Topical treatments for fungal infections of the skin and nails of the foot

Fay Crawford; Sally Hollis

doi:10.1002/14651858.CD001434.pub2

. 2007 Jul 18;2007(3):CD001434. doi: 10.1002/14651858.CD001434.pub2

Topical treatments for fungal infections of the skin and nails of the foot.

Fay Crawford ^1,^✉, Sally Hollis ²

Editor: Cochrane Skin Group

PMCID: PMC7073424 PMID: 17636672

Abstract

Background

Fungal infections of the feet normally occur in the outermost layer of the skin (epidermis). The skin between the toes is a frequent site of infection which can cause pain and itchiness. Fungal infections of the nail (onychomycosis) can affect the entire nail plate.

Objectives

To assess the effects of topical treatments in successfully treating (rate of treatment failure) fungal infections of the skin of the feet and toenails and in preventing recurrence.

Search methods

We searched the Cochrane Skin Group Specialised Register (January 2005), the Cochrane Central Register of Controlled Trials (The Cochrane Library Issue 1, 2005), MEDLINE and EMBASE (from inception to January 2005). We screened the Science Citation Index, BIOSIS, CAB ‐ Health and Healthstar, CINAHL DARE, NHS Economic Evaluation Database and EconLit (March 2005). Bibliographies were searched.

Selection criteria

Randomised controlled trials (RCTs) using participants who had mycologically diagnosed fungal infections of the skin and nails of the foot.

Data collection and analysis

Two authors independently summarised the included trials and appraised their quality of reporting using a structured data extraction tool.

Main results

Of the 144 identified papers, 67 trials met the inclusion criteria. Placebo‐controlled trials yielded the following pooled risk ratios (RR) of treatment failure for skin infections: allylamines RR 0.33 (95% CI 0.24 to 0.44); azoles RR 0.30 (95% CI 0.20 to 0.45); ciclopiroxolamine RR 0.27 (95% CI 0.11 to 0.66); tolnaftate RR 0.19 (95% CI 0.08 to 0.44); butenafine RR 0.33 (95% CI 0.24 to 0.45); undecanoates RR 0.29 (95% CI 0.12 to 0.70). Meta‐analysis of 11 trials comparing allylamines and azoles showed a risk ratio of treatment failure RR 0.63 (95% CI 0.42 to 0.94) in favour of allylamines. Evidence for the management of topical treatments for infections of the toenails is sparser. There is some evidence that ciclopiroxolamine and butenafine are both effective but they both need to be applied daily for prolonged periods (at least one year). The six trials of nail infections provided evidence that topical ciclopiroxolamine has poor cure rates and that amorolfine might be substantially more effective but more research is required.

Authors' conclusions

Placebo‐controlled trials of allylamines and azoles for athlete's foot consistently produce much higher percentages of cure than placebo. Allylamines cure slightly more infections than azoles and are now available OTC. Further research into the effectiveness of antifungal agents for nail infections is required.

Plain language summary

Creams, lotions and gels (topical treatments) for fungal infections of the skin and nails of the foot

We found lots of evidence to show fungal skin infections of the skin of the feet (athlete's foot or tinea pedis) are effectively managed by over the counter topical antifungal creams, lotions and gels. The most effective topical agent was terbinafine. Other topical agents such as azoles, ciclopiroxolamine, butenafine, tolnaftate and undecanoate were also effective in curing athlete's foot.

Evidence for the management of topical treatments for management of dermatophyte infections of the toenails was sparser and the studies are small. There was some evidence that ciclopiroxolamine and butenafine are both effective but they both needed to be applied daily for prolonged periods (at least one year).

Background

Description of the condition

Dermatophyte is a collective term for the most common type of fungi which cause infection of the skin and nail. Dermatophytes have the ability to invade keratinised tissue (skin, hair and nails). Infection is normally restricted to the outermost layer of the skin (epidermis). The skin between the toes is a frequent site of fungal infection (athlete's foot or tinea pedis and this can cause pain and itchiness. The skin may become white and macerated and vesicles (small blisters) may form. These can erupt and spread to other areas of the foot especially the soles where the area becomes reddened and raw. Additionally, patches of hard thickened skin occur on the soles, heels and side of the feet. This can lead to splits (fissures) in the skin. Fungal infections of the nail (onychomycosis) can affect the entire nail plate, and one, several or all nails of the feet can be infected simultaneously. The nails often appear changed in colour; they may be thickened and changed in texture (Beaven & Brooks 1994). Fungal infections of the nail are often associated with a skin infection, in which case they can act as a source of reinfection if only the skin is treated (Petit 1983).

Epidemiology

Fungal infections of the skin and nails of the foot are common, reflecting the contagious nature of the organisms. They are thought to occur when individuals regularly use communal changing rooms and swimming pools. Some groups of workers, e.g. coal miners, have been found to have a prevalence of 80% (Roberts 1992). However, people living in institutions with shared bathing facilities such as boarding schools and long term care hospitals also show a higher than average prevalence of this condition (Roberts 1992). The prevalence of onychomycosis has been suggested to increase with age (Roberts 1992) and to be present at a rate of about 5% in people aged 55 years and older.

Description of the intervention

Clinicians faced with a public demand for effective treatment for these conditions face a difficult task as the conditions can be resistant to treatment (Brautigam et al 1995). Whilst these superficial infections are not life threatening, chronic fungal infections of the skin and nails carry a considerable morbidity.

There is wide variation in the methods of treating fungal infections of the skin and toe nails of the foot which reflects the uncertainty surrounding efficacy. Uncertainty also extends to the optimal period of treatment, appropriate dosage of drug and frequency of application. Topical preparations are much less costly than orally administered antifungal drugs and cause minimal adverse side effects. However, whilst they may be helpful in treating the symptoms in localised skin infections, uncertainty exists as to their effectiveness in the complete eradication of the infecting organisms. The ideal topical antifungal for the treatment of fungal infection should be fungicidal so that treatment can be of short duration, it should obtain high cure rates, minimise relapses, be conducive to participant compliance and have minimal adverse effects.

Why it is important to do this review

Fungal infections of the feet are treated by dermatologists, general practitioners and podiatrists. A systematic review of the various therapies used in the topical treatment of fungal infections affecting the skin and nails of the foot will help to inform the treatment approach of all these professionals.

Objectives

To identify and evaluate the evidence for topical treatments for fungal infections of the skin and nails of the foot.   To establish the effectiveness of topical treatments used for fungal infections of the skin and nails compared with other treatments or untreated controls.

Methods

Criteria for considering studies for this review

Types of studies

All randomised controlled studies of topical treatment for fungal infections of the skin and nails of the foot.

Types of participants

All men and women of any age who have a fungal infection of the skin or nails of human foot which has been identified by microscopy and growth of dermatophytes in culture.

Types of interventions

Any programme of treatments administered topically to treat fungal infections of the feet compared with other treatments, placebo or no treatments. All types of intervention were considered.

Types of outcome measures

Primary outcomes

1. For each trial we calculated the treatment failure rate at follow up from the reported mycological results (defined as negative results on microscopy and no growth of dermatophyte in culture). In the update of the review we have only included studies which subject all skin and nail samples to KOH (potassium hydroxide) and culture).  2. Quality of life as measured by the cosmetic acceptability of the end result to the participant, absence of itchiness, independence from medical treatment and advice with respect to the condition.

Secondary outcomes

1. Measurement of recurrence of the condition in:  (a) skin ‐ maintenance of cure 12 weeks after initiation of intervention;  (b) nail ‐ maintenance of cure 36 weeks after initiation of intervention.  2. Side effects as measured by the frequency of reported adverse events.

Search methods for identification of studies

Electronic searches

We searched the Cochrane Skin Group Specialised Register (January 2005) search strategy in Appendix 1:

We searched the Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library Issue 1, 2005) using the search strategy in Appendix 2

We searched MEDLINE (OVID) (from inception to January 2005) using the strategy in Appendix 3.

We searched EMBASE (from inception to January 2005) using the following keywords: athlete's foot, tinea pedis, topical treatment and onychomycosis.

Other databases

We searched the following databases in March 2005 using the term 'athlete's foot' in each:  Science Citation Index and Social Science Citation Index within BIDS  CAB‐Health and Healthstar  The online versions of DARE  NHS Economic Evaluation Database  EconLit  Online ARC version of CINAHL

Searching other resources

References from published studies

We searched the bibliographies of all the papers identified by these strategies.

Handsearching of journals

We handsearched podiatry journals which were not listed in the electronic databases, namely, The Foot, The Journal of British Podiatric Medicine and The Journal of the American Podiatry Association.

Unpublished literature

We searched for unpublished or unlisted studies by contacting all Schools of Podiatry in the UK and made a request for dissertation bibliographies. Where possible we established personal communication with corresponding trial authors of papers identified through the above strategy to enquire about other relevant studies. We contacted the pharmaceutical industry to request reports of further published and unpublished trials.  We searched the Current Controlled Trials Register (on www.controlled‐trials.com) and www.clinicaltrials.gov in January 2005 for ongoing trials.

Language

We did not impose any language restrictions and sought translations where necessary.

Adverse Effects

We looked at the included RCTs for reports of adverse effects of the interventions. We have summarised our findings in the body of the review.

Data collection and analysis

Selection of studies

One author searched for trials (FC). Two authors checked titles and abstracts identified from the searches. One author (FC) obtained the full text of all studies of possible relevance for independent assessment. One author decided which trials fit the inclusion criteria and recorded their methodological quality using a structured data extraction tool. Any disagreement was resolved by discussion between the authors. Excluded studies and reasons for exclusion are stated.

Assessment of risk of bias in included studies

Assessment of methodological quality of included studies

Assessment of methodological quality was performed. The following areas were addressed, since these may be associated with biased estimates of treatment effect (Juni 2001):  (a) the method of allocation;  (b) the identity of study participants who were blind;  (d) the loss to follow‐up and exclusions;  (e) selective reporting;  (f) other forms of bias as detailed below:  (i) whether the aims were clearly defined;  (ii) whether a prior sample size calculation was reported;  (iii) whether the inclusion and exclusion criteria were defined  (iv) whether the baseline comparability of groups was reported (based on age, sex, and duration of complaint);  (v) whether interventions were defined;  (vi) whether the compliance was assessed;  (vii) were the infecting fungi identified;  (viii) was the distribution of species between groups stated;  (ix) were adverse events reported.

These items are reported in the Characteristics of included studies.

Measures of treatment effect

We reported results are reported as risk ratios (RR) of treatment failure with 95% confidence intervals (CI). To estimate differences between treatment regimens, we pooled trials that evaluated similar interventions. Since it was anticipated that there would be substantial heterogeneity between trials, we used random‐effects models when pooling.

Assessment of heterogeneity

Heterogeneity was assessed using I². If substantial heterogeneity (I² > 50%) existed between studies for the primary outcomes, reasons for heterogeneity, such as language, differences in health care systems, dosage and duration of treatment, were explored.

For major comparisons where overall pooling of the results was considered potentially appropriate, the results were also illustrated using L'Abbé plots (L'Abbe 1985). For each study, This plot showed the observed treatment failure rate in one group plotted against observed treatment failure rate in the other group. This was useful to illustrate the range of treatment failure rates among the trials, and the amount of heterogeneity between trials and sub groups. The diagonal line indicates no treatment effect (RR = 1). Points below this line correspond to trials where the treatment on the vertical axis has a lower treatment failure rate. The dashed line shows the pooled treatment effect. Each trial is plotted using a circle with size proportional to the standard error of the treatment effect so that trials providing more precise estimates are shown using larger circles.

