Summary of the evidence
The review protocol included a population of adults, young people and children with infected leg ulcers. There was minimal evidence for this population (2 small studies), therefore the population was expanded to people with leg ulcers that had an unclear infection status or were not infected. For antiseptic and antibiotics, the results have been presented separately for people with:
an infected leg ulcer
a leg ulcer with unclear infection status
an uninfected leg ulcer.
All the evidence is based on 1 systematic review of antibiotics and antiseptics for venous leg ulcers (O'Meara et al. 2014), which included 45 randomised controlled trials (RCTs). Nine of these were not included in the review because 8 contained ineligible interventions and 1 study was withdrawn from publication. Seven RCTs included people exclusively with leg ulcer infection (however 5 of these RCTs had an uncertain definition of infection); 14 RCTs included people with leg ulcers of unclear infection status and 15 studies included people with leg ulcers that were not infected.
No studies included in the review stated that they included children. The committee discussed that leg ulcer infection in children and young people is extremely rare, and usually a result of an underlying illness that requires specialist management. Therefore, they agreed not to extrapolate the evidence to children and young people.
Standard care is the care given in addition to the intervention and/or the control. The included studies were limited because the definition of standard care for each study varied widely, full details of what composed standard care is noted in the GRADE tables (appendix H of the evidence review).
In a single RCT, cadexomer-iodine was significantly better than standard care at reducing the average size of the ulcers, the amount of pain experienced from the ulcers and reducing or eliminating the presence of Staphylococcus aureus at 6 weeks.
There was no significant difference for:
cadexomer-iodine compared with silver dressing for the frequency of complete healing at 12 weeks and for participant satisfaction (neither group reported any adverse effects)
povidone-iodine plus compression compared with moist or foam dressings plus compression for complete healing at 4 months.
Cadexomer-iodine (topical application) was significantly better than standard care (varied by RCT) for the frequency of complete healing at 4 to 12 weeks, mean percentage change ulcer area and mean rate of ulcer healing. However, adverse events were significantly more common in the cadexomer-iodine group.
Cadexomer-iodine was not significantly different from hydrocolloid dressing or paraffin gauze for the frequency of complete healing at 12 weeks, neither group reported any adverse effects.
Povidone-iodine plus compression was not significantly different from hydrocolloid plus compression for the frequency of complete healing at 4 months.
Povidone-iodine 10% solution plus compression was significantly better for time to healing than hydrocolloid plus compression.
Benzoyl peroxide (10% and 20%) was significantly better than a saline dressing for reducing average ulcer size at 42 days. Data on adverse effects were limited and poorly reported.
Honey (calcium alginate dressing impregnated with Manuka honey) was not significantly different from standard care for:
complete healing at 12 weeks
incidence of ulcer infection during 12 weeks of treatment.
There were significantly more adverse effects in the honey group than the standard care group.
Silver dressing plus compression was significantly better than non‑adhesive plus compression dressing for:
complete healing at 9 weeks
proportion of adults who were pain free at the end of the trial.
Silver dressings were not significantly different from non-adhesive dressings for adverse effects.
There was no significant difference between the following comparisons for complete healing (4 to 12 weeks):
silver sulfadiazine (1% cream) plus compression compared with non-adherent dressing plus compression
silver impregnated dressings (with or without compression) compared with non-antimicrobial dressings (with or without compression)
silver-impregnated polyurethane foam dressing plus compression compared with 5-layer silver impregnated dressing plus compression.
Silver dressings were not significantly different from non-antimicrobial dressings for adverse effects.
There was no significant difference for:
silver sulfadiazine (1% cream) with non-adhesive foam dressing and compression compared with placebo cream with non-adherent dressing and compression for complete healing at 4 weeks
silver sulfadiazine (1% cream) compared with standard care for median time to healing
silver dressing plus compression compared with low-adherent dressing for complete healing at 4 to 12 weeks, 6 months or 12 months, or for ulcer recurrence within 12 months
silver dressings compared with non-antimicrobial dressings for adverse effects.
Silver dressing plus compression was significantly better than non-antimicrobial dressings plus compression for reducing ulcer surface area when measured using square centimetres at 4 weeks, but was not significantly different when measured as a percentage change. The healing rate (cm2 per day) in these 2 RCTs was not significantly different.
Ciprofloxacin was not significantly different from standard care for the frequency of complete healing, emergence of antibiotic-resistant strains or bacterial eradication at 3 months.
For the frequency of complete healing (unclear follow-up time), there was no significant difference between:
ciprofloxacin and placebo or
trimethoprim and placebo.
Emergence of resistance was significantly higher with ciprofloxacin than with placebo, but there was no significant difference in the emergence of resistance with trimethoprim compared with placebo.
There was no significant difference between:
systemic antibiotics (co-trimoxazole, gentamicin or amikacin according to sensitivities) and standard care for the outcomes of complete healing at 3 weeks, complete eventual healing or bacterial eradication
topical mupirocin compared with standard care for frequency of complete healing at 12 weeks or for the eradication of gram-positive bacteria.
Data on adverse effects were limited and poorly reported.
Ciprofloxacin was not significantly different from trimethoprim for the frequency of complete healing.
Limited data on adverse effects were reported. However, ciprofloxacin and trimethoprim increased the emergence of antimicrobial resistance compared with standard care or placebo. This finding was statistically significant for ciprofloxacin, but did not reach significance for trimethoprim.