A parallel-group double-blind study was carried out which compared the efficacy of chemical avulsion of affected nail by urea 40% and bifonazole 1% cream alone with that of the same local therapy combined with short-term oral griseofulvin in onychomycosis. A total of 120 patients were included in the study. Patients' characteristics were comparable in both treatment groups. Of the 98 patients fully evaluated, 91 had toenail involvement and only seven had fingernail involvement. Forty-six of the patients were men and 51 were women. The mean age of the patients was 47.14 +/- 13.84 years (range 17-80 years). The duration of onychomycosis was for more than 1 year in 96 patients and for 3 months duration in only one patient, who was in the placebo group. Forty patients had received different previous therapies. All topical treatments were discontinued for at least 2 weeks and oral therapy for at least 2 months prior to the beginning of the study. The diagnosis was confirmed by positive mycologic cultures. Trychophyton rubrum was identified as the pathogen in 90 patients, 45 in each group, T. tonsurans in four patients, two in each group, and T. mentagrophytes in three patients, two in the griseofulvin treated group, and one in the placebo group. The first phase of treatment given to all patients consisted of occlusive dressing every 24 h with urea 40% and bifonazole 1% ointment until the infected nail became completely detached. Subsequently, in the second phase bifonazole 1% cream was applied to the nail ped every 24 h for 4 weeks. In addition, concomitantly with the bifonazole cream the patients were randomly allocated to a daily oral double-blind treatment with griseofulvin 500 mg or placebo, for 4 weeks. Clinical and mycologic evaluations were carried out at baseline, immediately after removal of the nail, and at 3 days, 4 weeks, and 4 months after the end of treatment with bifonazole cream and griseofulvin/placebo tablets. Mycologic examination included identification of fungi by KOH preparation and culture on potato dextrose agar. Positive cultures were transfered for identification on Sabouraud's. Criteria for evaluation of efficacy comprised: "cure" defined as clinical and mycologic cure (fresh specimen and culture negative) at both investigation times after the end of treatment; "late cure" defined as mycologic cure at both investigation times after the end of treatment, clinical clearing of the nail only 4 months after the end of treatment; "improvement" defined as mycologic cure and only partial clinical improvement at both times after the end of treatment; "failure" indicating no mycologic cure (fresh specimen and/or culture positive); and "relapse" signifying a change from negative findings 1 month after the end of treatment to positive findings 4 months after the end of treatment. Adverse reactions were evaluated on each visit. Only those patients who had completed clinical and mycologic evaluation during the entire study were included in the final statistical analysis. Those patients with partial evaluation were included only in the evaluation of adverse events. Based on the assumptions of a failure rate (failure and relapse) of 30% with bifonazole cream alone and of 10% with bifonazole cream and griseofulvin tables, a = 0.05 and b = 0.2 the required sample size was at least 58 patients for each treatment group (Casagrande formula, one-sided test). The primary efficacy variable "assessment of treatment" (cure and improvement versus failure and relapse) was tested for treatment differences by Fisher's exact test (a = 0.05, one-sided test; Ho, no advantage with additional systemic therapy of griseofulvin). Additionally, the relapse rates of both treatments were tested exploratively in the same way as the primary efficacy variable. All other data were analyzed descriptively.