We listed non‐randomised controlled studies but did not discuss them further.

Data synthesis

Statistical analysis

The primary outcome i.e. rate of treatment failure, was extracted for three time points, considered to reflect clinically important timings. Each time point is analysed separately with sub comparisons within each treatment comparison consistently numbered as below.

1. Short‐term (two weeks): the order of preference was two, one, three weeks.  2. Medium‐term (six weeks): the time point closest to six weeks follow‐up was used, provided that it was within four to eight weeks. If two time points equally close to six weeks were available then the longer follow‐up was used (e.g. eight rather than four weeks). Where a final follow‐up endpoint with more participants included was available with four to eight weeks, this was used preferentially. Only > 80% follow‐up included. A sensitivity analysis on the medium term data, restricted to data where there was clear documentation of at least 80% follow‐up of randomised participants (where follow‐up is given by group, at least 80% follow‐up was required in both groups).  3. Long‐term (12 weeks onwards): longest follow‐up of at least 12 weeks.

Sub comparisons were used to group together treatments by duration (note that for short‐term outcomes, grouping by duration was not necessary, since all durations were at least two weeks).

Sensitivity analysis

Sensitivity analyses were also conducted to examine the effects of excluding studies with poor quality.

Results

Description of studies

Results of the search

We considered all RCTs that evaluated topical treatments for fungal infections of the skin and nails of the foot. For skin infections we included only trials that used microscopy and culture to establish the presence of dermatophytes. For nail infections we included only trials that used culture to do so. We included duplicate trials only once. We excluded trials on fungal infections that contained data on infections at various body sites if foot‐specific data could not be extracted separately.

(a) Identified trials relating to skin of the foot

Included studies

We identified 144 papers reporting trials of topical treatments for fungal skin infections and included 67 (Ablon 1996; Akers 1989; Aly 2003; Bagatell 1986; Bagatell 1991a; Bagatell 1991b; Bergstresser 1993; Berman 1992; Bojanovsky 1985; Carter 1972; Chretien 1980; Coffey 1986; Del Palacio 1989; Dobson 1989; Elewski 1996; Ellis 1989; Evans 1991; Evans 1993a; Evans 1993b; Evans 1994; Friederich 1992; Fuerst 1980; Gentles 1974; Gomez 1986; Haas 1985; Hollmen 2002; Holti 1970; Ison 1990; Izuno 1986; Kagawa 1985; Klaschka 1984; Kligman 1985a; Kligman 1985b; Korting 1997; Kuhlwein 1990; Ledezma 2000; Leenutaphong 1999; Mandy 1974; Pereda 2003; Plotkin 1990; Qadripur 1979; Roberts 1985; Sanchez 1994; Satchell 2002; Savin 1990; Savin 1994; Savin 1997; Schachner 1990; Schopf 1999; Smith 1977; Smith 1986; Smith 1988a; Smith 1988b; Smith 1988c; Smith 1990a; Smith 1990b; Smith 1992; Spiekermann 1976a; Spiekermann 1976b; Sushka 2001; Syed 2000; Tong 1992; Tschen 1997; Vermeer 1996; Weller 1998; Woscoff 1986; Zaug 1992).

Comparisons

Twenty‐nine trials compared a single active treatment with placebo (Akers 1989; Aly 2003; Bagatell 1986; Bagatell 1991a; Bagatell 1991b; Berman 1992; Chretien 1980; Coffey 1986; Dobson 1989; Evans 1991; Gentles 1974; Gomez 1986; Ison 1990; Izuno 1986; Klaschka 1984; Kligman 1985a; Korting 2001; Mandy 1974; Savin 1990; Savin 1994; Savin 1997; Schachner 1990; Smith 1977; Smith 1986; Smith 1988a; Smith 1990a; Spiekermann 1976a; Spiekermann 1976b; Tschen 1997).

Twenty‐five trials compared two active treatment regimens (Bojanovsky 1985; Carter 1972; Del Palacio 1989; Evans 1993a; Evans 1993b; Friederich 1992; Haas 1985; Holti 1970; Kagawa 1985; Kligman 1985b Kuhlwein 1990; Leenutaphong 1999; Plotkin 1990; Pereda 2003; Qadripur 1979; Roberts 1985; Sanchez 1994 Schopf 1999 Smith 1988b; Smith 1988c; Smith 1992; Sushka 2001; Vermeer 1996; Weller 1998; Woscoff 1986).

Thirteen trials compared more than two treatment regimens within the same trial (Ablon 1996; Bergstresser 1993; Elewski 1996; Ellis 1989; Evans 1994; Fuerst 1980; Korting 1997;Ledezma 2000; Satchell 2002; Smith 1990b; Syed 2000; Tong 1992; Zaug 1992).

Demographic information is presented for all studies in the Characteristics of included studies table.

Excluded studies

We excluded 77 trials and present the reasons for these exclusions in the Characteristics of excluded studies table.

Twelve studies evaluating topical treatments for skin infections which were included in the previous version of this review were excluded in this update (Daily 1985; Duncan 1975; Thomas 1976; Ortiz 1978; Tschen 1979; Smith 1977b; Fredriksson 1982; Privat 1982; Thomas 1986; Greer 1986; Tanenbaum 1982; Tsuboi 1996). These exclusions were made after a tightening of the review inclusion criteria to include only those studies that used both microscopy and culture to diagnose dermatophyte infections.

(b) Results of the search for trials relating to nails of the foot

Included studies

We identified 11 trials evaluating the efficacy of topical treatments for nails and included 6 in the review (Buck 1994; Gupta 2000a; Gupta 2000b; Mensing 1992; Montana 1994; Syed 1999). Please see Characteristics of included studies.

Excluded studies

The other five were excluded since they reported combined data from fingernails and toenails (Lauharanta 1992; Reinel 1992; Reinel 1992a; Ruping 1993; Terragni 1993). We also excluded four trials evaluating a combination of systemic and topical treatments for infected nails (Arenas 1991; Baran 2000; Barnetson 1998; Friedman 1997; Zaug 1995). Please see Characteristics of excluded studies.

Risk of bias in included studies

Allocation

The method of allocation was reported in only 17 trials (Ablon 1996; Akers 1989; Buck 1994; Chretien 1980; Evans 1993a; Evans 1993b; Evans 1994; Fuerst 1980; Gentles 1974; Holti 1970; Kligman 1985a; Kligman 1985b; Korting 1997; Mandy 1974; Plotkin 1990; Savin 1997; Smith 1977).

Blinding

Blinded outcome assessment was reported in only 12 trials (Ablon 1996; Buck 1994; Carter 1972; Evans 1993a; Evans 1994; Gentles 1974; Holti 1970; Korting 2001; Mensing 1992; Montana 1994; Savin 1990; Savin 1997). However only 12 did not report blinding of participants (Ablon 1996; Bojanovsky 1985; Friederich 1992; Leenutaphong 1999; Mensing 1992; Montana 1994; Roberts 1985; Sanchez 1994; Satchell 2002; Satchell 2002; Smith 1988b; Smith 1988c). One trial was reported to be single blind (Kagawa 1985). However only seven trials did not report blinding of subjects (Smith 1988a; Smith 1988b; Smith 1988c; Friederich 1992; Sanchez 1994; Roberts 1985; Friederich 1992).

Incomplete outcome data

Follow up and exclusions

A diagnosis of athlete's foot based on clinical signs and symptoms alone can be inaccurate because there are non‐fungal skin conditions which have a similar appearance, e.g. erythrasma. This review only included trials which reported the use of microscopy and culture tests to confirm the presence of fungi. The results from these lab‐based diagnostic tests can take up to several weeks to obtain, and often trial participants are randomised to an allocation and begin treatment before the test results are available. When the results show no fungi are present (negative test result) participants are then withdrawn from the study. In order to reduce bias from trials with high loss to follow up whilst recognising the practical constraints in which RCTs of athlete's foot generally take place, we performed a sensitivity analysis only including data which reported follow up data for at least 80% of the randomised sample.

Selective reporting

We found no evidence that selective reporting had occurred in any of the included trials.

Other potential sources of bias

Some trials were funded by industry however a sensitivity analysis of data from trials comparing allylamines versus azoles found no statistical differences between industry funded and non‐industry funded studies.

Effects of interventions

(a) Skin trials

(i) Placebo controlled trials

Allylamines Versus Placebo

Two different allylamines (naftifine 1% and terbinafine 1%) used for 1 to 4 weeks were evaluated in 11 placebo controlled randomised trials.

Short‐term outcome (two weeks)

Naftifine  Short‐term outcomes were available for all 5 trials using naftifine (n = 612) Klaschka 1984; Dobson 1989; Schachner 1990; Bagatell 1991a; Bagatell 1991b. Overall the observed relative reduction in risk of treatment failure was 25% (RR 0.75, 95% CI 0.60 to 0.93; Analysis 1.1), although there was substantial variation in the individual study results (I² = 79%).

1.1 — Comparison 1 Allylamines vs Placebo, Outcome 1 Short term (2 weeks) treatment failure.

Terbinafine  For terbinafine, short‐term outcomes were available for 4 trials (n = 316, Berman 1992; Evans 1991; Korting 1997; Syed 2000). The results were inconsistent between studies, giving an overall relative reduction in treatment failure of 42% which was not statistically significant (RR 0.58, 95% CI 0.31 to 1.08; Analysis 1.1).

Across all 9 trials providing short‐term outcome of 1% allylamines for a period of 1 to 2 weeks compared with placebo, there was a pooled relative reduction in treatment failure at 2 weeks of 31% (RR 0.69, 95% CI 0.56 to 0.87; Analysis 1.1), with substantial variation in individual study results (I² = 79%).

Medium‐term outcome (six weeks)

Naftifine  Naftifine (1%) used for 4 weeks was evaluated in 5 trials (n = 607, Bagatell 1991a; Bagatell 1991b; Dobson 1989; Klaschka 1984; Schachner 1990), a 58% relative reduction in treatment failure was observed (RR 0.42 95% CI 0.30 to 0.59; Analysis 1.2, Naftifine (tx 4 weeks) versus Placebo), with substantial variation in individual study results (I² = 68%).

1.2 — Comparison 1 Allylamines vs Placebo, Outcome 2 Medium term (6 weeks) treatment failure.

Terbinafine  Terbinafine (1%) was used for 1 week (2 trials, n = 229, Berman 1992; Korting 2001), 2 weeks (2 trials, n = 240, Evans 1991; Savin 1994) and 4 weeks (2 trials, n = 40, Savin 1990; Smith 1990a). A statistically significant reduction in risk of treatment failure was observed with each treatment duration (RR 0.16, 0.36, 0.18 respectively; Analysis 1.2). Pooling across all durations, a 77% relative reduction in treatment failure was observed (RR 0.23, 95% CI 0.15 to 0.38, pooled result not shown in the Forest plots (MetaView), with moderate variation in individual study results (I² = 50%).

