 25. Activated-charcoal dressing containing silver (Actisorb Plus®) vs conventional phase-adapted therapy using diverse topical modalities

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

 1 Number of wounds healed (6 weeks)138Risk Difference (M-H, Fixed, 95% CI)0.21 [-0.04, 0.46]

 
Comparison 26. Hydrofibre dressing containing ionic silver (Aquacel® Ag) vs calcium alginate dressing (Algosteril®)

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

 1 Number of patients that developed wound infection1134Risk Difference (M-H, Fixed, 95% CI)0.04 [-0.07, 0.16]

 2 Time to complete healing1134Mean Difference (IV, Fixed, 95% CI)-5.10 [-5.69, -4.51]

 3 Number of wounds completely healed during study1134Risk Difference (M-H, Fixed, 95% CI)0.09 [-0.06, 0.24]

 4 Percentage ulcer area reduction in 8 weeks1134Mean Difference (IV, Fixed, 95% CI)-2.40 [-18.72, 13.92]

 5 Ulcer depth reduction in 8 weeks (cm)1100Mean Difference (IV, Fixed, 95% CI)0.12 [-0.05, 0.29]

 6 Number of patients that experienced adverse effects1134Risk Difference (M-H, Fixed, 95% CI)-0.01 [-0.18, 0.15]

 
Comparison 27. SSD cream (1%)/hydrocolloid vs hydrocolloid

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

 1 Number of patients that developed wound infection1166Risk Difference (M-H, Fixed, 95% CI)-0.02 [-0.06, 0.02]

 
Comparison 28. SSD cream (1%)/hydrocolloid vs non-occlusive paraffin-impregnated gauze

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

 1 Number of patients that developed wound infection1184Risk Difference (M-H, Fixed, 95% CI)-0.06 [-0.10, -0.01]