Epidemiology of vaginal Candida infection: significance of numbers of vaginal yeasts and their biotypes.

The prevalence, quantity and biotypes of vaginal yeasts were determined for 1082 randomly selected nonpregnant patients attending two genitourinary medicine clinics. The overall yeast prevalence was 22.1% with a geometric mean of 40 and a median of 75 yeast colonies per positive isolate (on 5-cm Petri dishes). There was no statistically significant variation in prevalence, quantity or biotype of yeasts with the patients' age, season of the year, stage of the menstrual cycle, recent antibiotic history, contraceptive use or main diagnosis (excluding candidosis). A clinical score for Candida infection, based on symptoms of pruritus and signs of Candida vulvovaginitis, showed significant variation with the prevalence and quantity of yeast isolates. Unequivocal clinical evidence of candidosis was strongly associated with high concentrations of vaginal yeasts. There was also some association between certain groups of Candida albicans biotypes and the clinical score. These observations reemphasize the need for consideration of both clinical and mycological factors in establishing a diagnosis of vulvovaginal Candida infection. It is suggested that isolation of fewer than 10 yeast colonies from a vaginal swab is usually unlikely to indicate an infection requiring treatment.