A meta‐analysis of data from all 11 trials (n = 1116) comparing 1% allylamines with placebo (treatment for a period of 1 to 4 weeks) provided an estimated relative reduction in the risk of treatment failure of 67% (RR 0.33, 95% CI 0.24 to 0.44; Analysis 1.2), with substantial variation in individual study results (I² = 67%). All of the results were based on at least 80% follow up except those from the trials evaluating terbinafine used for 2 weeks. A sensitivity analysis based on the exclusion of data collected in these two trials (Evans 1991; Savin 1994) showed a similar overall estimate of effectiveness (RR of treatment failure 0.31, 95% CI 0.21 to 0.45; Analysis 1.3). The variation in individual study results remained substantial (I² = 75%).

1.3 — Comparison 1 Allylamines vs Placebo, Outcome 3 Medium term (6 weeks) treatment failure. Only >80% follow‐up included.

A L'Abbé plot of the outcomes at 6 weeks (Figure 1) demonstrates that the allylamines generally had treatment failure rates of around 30%, compared to around 85% for placebos, though there was considerable variation in individual trial results (see MeMethods, Assessment of heterogeneity for more details of L'Abbé plots and how to interpret them) .

Azoles Versus Placebo

Six different azoles (bifonazole, clotrimazole, miconazole nitrate, oxiconazole nitrate, sulconazole nitrate and tioconazole) were evaluated in 13 placebo controlled randomised trials. The concentration of these drugs was generally 1%, but 2% for miconazole, they were used for 4 to 6 weeks.

Short‐term outcome (two weeks)

Bifonazole  Short‐term outcome was reported only in the four trials of bifonazole, and one trial of oxiconazole. A meta‐analysis of the 4 trials evaluating bifonazole (n = 176, Bagatell 1986; Coffey 1986; Izuno 1986; Smith 1986) estimated the relative reduction in treatment failure at 2 weeks to be 48% (RR 0.52, 95% CI 0.37 to 0.73; Analysis 2.1), with good consistency between the individual trial results (I² = 18%).

2.1 — Comparison 2 Azoles (tx 4‐6 weeks) vs Placebo, Outcome 1 Short term (2 weeks) treatment failure.

Oxiconazole  The trial of oxiconazole (n = 155, Ellis 1989) had an observed relative reduction in treatment failure at 2 weeks of 21% (RR 0.79, 95% CI 0.61 to 1.02; Analysis 2.1) which was not quite statistically significant. Overall the pooled estimated relative risk of treatment failure (n = 329, 5 trials) of 1% azoles (bifonazole or oxiconazole) versus placebo at 2 weeks was 41% (RR 0.59, 95% CI 0.43 to 0.82; Analysis 2.1), though there was considerable variation between the results for the 2 different azoles (overall I² = 50%).

Medium‐term outcome (six weeks)

All azoles versus placebo  There was statistically significant evidence of effectiveness for each of the azoles individually except miconazole nitrate, which was studied in only 54 people, providing inconclusive results (RR of treatment failure 0.41, 95% CI 0.14 to 1.14; Analysis 2.2) (Gentles 1974; Mandy 1974). The results were generally fairly consistent both between individual studies for each azole, and between different azoles. Overall, a meta‐analysis of data from 13 trials (n = 1235, Akers 1989; Bagatell 1986; Coffey 1986; Elewski 1996; Ellis 1989; Izuno 1986; Gentles 1974; Mandy 1974; Smith 1977; Smith 1986; Smith 1988a; Spiekermann 1976a; Spiekermann 1976b) comparing azoles with placebo estimated the pooled relative reduction in treatment failure as 60% (RR 0.40, 95% CI 0.35 to 0.46; Analysis 2.2).

2.2 — Comparison 2 Azoles (tx 4‐6 weeks) vs Placebo, Outcome 2 Medium term (6 weeks) treatment failure.

Sensitivity analysis

A sensitivity analysis based only on data collected in the 6 trials with at least 80% follow up at 6 weeks (n = 448, Akers 1989; Ellis 1989; Gentles 1974; Mandy 1974; Smith 1977; Smith 1988a) showed very similar results (RR of treatment failure 0.43 (95% CI 0.34 to 0.53; Analysis 2.3).

2.3 — Comparison 2 Azoles (tx 4‐6 weeks) vs Placebo, Outcome 3 Medium term (6 weeks) treatment failure. Only >=80% follow‐up included.

A L'Abbé plot of the outcomes at 6 weeks (Figure 2) demonstrates that the azoles generally had treatment failure rates of around 25%, compared to around 50 to 90% for placebos (see Methods, statistical analysis for more details of L'Abbé plots and how to interpret them).

Short‐term outcomes (two weeks)

Butenafine  Outcomes collected at two weeks in three placebo controlled trials of butenafine 1% used for 1 to 2 weeks (n = 271, Savin 1997; Tschen 1997; Syed 2000) showed a statistically significant relative reduction in treatment failure at 2 weeks of 57% (RR 0.43, 95% CI 0.23 to 0.78; Analysis 3.1).

3.1 — Comparison 3 Other topical antifungal treatments vs Placebo, Outcome 1 Short term (2 weeks) treatment failure.

Ciclopiroxolamine  A placebo controlled trial of 1% ciclopiroxolamine (n = 168, Kligman 1985a) shows a statistically significant relative reduction in treatment failure at 2 weeks of 57% (RR 0.43, 95% CI 0.27 to 0.68; Analysis 3.1).

Undecanoates  A placebo controlled trial of undecanoates (n = 168, Chretien 1980) also showed a statistically significant effect with relative reduction in treatment failure at 2 weeks of 86% (RR 0.14, 95% CI 0.06 to 0.31; Analysis 3.1).

Medium‐term outcome (six weeks)

Butenafine  A statistically significant effect of butenafine(1%) was observed when it was used for 1 week and for 4 weeks. Butenafine used for 1 week versus placebo was evaluated (n = 271) in one trial (Savin 1997). A statistically significant relative reduction in treatment failure of 67% was observed (RR 0.33, 95% CI 0.24 to 0.45; Analysis 3.2). Butenafine used for 4 weeks versus placebo was evaluated (n = 80) in another trial (Tschen 1997), giving a statistically significant relative reduction in treatment failure of 81% (RR 0.19, 95% CI 0.08 to 0.43; Analysis 3.2). Neither of these trials achieved at least 80% follow‐up.

3.2 — Comparison 3 Other topical antifungal treatments vs Placebo, Outcome 2 Medium term (6 weeks) treatment failure.

Ciclopiroxolamine  Ciclopiroxolamine (1% and 0.77%) used for 4 weeks was evaluated in 2 placebo controlled trials (n = 144, Kligman 1985a; n = 317, Aly 2003). A statistically significant relative reduction in treatment failure was observed (RR of treatment failure = 0.27, 95% CI 0.11 to 0.66; Analysis 3.2).

Tee tree oil  Tea tree oil (10%) used for 4 weeks was evaluated in 2 placebo controlled trials (n = 185, Satchell 2002; Tong 1992). Although one of the individual trials showed a statistically significant effect (Satchell 2002), the results of the second trial were less favourable, and combining data from both trials did not show a statistically significant effect (RR of treatment failure 0.73, 95% CI 0.48 to 1.11;Analysis 3.2). Only the trial with the less favourable results (Tong 1992) had at least 80% follow‐up.

Tolciclate  Tolciclate (1%) used for 6 weeks was evaluated in a small placebo controlled trial (n=40, Gomez 1986) and produced a relative reduction of treatment failure of 0.04 (95% CI 0.00 to 0.63; Analysis 3.2), a statistically significant effect, with at least 80% follow‐up.

Tolnaftate  Tolnaftate (1%) used for 4 weeks was compared with placebo in 2 trials (n = 115, Fuerst 1980; Tong 1992) and a statistically significant relative reduction in treatment failure of 70% was found (RR 0.30, 95% CI 0.13 to 0.72; Analysis 3.2), with at least 80% follow‐up in both trials.

Undecanoates  Undecanoates (Undecylenic acid, zinc undecylenic acid) were compared with placebo in two trials (n = 125, Chretien 1980; Fuerst 1980) and a statistically significant relative reduction in treatment failure of 71% was found (RR 0.29, 95% CI 0.12 to 0.70; Analysis 3.2), with at least 80% follow‐up in both trials.

(ii) Treatment versus treatment comparisons

Comparisons between Different Allylamines or Allylamine Regimens

Four trials compared the rate of treatment failure of different allylamines or allylamine regimens (all 1%, Ablon 1986; Bergstresser 1993; Evans 1994; Smith 1990b).

Short‐term outcome (two weeks)

Naftifine twice daily  In the trial by Smith 1990b (n = 117) no statistically significant difference was observed at 2 weeks between naftifine used once daily or twice daily (RR of treatment failure at 2 weeks 0.92, 95% CI 0.72 to 1.17; Analysis 4.1).

4.1 — Comparison 4 Comparisons between different allylamines or allylamine regimes, Outcome 1 Short term (2 weeks) treatment failure.

Naftifine versus terbinafine  In the evaluation of naftifine compared with terbinafine (n = 62, Ablon 1996) there were similar treatment failure rates at 2 weeks for each of the 2 treatments (RR of treatment failure 0.98, 95% CI 0.69 to 1.41; Analysis 4.1).

Medium‐term outcome (six weeks)

Naftifine twice daily versus naftifine once daily  Smith 1990b compared 1% naftifine once daily to twice daily both for 4 weeks (n = 101) and found fewer treatment failures with twice daily, thought the difference was not statistically significant (RR of treatment failure = 1.74, 95% CI 0.82 to 3.67; Analysis 4.2).

4.2 — Comparison 4 Comparisons between different allylamines or allylamine regimes, Outcome 2 Medium term (6 weeks) treatment failure.

Naftifine versus Terbinafine  Ablon 1996 compared naftifine with 1% terbinafine, both applied for 2 weeks (n = 62), the results favoured terbinafine but the difference was not statistically significant (RR of treatment failure 2.05, 95% CI 0.77 to 5.42; Analysis 4.2). Evans 1994 compared terbinafine applied for 5 to 7 days with terbinafine applied for 1 to 2 days (n = 65), no statistical difference was detected in the treatment failure rates (RR of treatment failure 0.89, 95% CI 0.31 to 2.50; Analysis 4.2).

Terbinafine four week versus terbinafine one week  Bergstresser 1993 compared terbinafine used for 4 weeks with terbinafine used for 1 week (n = 83) but also did not detect any difference between the rate of treatment failure in the 2 groups (RR of treatment failure 1.06, 95% CI 0.42 to 2.66; Analysis 4.2). All of these results are based on at least 80% follow‐up except those of Bergstresser 1993.

Long‐term outcome (12 weeks)

Terbinafine one to three versus five to seven days  Data collected in the trial of terbinafine used for 5 to 7 days versus terbinafine used for 1 to 3 days (n = 65, Evans 1994) found slightly less treatment failures with 5 to 7 days treatment, but the difference was not statistically significant (RR of treatment failure at 12 weeks 0.53, 95% CI 0.15 to 1.88).

Terbinafine four weeks versus terbinafine one week

A trial comparing outcomes from 1% terbinafine used for 4 weeks versus 1% terbinafine used for 1 week (n = 80, Bergstresser 1993) found less slightly treatment failures with 4 weeks treatment, but the difference was not statistically significant (RR of treatment failure at 12 weeks 0.60, 95% CI 0.24 to 1.54).

Comparisons between different azoles or azole regimens

Thirteen trials reported data collected in evaluations of one azole versus another (Bergstresser 1993; Elewski 1996; Ellis 1989; Friederich 1992; Korting 1997; Kuhlwein 1990; Pereda 2003; Qadripur 1979; Roberts 1985; Smith 1988b; Smith 1988c; Sushka 2001; Woscoff 1986).

