Search Strategy and Selection Criteria
A systematic review was conducted according to the Cochrane Collaboration guidelines. We performed a comprehensive search for randomized controlled trials (RCTs) using as descriptors the names of the antifungals of interest (amorolfine, bifonazole, butenafine, ciclopiroxolamine, clotrimazole, econazole, fenticonazole, flutrimazole, isoconazole, ketoconazole, miconazole, naftifine, oxiconazole, sertaconazole, terbinafine and tioconazole) combined with the terms 'random', 'controlled trial' and 'controlled clinical trial'. The terms 'vaginal', 'vulvovaginal' and 'oropharyngeal' were included in the search preceded by the Boolean operator 'NOT', so as to include only studies assessing the topical use of interventions. The search strategies used for each database are described in Appendix S1 (see Supporting information).
The following databases were included in the search: Medline, Cochrane Central Register of Controlled Trials, EMBASE, Lilacs and the International Pharmaceutical Abstracts. Studies published up to July 2010 in English, Spanish or Portuguese, that compared the use of topical antifungals in the treatment of dermatomycosis with the use of other antifungals or with placebo were included.
Two reviewers (I.R. and A.S.) independently selected studies based initially on their title and abstract. Only RCTs that evaluated the treatment of any dermatomycosis, including cutaneous candidiasis and that of tinea pedis, corporis, cruris and versicolor were included. We excluded onychomycosis from the study as its treatment duration is much longer when compared with other dermatomycosis. Tinea capitis was also excluded as it is not treated with topical antifungals.
Only trials that evaluated mycological cure based on microscopy and/or culture to establish the presence of fungal species were included. The intervention consisted of any topical antifungal regardless of pharmaceutical dosage form, concentration, drug regimen, duration of therapy or pharmacological class. We excluded studies that used a crossover methodology.
Data Extraction and Assessment of Trial Quality
Data extraction was performed by two reviewers (I.R. and A.S.) independently and included study design, baseline characteristics, health intervention, drug regimen, efficacy, safety and tolerability. Any discrepancies in data collection were resolved through consensus and a third reviewer (C.J.C.) was consulted when necessary.
The efficacy outcomes evaluated were mycological cure at the end of treatment, defined as cure obtained until 7 days following the end of treatment, and sustained cure, defined as cure maintained for at least 14 days after the conclusion of treatment. When more than one sustained cure result was described, we prioritized those for which the time interval of monitoring was greater. For both outcomes, the cure had to be confirmed by culture and/or microscopy. The clinical cure rate was not evaluated as this is a subjective outcome. Trials that only showed this rate were excluded.
The safety outcomes evaluated were the occurrence of adverse events and tolerability, including withdrawals due to adverse events.
The methodological quality of each RCT included was evaluated using a method assessment tool published by Jadad et al., described in Appendix S2 (see Supporting information). Only studies with a score of 3 or more were included. To assess the risk of bias in the included studies, we used the Cochrane Collaboration tool, an evaluation in which critical assessments are separated into different domains in order to consider the following types of bias: selection, performance, detection, attrition, reporting and other biases.
For the efficacy outcome, we used the random-effects model and inverse variance method to pool the odds ratios (ORs) from individual studies. For the safety and tolerability outcomes, we employed the statistical method of Peto, which is appropriate when the number of observed events is small and similar in both experimental and control groups, as expected when evaluating adverse events associated with topical therapy with antifungal agents. The OR and the corresponding 95% confidence interval (CI) were calculated and pooled using a fixed-effects model.
The heterogeneity of treatment effects was evaluated by the inconsistency index (I 2). Values of I 2 lower than 25% were considered to show low heterogeneity, whereas values between 25 and 50% were considered moderate to high. In meta-analyses with I 2 > 50% (high heterogeneity), sensitivity analyses were performed to determine whether the study characteristics and statistical methods could have influenced the results. A sensitivity analysis was conducted by the hypothetical removal of each study from the meta-analysis and evaluation of its impact on the overall result. Moreover, the studies were pooled into subgroups based on the dermatomycosis evaluated. All analyses were carried out using Review Manager v. 5·1 statistical software (https://ims.cochrane.org/revman/download).
The British Journal of Dermatology. 2012;166(5):927-933. © 2012 Blackwell Publishing