Short‐term outcome (two weeks)

Clotrimazole versus econazole  In 2 trials of 1% clotrimazole versus 1% econazole (n = 497, Korting 1997; Qadripur 1979) combined data did not show a statistically significant difference (RR of treatment failure at 2 weeks 1.13, 95% CI 0.92 to 1.39; Analysis 5.1).

5.1 — Comparison 5 Comparisons between different azoles or azole regimes, Outcome 1 Short term (2 weeks) treatment failure.

Econazole gel versus econazole cream  In 2 arms of the trial by Korting 1997 (n = 229) 1% econazole gel was compared with gel 1% econazole cream, there was no statistically significant difference (RR of treatment failure 0.86, 95% CI 0.66 to 1.23; Analysis 5.1).    Miconazole nitrate versus sulconazole nitrate  Woscoff 1986 evaluated 2% miconazole nitrate versus 1% sulconazole nitrate in a small trial (n = 41), the results favoured sulconazole nitrate but were not statistically significant (RR of treatment failure at 2 weeks 6.30, 95% CI 0.83 to 47.80; Analysis 5.1).    Oxiconazole once versus twice daily  The trial of 1% oxiconazole used twice daily versus once daily (Ellis 1989) did not detect a statistically significant difference between the 2 regimens; RR of treatment failure at 2 weeks 1.05 (95% CI 0.74 to 1.49; Analysis 5.1).

Medium‐term outcome (six weeks)

Most trials achieved at least 80% follow‐up at 6 weeks, apart from Bergstresser 1993 and Smith 1988c.

Bifonazole  Bifonazole versus croconazole  One trial (Kuhlwein 1990) found 100% rate of treatment success in both arms of a small trial (n = 36) comparing 1% bifonazole for 3 weeks with 1% croconazole for 3 weeks.

Bifonazole used for three weeks versus bifonazole used for one hour for three weeks  Another trial evaluated 1% bifonazole used for 3 weeks versus 1% bifonazole removed after one hour for 3 weeks (n = 73, Friederich 1992) but did not detect any difference between the rates of treatment failure (RR 4.34, 95% CI 0.51 to 37.01; Analysis 5.2).

5.2 — Comparison 5 Comparisons between different azoles or azole regimes, Outcome 2 Medium term (6 weeks) treatment failure.

Bifonazole versus miconazole  A third trial, Roberts 1985, compared 1% bifonazole with 2% miconazole used for 3 weeks but did not observe a statistically significant difference between the 2 treatments (RR of treatment failure 0.80, 95% CI 0.27 to 2.37; Analysis 5.2).

Bifonazole versus flutrimazole  A fourth trial Pereda 2003 (n = 264) compared bifonazole 1% powder with fluconazole 1% powder and found bifonazole to produce statistically significantly fewer treatment failures (0.21, 95% CI 0.14 to 0.31; Analysis 5.2).

Clotrimazole  Clotrimazole for four weeks versus clotrimazole for one week  Four trials evaluated 1% clotrimazole (Bergstresser 1993; Korting 1997; Qadripur 1979; Sushka 2001). Bergstresser 1993 compared 4 weeks of 1% clotrimazole with 1 week of 1% clotrimazole (n = 89) and found a statistically significant relative reduction in treatment failure of 64% in favour of 4 weeks treatment (RR of treatment failure = 0.36, 95% CI 0.18 to 0.72; Analysis 5.2), less than 80% follow‐up was achieved in this trial.

Clotrimazole versus econazole  Two trials (n = 497, Qadripur 1979; Korting 1997) comparing 1% clotrimazole versus 1% econazole each individually had statistically significant results favouring each of the treatments, the pooled results did not show a statistically significant difference (RR of treatment failure = 0.95, 95% CI 0.31 to 2.88; Analysis 5.2).

Clotrimazole versus ketoconazole  Sushka 2001 compared clotrimazole 1% used once daily with 2% ketoconazole 2% used twice daily but the results did not show a statistically significant difference between the 2 creams (RR of treatment failure = 1.06, 95% CI 0.63 to 1.76; Analysis 5.2).

Econazole gel versus econazole cream  One trial (n = 313, Korting 1997) evaluated 1% econazole gel versus 1% econazole cream but did not show a statistically significant difference in rates of treatment failure (RR of treatment failure = 0.78, 5% CI 0.53 to 1.16; Analysis 5.2).

Miconazole versus ticonazole  Combined data from 2 trials comparing 2% miconazole used for 6 weeks versus 1% tioconazole used for 6 weeks (n = 220, Smith 1988a; Smith 1988c) but did not show a statistically significant difference in the two treatments, RR of treatment failure 1.29 (95% CI 0.77 to 2.16; Analysis 5.2).

Oxiconazole + fluctic versus oxiconazole  A trial of 1% oxiconazole + fluctic used for 4 weeks versus 1% oxiconazole used for 4 weeks (n = 201, Elewski 1996) produced a statistically significant relative reduction in treatment failure of 47% (RR 0.53, 95% CI 0.33 to 0.87; Analysis 5.2), favouring oxiconazole + fluctic.

Oxiconazole once per day versus oxiconazole twice per day  Ellis 1989 evaluated oxiconazole once per day compared with oxiconazole twice per day (n = 110) but found no statistically significant difference in the treatment failure rates (RR 1.23, 95% CI 0.61 to 2.46; Analysis 5.2).

Long‐term outcome (12 weeks+)

Clotrimazole four weeks versus clotrimazole one week  In a trial comparing 4 weeks of 1% clotrimazole with 1 week of 1% clotrimazole Bergstresser 1993 found a statistically significant relative reduction in treatment failure of 53% favouring the longer treatment time (RR at 12 weeks 0.47, 95% CI 0.28 to 0.78; Analysis 5.4).

5.4 — Comparison 5 Comparisons between different azoles or azole regimes, Outcome 4 Long term (12 weeks onwards) treatment failure.

Allylamines one to two weeks versus azoles one to two weeks

Short‐term outcome (two weeks)

Ten trials (n = 1519, Ablon 1996; Bojanovsky 1985; Evans 1993a; Evans 1993b; Haas 1985; Leenutaphong 1999; Sanchez 1994 Schopf 1999; Smith 1990b; Smith 1992) comparing 1% allylamines (naftifine, terbinafine) with 1% azoles (bifonazole, clotrimazole, oxiconazole) used for 1 to 2 weeks showed a small, not statistically significant, difference in favour of allylamines; RR of treatment failure at 2 weeks 0.86 (95% CI 0.70 to 1.06; Analysis 6.1). There was considerable variation in the results of the individual trials (I² = 60%).

6.1 — Comparison 6 Allylamines 1‐2 weeks vs Azoles 1‐2 weeks, Outcome 1 Short term (2 weeks) treatment failure.

A L'Abbé plot of the outcomes at two weeks (Figure 3) demonstrates that the allylamines generally had similar treatment failure rates to the azoles, with a wide variation in failure rates.

Medium‐term outcome (six weeks)

Data collected at 6 weeks from 2 trials (n = 1730, Ablon 1996; Bergstresser 1993) comparing 1% allylamines (naftifine, terbinafine) versus 1% azoles used for 1 to 2 weeks (oxiconazole, clotrimazole) produced a statistically significant relative reduction in treatment failure of 66% (RR 0.34, 95% CI 0.22 to 0.52; Analysis 6.2), favouring allylamines. Only one of these trials had follow up data of at least 80%of those originally randomised (Ablon 1996), but the results of both trials were very similar and these data provide a similar relative risk of treatment failure (0.33, 95% CI 0.20 to 0.56; Analysis 6.3).

6.2 — Comparison 6 Allylamines 1‐2 weeks vs Azoles 1‐2 weeks, Outcome 2 Medium term (6 weeks) treatment failure.

6.3 — Comparison 6 Allylamines 1‐2 weeks vs Azoles 1‐2 weeks, Outcome 3 Medium term (6 weeks) treatment failure. Only >=80% follow‐up included..

A L'Abbé plot of the outcomes at 6 weeks (Figure 4) shows the allylamines had treatment failure rates of around 15% to 20%, compared to 50% to 70% for the azoles (Methods, Assessment of heterogeneity for more details of L'Abbé plots and how to interpret them).

Long term outcome (12 weeks)

Data collected at 12 weeks from one trial (n = 75, Bergstresser 1993) comparing 1% allylamines (terbinafine) with 1% azoles (clotrimazole) used for 1 to 2 weeks produced a statistically significant relative reduction in treatment failure of 72% (RR 0.28, 95% CI 0.14 to 0.58; Analysis 6.4), favouring allylamines.

6.4 — Comparison 6 Allylamines 1‐2 weeks vs Azoles 1‐2 weeks, Outcome 4 Long term (12 weeks onwards) treatment failure.

Allylamine one week versus azoles four weeks

Medium‐term outcome (six weeks)

Data collected at 6 weeks in 5 trials (n = 962, Bergstresser 1993; Evans 1993a; Leenutaphong 1999; Schopf 1999; Vermeer 1996) which compared a 1% allylamine (terbinafine) used for 1 week with an 1% azole used for 4 weeks (clotrimazole, miconazole) did not show a statistically significant difference in treatment failure (RR 0.75, 95% CI 0.33 to 1.72; Analysis 7.1). There was considerable variation in the results of the individual trials (I² = 60%). Combining data from 3 of the trials which had at least 80% follow‐up (n = 685, Evans 1993b; Leenutaphong 1999; Schopf 1999) also did not show a statistically different difference in the treatment failure rates (RR 0.50, 95% CI 0.10 to 2.54; Analysis 7.2).

7.1 — Comparison 7 Allylamines 1 week vs Azoles 4 weeks, Outcome 1 Medium term (6 weeks) treatment failure.

7.2 — Comparison 7 Allylamines 1 week vs Azoles 4 weeks, Outcome 2 Medium term (6 weeks) treatment failure. Only >=80% follow‐up included..

A L'Abbé plot of the outcomes at six weeks (Figure 5) demonstrates that the allylamines had broadly similar treatment failure rates to the azoles, with a wide variation in failure rates.

Allylamines versus azoles four to six weeks

Medium‐term outcome (six weeks)

A meta analysis of data collected in nine trials (n = 1003, Bergstresser 1993; Bojanovsky 1985; Evans 1993a; Haas 1985; Kagawa 1985; Plotkin 1990; Sanchez 1994; Smith 1990b; Smith 1992) found a statistically significant difference between 1% allylamines (naftifine, terbinafine) and 1% to 2% azoles (bifonazole, clotrimazole) used for 4 to 6 weeks, with a relative reduction in treatment failure of 37% favouring allylamines (RR 0.63, 95% CI 0.42 to 0.94; Analysis 8.1). There was considerable variation in the results of the individual trials (I² = 68%) Six of those trials (Evans 1993a; Haas 1985; Kagawa 1985; Smith 1990a; Smith 1992) reported outcomes for at least 80% of those randomised (n = 896) and produced a relative risk of treatment failure of 0.55 (95% CI 0.34 to 0.89; Analysis 8.2) for the superiority of allylamines over azoles. The variation in the results of the individual trials was still substantial (I² = 70%).

8.1 — Comparison 8 Allylamines 4‐6 weeks vs Azoles 4‐6 weeks, Outcome 1 Medium term (6 weeks) treatment failure.

8.2 — Comparison 8 Allylamines 4‐6 weeks vs Azoles 4‐6 weeks, Outcome 2 Medium term (6 weeks) treatment failure. Only >=80% follow‐up included..

A L'Abbé plot of the outcomes at six weeks (Figure 6) demonstrates a wide variation in observed treatment failure rates.

Long‐term outcome (12 + weeks)

Two trials (n = 141, Bergstresser 1993; Evans 1993a) collected long term outcomes from comparisons of allylamines versus azoles, the combined data produced a relative risk of treatment failure of 0.47 (95% CI 0.22 to 1.02; Analysis 8.3) which favoured allylamines, but did not quite reach statistical significance.

8.3 — Comparison 8 Allylamines 4‐6 weeks vs Azoles 4‐6 weeks, Outcome 3 Long term (12 weeks onwards) treatment failure.

Allylamines versus other antifungal topical skin treatments

Two small trials compared an allylamine (terbinafine) with another topical skin treatment.

Short‐term outcome (two weeks)

Terbinafine versus butenafine  One trial comparing 1% terbinafine versus 1% butenafine (Syed 2000 n = 40) found no statistically significant difference in treatment failure at 2 weeks (RR 2.00, 95% CI 0.41 to 9.71; Analysis 9.1)

9.1 — Comparison 9 Allylamines vs Other antifungal topical skin treatments, Outcome 1 Short term (2 weeks) treatment failure.

Medium‐term outcome (six weeks)

Terbinafine versus ajoene  Ledezma 2000 compared terbinafine 1% for 1 week with ajoene 0.6% and 1.0% (n = 47). The treatment failure rate for terbinafine was between that for 0.6% and 1.0% ajoene, but neither concentration was significantly different to terbinafine.

Azoles versus other antifungal topical skin treatments

Medium‐term outcome (six weeks)

Bifonazole versus amorolfine  In a very small trial comparing 1% bifonazole versus 0.5% amorolfine applied for 6 weeks (n = 9, Del Palacio 1989) all participants in each group were cured (no treatment failure) at 6 weeks; 6/6 in the bifonazole arm and 3/3 in the amorolfine arm.

Clotrimazole versus ciclopiroxolamine  A trial comparing 1% clotrimazole versus 1% ciclopiroxolamine applied to the skin for 4 weeks (n = 87, Kligman 1985b) produced a relative risk of treatment failure of 1.41 (95% CI 0.67 to 2.95; Analysis 10.1) in outcomes taken at 2 weeks. Outcomes taken at 6 weeks found a relative risk of 1.78 (95% CI 0.59 to 5.38; Analysis 10.2), but less than 80% follow up was available at this time. At 4 weeks there was 100% follow‐up, and an RR of treatment failure of 1.71 (95% CI 0.54 to 5.42; Analysis 10.3).

10.1 — Comparison 10 Azoles vs Other antifungal topical skin treatments, Outcome 1 Short term (2 weeks) treatment failure.

10.2 — Comparison 10 Azoles vs Other antifungal topical skin treatments, Outcome 2 Medium term (6 weeks) treatment failure.

10.3 — Comparison 10 Azoles vs Other antifungal topical skin treatments, Outcome 3 Medium term (6 weeks) treatment failure. Only >=80% follow‐up included..

Comparison between other topical treatments

Short‐term outcome (two weeks)

Salicylic acid plus nitrate versus salicylic acid  Weller 1998 reported outcomes collected at 2 weeks with statistically significantly fewer treatment failures when nitrite was added to salicylic acid, RR of treatment failure 0.09 (95% CI 0.01 to 0.62; Analysis 11.1).

11.1 — Comparison 11 Comparisons between other topical antifungal treatments, Outcome 1 Short term (2 weeks) treatment failure.

Medium‐term outcome (six weeks)

Ajoene 0.6% and 1.0%  Ledezma 2000 compared ajoene 0.6% and 1.0% (n = 29), and found less treatment failures with 1.0% although the difference was not quite statistically significant (RR 0.07, 95% CI 0.00 to 1.17; Analysis 11.2). There was less than 80% follow‐up.

11.2 — Comparison 11 Comparisons between other topical antifungal treatments, Outcome 2 Medium term (6 weeks) treatment failure.

Amorolfine 0.125, 0.25 and 0.5.  Zaug 1992 compared amorolfine 0.125, 0.25 and 0.5 (n = 337), and found little difference in any of the treatment failure rates, follow‐up was at least 80%.

Halprogen versus tolnaftate  In a trial of halprogen compared with 1% tolnaftate, both used for 4 weeks (n = 69), Carter 1972 reported outcomes at 6 weeks and found halprogen to provide a statistically significant relative reduction in treatment failure of 64% (RR 0.36, 95% CI 0.19 to 0.69; Analysis 11.4), follow‐up was at least 80%.

11.4 — Comparison 11 Comparisons between other topical antifungal treatments, Outcome 4 Medium term (6 weeks) treatment failure. Only >=80% follow‐up included..

Salicylic acid plus nitrate versus salicylic acid  Weller 1998 evaluated salicylic acid plus nitrite versus salicylic acid both used for 4 weeks (n = 35) and found that the addition of nitrite provided a statistically significant relative reduction in treatment failure of 54% (RR 0.46, 95% CI 0.22 to 0.96; Analysis 11.4), follow‐up was at least 80%.

Tea tree oil versus tea tree oil  A trial comparing 50% tea tree oil versus 25% tea tree oil (n = 69, Satchell 2002) did not show a statistically significant difference in treatment failures (RR 0.79, 95% 0.45 to 1.41; Analysis 11.2).

Tea tree oil versus tolnaftate  In a comparison of 10% tea tree oil versus 1% tolnaftate (n = 70), Tong 1992 found tolnaftate to be associated with statistically significantly fewer treatment failures (RR of treatment failure = 4.64 (95% 2.01 to 10.68; Analysis 11.2), follow‐up was at least 80%.

Tolnaftate versus undecanoates  In a comparison of 1% tolnaftate versus undecanoates (n = 40), Fuerst 1980 found no difference in treatment failure rates (RR 1.07, 95% 0.42 to 2.68; Analysis 11.2).    Long‐term outcome (12+ weeks)  Whitfield's ointment versus variotin  A very small trial comparing Whitfield's ointment versus variotin applied for 8 weeks (n = 10, Holti 1970) did not detect a statistically significant difference between the 2 treatments (RR of treatment failure 1.33, 95% CI 0.17 to 10.25; Analysis 11.3).

11.3 — Comparison 11 Comparisons between other topical antifungal treatments, Outcome 3 Long term (12 weeks onwards) treatment failure.

Other outcome measures

Primary

No trials reported quality of life as measured by the cosmetic acceptability of the end result to the participant, absence of itchiness, independence from medical treatment and advice with respect to the condition.

Secondary

1. Measurement of recurrence of the condition in:

(a) skin ‐ maintenance of cure 12 weeks after initiation of intervention.

Clotrimazole four weeks versus clotrimazole one week

In a trial comparing 4 weeks of 1% clotrimazole with 1 week of 1% clotrimazole Bergstresser 1993 found a statistically significant relative reduction in treatment failure of 53% favouring the longer treatment time (RR at 12 weeks 0.47, 95% CI 0.28 to 0.78; Analysis 5.4).

Terbinafine one to three versus five to seven days

Data collected in the trial of terbinafine used for 5‐7 days versus terbinafine used for 1‐3 days (n = 65, Evans 1994) found slightly less treatment failures with 5‐7 days treatment, but the difference was not statistically significant (RR of treatment failure at 12 weeks 0.53, 95% CI 0.15 to 1.88; Analysis 4.4).

4.4 — Comparison 4 Comparisons between different allylamines or allylamine regimes, Outcome 4 Long term (12 weeks onwards) treatment failure.

Terbinafine four weeks versus terbinafine one week

Allylamines versus azoles four to six weeks

Long‐term outcome (12+ weeks)

Whitfield's ointment versus variotin

A very small trial comparing Whitfield's ointment versus variotin applied for 8 weeks (n = 10, Holti 1970) did not detect a statistically significant difference between the 2 treatments (RR of treatment failure 1.33, 95% CI 0.17 to 10.25; Analysis 11.3).

Adverse events

Fifty included trials mentioned adverse events in the report; Ablon 1996; Bagatell 1986; Bagatell 1991a; Bagatell 1991b; Bojanovsky 1985; Carter 1972; Del Palacio 1989; Dobson 1989; Elewski 1996; Ellis 1989; Evans 1991; Evans 1993a; Evans 1993b; Evans 1994; Fuerst 1980; Gomez 1986; Gupta 2000a; Gupta 2000b; Holti 1970; Ison 1990; Izuno 1986; Kagawa 1985; Klaschka 1984; Kligman 1985a; Kligman 1985b; Kuhlwein 1990; Ledezma 2000; Mandy 1974; Plotkin 1990; Qadripur 1979; Sanchez 1994; Savin 1990; Savin 1994; Savin 1997; Schachner 1990; Smith 1977; Smith 1986; Smith 1988a; Smith 1988b; Smith 1988c; Smith 1990a; Smith 1990b; Smith 1992; Spiekermann 1976a; Spiekermann 1976b; Syed 1999; Tong 1992; Tschen 1997; Weller 1998; Woscoff 1986; Zaug 1992. Few serious adverse events were reported with the exception of Savin 1997 which found an increase in liver enzymes in one person in the placebo arm of the trial and Smith 1990a which reported elevated liver enzymes in both terbinafine 1% cream and placebo arms of the trial and neutropaenia in one participant in the placebo arm. Most trials which gave detail of the adverse events experience by people participating in the trials reported burning, stinging, itching sensations. There are plans to extend the analysis of adverse events in future review updates.

(b) Nails trials

Placebo comparisons

Ciclopiroxolamine versus placebo

Ciclopiroxolamine lacquer (1%) was compared to placebo in 2 trials (Gupta 2000a; Gupta 2000b) relative risk 0.32; 95% CI 0.20 to 0.52 at 48 weeks (Analysis 12.1).

12.1 — Comparison 12 Treatments for onychomycosis (nail infections), Outcome 1 Treatments versus placebo, treatment failure.

Fungoid tincture versus placebo

A comparison of fungoid tincture versus placebo found a relative risk 0.17; 95% CI 0.02 to 1.14 at 12 months; Analysis 12.1 (Montana 1994)

Butenafine + tea tree oil versus placebo

A comparison of 2% butenafine and 5% tea tree oil versus placebo produced a relative risk 0.03; 95% CI 0.00 to 0.47 at 36 weeks; Analysis 12.1, showing butenafine and tea tree oil to be statistically significantly more effective than placebo alone (Syed 1999).

Treatment versus treatment comparisons

Clotrimazole versus tea tree oil

A trial of 1% clotrimazole solution performed better than 100% tea tree oil, though not significantly (Buck 1994).

Amorolfine 5% + methylene versus amorolfine 5% + ethanol

In 1 small trial 2 amorolfine 5% nail lacquer formulations with different vehicles used twice weekly for 4 weeks both achieved a relative risk of 1.00; 95% CI 0.85 to 1.18; Analysis 12.2, at both 3 and 14 days after the end of treatment. (Mensing 1992)

12.2 — Comparison 12 Treatments for onychomycosis (nail infections), Outcome 2 Comparisons between active treatments, treatment failure.

Adverse events

Adverse events were reported in four trials evaluating topical nail treatments. In the trial of 2% butenafine with 5% tea tree oil 93.3% of participants had no drug‐related negative side effects. Four participants in the active cream group reported mild inflammation which did not lead to discontinuation or interruption of treatment (Syed 1999). In the trials of ciclopiroxolamine 8% nail lacquer adverse events considered by the investigator to be possibly related to the nail lacquer or vehicle were reported to be as follows: 16 participants in the ciclopiroxolamine group developed a rash compared to 3 participants in the vehicle group; 5 participants in the ciclopiroxolamine group developed nail disorders e.g. changes in the nail shape or colour compared to 5 participants in the vehicle group; 3 participants in the ciclopiroxolamine group showed an application site disorder compared to 3 participants in the vehicle group (Gupta 2000a; Gupta 2000b). In the placebo controlled trial of fungoid tincture minimal adverse effects were noted, mild peeling occurred in eight participants and erythema occurred in one. It is not clear from the report to which arm of the trial the affected participants belonged Montana 1994.

Discussion

Summary of main results

Superficial fungal infections affecting the foot are common and often the first line management strategy is the use of topical agents on both skin and nail infections. The review identified good evidence that allylamines, azoles, butenafine, ciclopiroxolamine, tolciclate and tolnaftate are all efficacious relative to placebo in the management of fungal infections of the skin. Allylamines produced evidence of greater effectiveness when used for longer and there is some evidence that the effect of allylamines increases over time. The observations were collected from a large number of participants (11 trials, n = 1116) providing strong evidence that allylamines are very much more effective than placebo in the management of athletes foot, with an estimated relative reduction in treatment failure at 6 weeks of 67% (RR 0.33, 95% CI 0.24 to 0.44; Analysis 1.1).

The effectiveness of azoles was also seen to improve over time; at outcomes collected six weeks after baseline greater effectiveness than outcomes taken earlier (two weeks after baseline). In common with the trials of allylamines versus placebo, azole creams are very much more effective than placebo, with an estimated relative reduction in treatment failure at 6 weeks of 60% (13 trials, n = 1235, RR 0.40, 95% CI 0.35 to 0.46; Analysis 2.2).

Given the strength of the evidence from a large number of trials and people, and the narrow confidence intervals around the estimates for both allylamines and azoles we would not recommend the use of placebo controls in future RCTs evaluating the use of topical treatments for athlete's foot.

Butenafine, ciclopiroxolamine, tolciclate and tolnaftate also showed greater effectiveness than placebo in the treatment of fungal skin conditions. These results are based on a limited number of trials including small numbers of people however and the evidence is consequently less strong than for allylamines and azoles.

Comparisons between different regimes of allylamines provided little evidence that any regime is more effective than another. Different types of allylamines or different doses were not found to have different treatment failure rates. This might be because there are genuinely no differences in the effectiveness of different types or regimes of topical allylamines, or it may because the trials included too few participants to detect differences between groups.

We did not detect any difference in treatment failure rates between any of the individual azoles, but there is some evidence that the length of treatment affects the success of azoles creams: clotrimazole used for four weeks instead of one week was shown to improve its effectiveness in one trial.    Direct comparisons of allylamines versus azoles show allylamines to be generally more efficacious than azoles. Trials directly comparing the two compounds demonstrate the superiority of allylamines. There is little evidence of superiority at 2 weeks but this effect becomes detectable in outcomes taken 6 weeks after treatment begins and appears to remain at 12 weeks. The meta analysis of 8 trials and outcomes from 962 participants supports the finding that allylamines are more effective than azoles when applied for between 4 to 6 weeks. At 6 weeks, there was a relative reduction in treatment failure with allylamines compared to azoles of 37% (RR 0.63 95% CI 0.42 to 0.94; Analysis 8.1).

Meta analyses of randomised controlled trials of allylamines used for one week versus azoles for four weeks demonstrated fairly similar outcomes at six weeks, though there is insufficient evidence to claim that these regimes are equivalent.

Comparisons of other antifungal agents have found that halprogen is more effective than tolnaftate and that combining salicylic acid with nitrite produces greater effectiveness than that of salicylic acid alone. The small trial of Whitfield's ointment and variotin applied for eight weeks only included ten people with athlete's foot and it is therefore unsurprising that the trialists failed to detect differences between these two compounds. Unfortunately there is little evidence to assess tolnaftate against placebo or to compare butenafine, ciclopiroxolamine, tolciclate and tolnaftate with each other.

The review did not find any evidence to support the use of tea tree oil in the management of athlete's foot. Tea tree oil did not produce a greater benefit than placebo, and was significantly less effective than tolnaftate. In a trial of different concentrations of tea tree oil Satchell 2002 did not detect statistically significant differences between concentrations of 50% and 25%. All the randomised evaluations included in this systematic review of tea tree oil suggest that it is ineffective in the management of fungal skin infections.

Evidence about the efficacy of topical treatments for nail infections is sparse. Combining data from 2 trials of ciclopiroxolamine versus placebo found treatments failure rates of 61% and 64% for ciclopiroxolamine. These outcomes followed long treatment times (48 weeks) and this makes ciclopiroxolamine a poor choice for nail infections. Better results were observed with the use of amorolfine lacquer; 6% treatment failure rates were found after 1 month of treatment but these data were collected on a very small sample of people and these high rates of success might be unreliable. Butenafine 2% produced a treatment failure rate of 20%. There is limited evidence about the efficacy of tea tree oil for skin infections; it was evaluated in only one small trial however it was found to be ineffective for fungal nail conditions when compared with topical butenafine.

Quality of the evidence

The randomised controlled trials in this review were generally well reported, and follow up rates were reasonable for such a condition.

Since no trial reported the species obtained from participants who were resistant to treatment we cannot draw conclusions about susceptibility to individual compounds to help clinical decision‐making. Little data was available about the long term outcomes associated with the use of antifungal creams and there is uncertainty about rates of reinfection and relapse.

Authors' conclusions

Implications for practice.

All antifungal compounds demonstrated some success in curing athlete's foot. The best results were observed with the use of allylamines and there is a small amount of evidence that butenafine may be similarly good. Azoles are also very effective and participants should be advised that although all azoles appear to be similarly effective, using an azole cream for four weeks is likely to produce better results than using it for one week. Azoles may also be more efficacious than tolnaftate but they seem no more efficacious than undecanoic acid. There is limited evidence about the efficacy of tea tree oil for skin infections.

There is little evidence that topical anti‐fungals are effective in the management of onychomycosis or fungally infected toe nails. The majority of available data demonstrate low cure rates after long treatment times with ciclopiroxolamine. Amorolfine and butenafine may be much more effective than ciclopiroxolamine and tea tree oil but only a few observations are available.

Implications for research.

The estimates of effectiveness of allylamines and azoles relative to placebo have conclusively demonstrated these drugs to be of greater effectiveness and we recommend that no more placebo controlled trials of allylamines or azoles should be conducted.

More direct comparisons of undecanoic acid and tolnaftate with allylamines and azoles for athlete's foot are required. Large randomised controlled trials comparing the effectiveness of topical amorolfine and butenafine are needed to establish an alternative to oral treatments for toe nail infections.

Feedback

Feedback on section 'Effects of interventions' part (b) Nails trials, 8 January 2009

Summary

In the section 'Effects of interventions' part (b) Nails trials, the authors say "A comparison of 2% butenafine versus 5% tea tree oil produced a relative risk 0.03; 95% CI 0.00 to 0.47 at 36 weeks showing butenafine + tea tree oil to be statistically significantly more effective than tea tree oil alone (Syed 1999).

The study was placebo‐controlled. So, the review should say that butenafine + tea tree oil is more effective than placebo. The way you put it provides evidence that tea tree oil is ineffective. The real situation is that there no evidence one way or the other.

You need to check the graphs and tables for the same error.

NOTE: The submitter agrees with the default conflict of interest statement: I certify that I have no affiliations with or involvement in any organization or entity with a financial interest in the subject matter of my feedback.

Reply

I wish to thank the person who has given us feedback about the mistake in the section 'Effects of interventions' part (b) Nails trials. I have referred to the original paper (Syed 1999) and can confirm that the criticism is valid; the comparisons were 2% Butenafine together with 5% melaleuca alternifolia versus placebo and I agree that the text should read "A comparison of 2% butenafine and 5% tea tree oil produced a relative risk 0.03; 95% CI 0.00 to 0.47 at 36 weeks showing butenafine and tea tree oil to be statistically significantly more effective than placebo alone”.

Fay Crawford (on behalf of all the authors of this review)

As a result of the above dialogue the editorial base made the appropriate changes to the text in the section 'Effects of interventions' part (b) Nails trials. The graph was moved from Analysis 12.2 'Comparisons between active treatments to Analysis12.1 'Treatments versus placebo'.

Contributors

Michael Power

What's new

Date	Event	Description
1 March 2016	Amended	Edited the published note about the updating of the review.

Date	Event	Description
9 February 2016	Amended	This review is going to be updated. We have written a published note to say that we decided to split the topic into topical and device‐based treatments for fungal infections of the toenails and topical treatments for athlete's foot, so a protocol and then a new review will be written.
25 March 2009	Feedback has been incorporated	Feedback in response to comments made by Micheal Power.
25 March 2009	Amended	As a result of the feedback received, Analysis 12.1 'Treatments versus placebo' was amended to include butenafine & tea tree oil versus placebo. The text in 'Effects of interventions' part (b) Nails trials was also amended to include this change.
20 July 2008	Amended	Converted to new review format.
23 May 2007	New citation required and conclusions have changed	Substantive amendment
9 April 2007	New search has been performed	Minor update
23 May 2005	Amended	Reformatted
1 March 2005	Amended	New studies found and included or excluded

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Short term (2 weeks) treatment failure	9	928	Risk Ratio (M‐H, Random, 95% CI)	0.69 [0.56, 0.87]
1.1 Naftifine vs Placebo	5	612	Risk Ratio (M‐H, Random, 95% CI)	0.75 [0.61, 0.93]
1.2 Terbinafine vs Placebo	4	316	Risk Ratio (M‐H, Random, 95% CI)	0.58 [0.31, 1.08]
2 Medium term (6 weeks) treatment failure	11	1116	Risk Ratio (M‐H, Random, 95% CI)	0.33 [0.24, 0.44]
2.1 Naftifine (tx 4 weeks) vs Placebo	5	607	Risk Ratio (M‐H, Random, 95% CI)	0.42 [0.30, 0.59]
2.2 Terbinafine (tx 1 week) vs Placebo	2	229	Risk Ratio (M‐H, Random, 95% CI)	0.16 [0.09, 0.26]
2.3 Terbinafine (tx 2 weeks) vs Placebo	2	240	Risk Ratio (M‐H, Random, 95% CI)	0.36 [0.27, 0.48]
2.4 Terbinafine (tx 4 weeks) vs Placebo	2	40	Risk Ratio (M‐H, Random, 95% CI)	0.18 [0.04, 0.82]
3 Medium term (6 weeks) treatment failure. Only >80% follow‐up included	9	876	Risk Ratio (M‐H, Random, 95% CI)	0.31 [0.21, 0.45]
3.1 Naftifine (tx 4 weeks) vs Placebo	5	607	Risk Ratio (M‐H, Random, 95% CI)	0.42 [0.30, 0.59]
3.2 Terbinafine (tx 1 week) vs Placebo	2	229	Risk Ratio (M‐H, Random, 95% CI)	0.16 [0.09, 0.26]
3.3 Terbinafine (tx 4 weeks) vs Placebo	2	40	Risk Ratio (M‐H, Random, 95% CI)	0.18 [0.04, 0.82]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Short term (2 weeks) treatment failure	5	329	Risk Ratio (M‐H, Random, 95% CI)	0.59 [0.43, 0.82]
1.1 Bifonazole vs Placebo	4	176	Risk Ratio (M‐H, Random, 95% CI)	0.52 [0.37, 0.73]
1.2 Oxiconazole vs Placebo	1	153	Risk Ratio (M‐H, Random, 95% CI)	0.79 [0.61, 1.02]
2 Medium term (6 weeks) treatment failure	13	1235	Risk Ratio (M‐H, Random, 95% CI)	0.40 [0.35, 0.46]
2.1 Bifonazole (tx 4 weeks) vs Placebo	4	182	Risk Ratio (M‐H, Random, 95% CI)	0.36 [0.20, 0.67]
2.2 Clotrimazole (tx 4‐6 weeks) vs Placebo	3	371	Risk Ratio (M‐H, Random, 95% CI)	0.42 [0.27, 0.64]
2.3 Miconazole nitrate (tx 4 weeks) vs Placebo	2	54	Risk Ratio (M‐H, Random, 95% CI)	0.41 [0.14, 1.14]
2.4 Oxiconazole (tx 4 weeks) vs Placebo	2	451	Risk Ratio (M‐H, Random, 95% CI)	0.37 [0.30, 0.46]
2.5 Sulconazole nitrate (tx 4‐6 weeks) vs Placebo	1	117	Risk Ratio (M‐H, Random, 95% CI)	0.48 [0.31, 0.75]
2.6 Ticonazole (tx 4‐6 weeks) vs Placebo	1	60	Risk Ratio (M‐H, Random, 95% CI)	0.37 [0.22, 0.62]
3 Medium term (6 weeks) treatment failure. Only >=80% follow‐up included	6	448	Risk Ratio (M‐H, Random, 95% CI)	0.43 [0.34, 0.53]
3.1 Clotrimazole (tx 6 weeks) vs Placebo	1	63	Risk Ratio (M‐H, Random, 95% CI)	0.56 [0.32, 0.98]
3.2 Miconazole nitrate (tx 4 weeks) vs Placebo	2	54	Risk Ratio (M‐H, Random, 95% CI)	0.41 [0.14, 1.14]
3.3 Oxiconazole (tx 4 weeks) vs Placebo	1	154	Risk Ratio (M‐H, Random, 95% CI)	0.34 [0.23, 0.52]
3.4 Sulconazole nitrate (tx 4‐6 weeks) vs Placebo	1	117	Risk Ratio (M‐H, Random, 95% CI)	0.48 [0.31, 0.75]
3.5 Ticonazole (tx 4‐6 weeks) vs Placebo	1	60	Risk Ratio (M‐H, Random, 95% CI)	0.37 [0.22, 0.62]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Short term (2 weeks) treatment failure	5		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
1.1 Butenafine vs Placebo	3	391	Risk Ratio (M‐H, Random, 95% CI)	0.43 [0.23, 0.78]
1.2 Ciclopiroxolamine vs Placebo	1	168	Risk Ratio (M‐H, Random, 95% CI)	0.43 [0.27, 0.68]
1.3 Undecanoates vs Placebo	1	85	Risk Ratio (M‐H, Random, 95% CI)	0.14 [0.06, 0.31]
2 Medium term (6 weeks) treatment failure	9		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
2.1 Butenafine (tx 1 week) vs Placebo	1	271	Risk Ratio (M‐H, Random, 95% CI)	0.33 [0.24, 0.45]
2.2 Butenafine (tx 4 weeks) vs Placebo	1	80	Risk Ratio (M‐H, Random, 95% CI)	0.19 [0.08, 0.43]
2.3 Ciclopiroxolamine (tx 4 weeks) vs Placebo	2	461	Risk Ratio (M‐H, Random, 95% CI)	0.27 [0.11, 0.66]
2.4 Tea tree oil (tx 4 weeks) vs Placebo	2	185	Risk Ratio (M‐H, Random, 95% CI)	0.73 [0.48, 1.11]
2.5 Tolciclate (tx 6 weeks) vs Placebo	1	40	Risk Ratio (M‐H, Random, 95% CI)	0.04 [0.00, 0.63]
2.6 Tolnaftate (tx 4 weeks) vs Placebo	2	115	Risk Ratio (M‐H, Random, 95% CI)	0.30 [0.13, 0.72]
2.7 Undecanoates (tx 4‐6 weeks) vs Placebo / no treatment	2	85	Risk Ratio (M‐H, Random, 95% CI)	0.29 [0.12, 0.70]
3 Medium term (6 weeks) treatment failure. Only >=80% follow‐up included.	5		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
3.1 Ciclopiroxolamine (tx 4 weeks) vs Placebo	1	144	Risk Ratio (M‐H, Random, 95% CI)	0.39 [0.23, 0.66]
3.2 Tea tree oil (tx 4 weeks) vs Placebo	1	71	Risk Ratio (M‐H, Random, 95% CI)	0.88 [0.68, 1.16]
3.3 Tolciclate (tx 6 weeks) vs Placebo	1	40	Risk Ratio (M‐H, Random, 95% CI)	0.04 [0.00, 0.63]
3.4 Tolnaftate (tx 4 weeks) vs Placebo	2	115	Risk Ratio (M‐H, Random, 95% CI)	0.30 [0.13, 0.72]
3.5 Undecanoates (tx 4‐6 weeks) vs Placebo / no treatment	2	125	Risk Ratio (M‐H, Random, 95% CI)	0.24 [0.08, 0.74]

Outcome or subgroup title	No. of studies	Statistical method	Effect size
1 Short term (2 weeks) treatment failure	2	Risk Ratio (M‐H, Random, 95% CI)	Totals not selected
1.1 Naftifine once daily vs Naftifine twice daily	1	Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
1.2 Naftifine vs Terbinafine	1	Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
2 Medium term (6 weeks) treatment failure	4	Risk Ratio (M‐H, Random, 95% CI)	Totals not selected
2.1 Naftifine once daily (tx 4 weeks) vs Naftifine twice daily (tx 4 weeks)	1	Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
2.2 Naftifine (tx 2 weeks) vs Terbinafine (tx 2 weeks)	1	Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
2.3 Terbinafine (tx 5‐7 days) vs Terbinafine (tx 1 ‐3 days)	1	Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
2.4 Terbinafine (tx 4 weeks) vs Terbinafine (tx 1 week)	1	Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
3 Medium term (6 weeks) treatment failure. Only >=80% follow‐up included.	3	Risk Ratio (M‐H, Random, 95% CI)	Totals not selected
3.1 Naftifine once daily (tx 4 weeks) vs Naftifine twice daily (tx 4 weeks)	1	Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
3.2 Naftifine (tx 2 weeks) vs Terbinafine (tx 2 weeks)	1	Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
3.3 Terbinafine (tx 5‐7 days) vs Terbinafine (tx 1‐3 days)	1	Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
4 Long term (12 weeks onwards) treatment failure	2	Risk Ratio (M‐H, Random, 95% CI)	Totals not selected
4.1 Terbinafine (tx 5‐7 days) vs Terbinafine (tx 1 ‐3 days)	1	Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
4.2 Terbinafine (tx 4 weeks) vs Terbinafine (tx 1 week)	1	Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Short term (2 weeks) treatment failure	10	1519	Risk Ratio (M‐H, Random, 95% CI)	0.86 [0.70, 1.06]
1.1 Naftifine vs Bifonazole	1	98	Risk Ratio (M‐H, Random, 95% CI)	1.07 [0.78, 1.47]
1.2 Naftifine vs Clotrimazole	4	617	Risk Ratio (M‐H, Random, 95% CI)	0.78 [0.53, 1.14]
1.3 Naftinfine/Terbinane vs Oxiconazole	1	90	Risk Ratio (M‐H, Random, 95% CI)	0.84 [0.65, 1.09]
1.4 Terbinafine vs Bifonazole	1	42	Risk Ratio (M‐H, Random, 95% CI)	0.31 [0.14, 0.66]
1.5 Terbinafine vs Clotrimazole	3	672	Risk Ratio (M‐H, Random, 95% CI)	1.16 [0.86, 1.55]
2 Medium term (6 weeks) treatment failure	2	173	Risk Ratio (M‐H, Random, 95% CI)	0.34 [0.22, 0.52]
2.1 Naftinfine/Terbinane vs Oxiconazole	1	90	Risk Ratio (M‐H, Random, 95% CI)	0.33 [0.20, 0.56]
2.2 Terbinafine vs Clotrimazole	1	83	Risk Ratio (M‐H, Random, 95% CI)	0.36 [0.17, 0.75]
3 Medium term (6 weeks) treatment failure. Only >=80% follow‐up included.	1	90	Risk Ratio (M‐H, Random, 95% CI)	0.33 [0.20, 0.56]
3.1 Naftinfine/Terbinane 2 weeks vs Oxiconazole 2 weeks	1	90	Risk Ratio (M‐H, Random, 95% CI)	0.33 [0.20, 0.56]
4 Long term (12 weeks onwards) treatment failure	1	75	Risk Ratio (M‐H, Random, 95% CI)	0.28 [0.14, 0.58]
4.1 Terbinafine vs Clotrimazole	1	75	Risk Ratio (M‐H, Random, 95% CI)	0.28 [0.14, 0.58]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Medium term (6 weeks) treatment failure	5	962	Risk Ratio (M‐H, Random, 95% CI)	0.75 [0.33, 1.72]
1.1 Terbinafine vs Clotrimazole	4	771	Risk Ratio (M‐H, Random, 95% CI)	0.64 [0.24, 1.75]
1.2 Terbinafine vs Miconazole	1	191	Risk Ratio (M‐H, Random, 95% CI)	1.45 [0.36, 5.91]
2 Medium term (6 weeks) treatment failure. Only >=80% follow‐up included.	3	685	Risk Ratio (M‐H, Random, 95% CI)	0.50 [0.10, 2.54]
2.1 Terbinafine vs Clotrimazole	3	685	Risk Ratio (M‐H, Random, 95% CI)	0.50 [0.10, 2.54]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Medium term (6 weeks) treatment failure	9	1003	Risk Ratio (M‐H, Random, 95% CI)	0.63 [0.42, 0.94]
1.1 Naftifine vs Bifonazole	1	97	Risk Ratio (M‐H, Random, 95% CI)	1.96 [0.63, 6.08]
1.2 Naftifine vs Clotrimazole	6	775	Risk Ratio (M‐H, Random, 95% CI)	0.53 [0.34, 0.81]
1.3 Terbinafine vs Bifonazole	1	42	Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]
1.4 Terbinafine vs Clotrimazole	1	89	Risk Ratio (M‐H, Random, 95% CI)	1.07 [0.44, 2.60]
2 Medium term (6 weeks) treatment failure. Only >=80% follow‐up included.	6	854	Risk Ratio (M‐H, Random, 95% CI)	0.55 [0.34, 0.89]
2.1 Naftifine vs Bifonazole	1	97	Risk Ratio (M‐H, Random, 95% CI)	1.96 [0.63, 6.08]
2.2 Naftifine vs Clotrimazole	5	757	Risk Ratio (M‐H, Random, 95% CI)	0.48 [0.30, 0.76]
3 Long term (12 weeks onwards) treatment failure	2	141	Risk Ratio (M‐H, Random, 95% CI)	0.47 [0.22, 1.02]
3.1 Naftifine vs Clotrimazole	1	56	Risk Ratio (M‐H, Random, 95% CI)	0.46 [0.08, 2.56]
3.2 Terbinafine vs Clotrimazole	1	85	Risk Ratio (M‐H, Random, 95% CI)	0.47 [0.20, 1.13]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Treatments versus placebo, treatment failure	4		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
1.1 Ciclopiroxolamine versus placebo. Outcomes at 48 weeks	2	460	Risk Ratio (M‐H, Random, 95% CI)	0.32 [0.20, 0.52]
1.2 Fungoid Tincture versus Placebo. Outcomes at 12 months	1	20	Risk Ratio (M‐H, Random, 95% CI)	0.17 [0.02, 1.14]
1.3 Butenafine 2% & tea tree oil vs placebo. Outcomes at 36 weeks	1	60	Risk Ratio (M‐H, Random, 95% CI)	0.03 [0.00, 0.47]
2 Comparisons between active treatments, treatment failure	1		Risk Ratio (M‐H, Random, 95% CI)	Totals not selected
2.1 Amorolfine 5% + methylene versus Amorolfine 5% + ethanol outcomes at 3 days	1		Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]

Methods	RCT  Study aim clearly defined: yes  Blinded assessor of primary outcome: yes  Double blind study: not stated  Participants comparable at baseline for age: not stated   Sex: not stated Duration of complaint: not stated  Inclusion and exclusion criteria specified: not stated  Interventions well described: yes  Assessment of compliance: not stated  Intention to treat analysis: yes  A Priori sample size calculation: not stated  Fungi identified: yes  Distribution of species between groups: yes  Adverse events reported: yes  Number of drop outs stated: 11
Participants	Total evaluable sample size: 90  Inclusion criteria: male  Exclusion criteria: onychomychosis  Treatment setting: Ambulatory care dermatology clinic
Interventions	a. Naftifine cream  Duration: two weeks  Frequency: applied once daily  b. Terbinafine cream  Duration: two weeks  Frequency: applied once daily  c. Oxiconazole lotion  Duration: two weeks  Frequency: applied once daily
Outcomes	Primary outcome: cure  Secondary Outcomes: Signs and symptoms  ‐ erythema, scaling, fissuring, exudation, pruritus, maceration, vesiculation, burning  Adverse events: none
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Allocation concealment (selection bias)	Low risk	A ‐ Adequate

Methods	RCT. Study aim clearly described: yes Assessor blind to treatment allocation; not stated  Blinding of subjects: not stated  Study groups comparable at baseline for age; yes, Sex: yes, duration; not stated  Were the inclusion/exclusion criteria specified; yes  Were the interventions welldescribed; no  Was there an assessment of compliance; not stated  Was data included from subjects who withdrew after randomisation;no  Was an A priori sample size calculation performed;yes
Participants	Total evaluable sample size: 137  Inclusion criteria: aged 14 +, intertriginious tinea pedis  Exclusion criteria: use of antifungal treatment within seven days (topical) or 6 months (systemic) prior to study
Interventions	a. Placebo (20% ethanol)  b. 25% Tea tree oil  c. 50% Tea tree oil
Outcomes	Primary outcome; cure  Secondary outcomes clinical cure (burnng, itching) graded as absent, mild, moderate, severe, or very severe.
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Allocation concealment (selection bias)	Unclear risk	B ‐ Unclear

Study	Reason for exclusion
Albanese 1992	Looks at preventative effects rather than curative
Alexander 1972	Combined data
Aly 1994	Nine patients with yeast only at baseline
Arenas 1991	Oral versus topical
Athow Frost 1986	Cannot separate mycological results for tinea pedis from those related to fungal infection elsewhere on the body
Avlia 1985	Cannot separate mycological results for tinea pedis from those related to fungal infection elsewhere on the body
Baran 2000	Compares oral versus topical treatments
Barnetson 1998	Oral versus topical treatments
Bjornberg 1986	Cannot separate mycological results for tinea pedis from tinea cruris
Clayton 1982	Cannot separate mycological results for tinea pedis from those related to fungal infection elsewhere on the body
Cuce 1980	Cannot separate mycological results for tinea pedis from those related to fungal infection elsewhere on the body
Cuce 1989	No results for culture are given, only microscopy
Daily 1985	Positive culture and or KOH at baseline
Del Palacio 1989	Cannot separate mycological results for tinea pedis from those related to fungal infection elsewhere on the body
Del Palacio 1991	Cannot separate mycological results for tinea pedis from those related to fungal infection elsewhere on the body
Del Palacio 1992	Cannot separate mycological results for tinea pedis from those related to fungal infection elsewhere on the body
Del Palacio 1999	Combined for candida and dermatophyte infections and sites.
Del Palacio 2000	Combined data for candida and dermatophyte infections
Del Palacio A 1992	Cannot separate mycological results for tinea pedis from those related to fungal infection elsewhere on the body
Duncan 1975	Only half participants had positive culture at entry
Effendy 1987	Cannot separate mycological results for dermatophytes from those related to yeasts and moulds
Evans E 1994	Duplicate study
Fredriksson 1972	Cannot separate mycological results for tinea pedis from those related to fungal infection elsewhere on the body
Fredriksson 1974	Cannot separate mycological results for tinea pedis from those related to fungal infection elsewhere on the body and those infected with candida infection
Fredriksson 1977	Cannot separate mycological results for tinea pedis from those related to fungal infection elsewhere on the body
Fredriksson 1982	Microscopy was not performed at baseline
Friedman 1997	Oral versus topical treatments. Combined data for finger and toe nails
Fulton 1975	Cannot separate mycological results for tinea pedis from those related to fungal infection elsewhere on the body
Galimberti 1984	Looks at preventative effects rather than curative
Gip 1983	Cannot separate mycological results for tinea pedis from those related to fungal infection elsewhere on the body
Greer 1986	Duplicate study
Greer 1987	+ve KOH and or culture
Grigoriu 1983	Cannot separate mycological results for tinea pedis from those related to fungal infection elsewhere on the body
Grooten 1992	Combined data
Kashin 1985	There is no definate end of treatment time from which data can be extracted
Katayama 1987	No culture assessment is made
Kates 1990	Looks at the effects on bacterial infection, not fungal infection
Katz 1972	Cannot separate mycological results for tinea pedis from those related to fungal infection elsewhere on the body
Koca 2001	Experimental agent is an antiperspirant (aluminium hyroxychloride).
Lassus 1983	Cannot separate mycological results for tinea pedis from those related to fungal infection elsewhere on the body
Lassus 1988	Cannot separate mycological results for tinea pedis from those related to fungal infection elsewhere on the body
Lauharanta 1992	Results not clear
Lebwohl 2001	Definitions of population unclear
Lestienne 1982	The written report of the study is too confusing to be able to extract the appropriate data
Li 2001	Combined data for hands and feet infections
Lison 1985	No mycological assessment was carried out
Lynfield 1974	Compares systemic treatment with topical treatment
Maibach 1978	Cannot separate mycological results for tinea pedis from those related to fungal infection elsewhere on the body
McVie 1986	Cannot separate mycological results for tinea pedis from those related to fungal infection elsewhere on the body
Meinicke 1985	Cannot separate mycological results for tinea pedis from those related to fungal infection elsewhere on the body
Mertens 1976	Study looks at bacterial infection as well as fungal infection and does not separate the results
Nolting 1992	Cannot separate mycological results for tinea pedis from those related to fungal infection elsewhere on the body
Nolting 1993	No clear treatment time given  Mycology results are not clear  Only treatment groups one and two are double blind, not group three
Ortiz 1978	Incomplete data for diagnosis at entry
Oyeka 1991	Cannot separate mycological results for tinea pedis of the skin from infection of the nail
Pariser 1994	Detailed numerical reults are not given
Patel 1999	One hundred and four (104/217) only had culture confirmed at baseline. No microscopy for any participant
Privat 1982	No microscopy
Qadripur 1984	Cannot separate mycological results for tinea pedis from those related to fungal infection elsewhere on the body
Reinel 1992	Combined finger and toenail data
Reinel 1992a	Duplicate report of Reinel 1992
Ruping 1993	Combined finger and toenail data
Saple 2001	Data combined for three sites; tinea pedis, tinea cruris, tinea corporis
Shellow 1982	Culture is used to identify pathogens at baseline but not to assess outcome
Smith 1974	Results for mycological cure do not separate out those for tinea pedis from tinea cruris
Smith 1977b	Data extracted for negative culture only
Stettendorf 1983	Cannot separate tinea pedis results from those of tinea corporis and inguilalis
Suschka 2001	Only 56% in Clotrimazole group and 60% in ketoconazole group had + ve test results for dermatophytes.
Tanenbaum 1982	Combined data for dermatophytes and candida
Tanenbaum 1983	Studies effect of treatment on candidiasis, not dermatophytes
Tanuma 2001	Fungi demnstrated by direct microscopic exam OR culture
Terragni 1993	Combined finger and toenail data
Thomas 1976	Unit of analysis sites not patients unclear as to the numbers of units in study
Thomas 1986	Unit of analysis site not patient
Tschen 1979	Only 63/90 had + culture at baseline
Tsuboi 1996	Diagnosis only established by KOH
Weil 1996	Study only looks at culture results and not mycology. The data combines that of dermatophytes and yeasts.
Zaug 1995	Oral versus topical

PERMALINK

Topical treatments for fungal infections of the skin and nails of the foot.

Fay Crawford

Sally Hollis

Abstract

Background

Objectives

Search methods

Selection criteria

Data collection and analysis

Main results

Authors' conclusions

Plain language summary

Background

Description of the condition

Epidemiology

Description of the intervention

Why it is important to do this review

Objectives

Methods

Criteria for considering studies for this review

Types of studies

Types of participants

Types of interventions

Types of outcome measures

Primary outcomes

Secondary outcomes

Search methods for identification of studies

Electronic searches

Other databases

Searching other resources

References from published studies

Handsearching of journals

Unpublished literature

Language

Adverse Effects

Data collection and analysis

Selection of studies

Assessment of risk of bias in included studies

Assessment of methodological quality of included studies

Measures of treatment effect

Assessment of heterogeneity

Data synthesis

Statistical analysis

Sensitivity analysis

Results

Description of studies

Results of the search

(a) Identified trials relating to skin of the foot

Included studies

Comparisons

Excluded studies

(b) Results of the search for trials relating to nails of the foot

Included studies

Excluded studies

Risk of bias in included studies

Allocation

Blinding

Incomplete outcome data

Follow up and exclusions

Selective reporting

Other potential sources of bias

Effects of interventions

(a) Skin trials

(i) Placebo controlled trials

Allylamines Versus Placebo

Short‐term outcome (two weeks)

1.1. Analysis.

Medium‐term outcome (six weeks)

1.2. Analysis.

1.3. Analysis.

1.

Azoles Versus Placebo

Short‐term outcome (two weeks)

2.1. Analysis.

Medium‐term outcome (six weeks)

2.2. Analysis.

Sensitivity analysis

2.3. Analysis.

2.