Predicting the occurrence of antibiotic-induced candidal vaginitis (AICV)

Fam Pract Res J. 1991 Sep;11(3):319-26.


This paper reports the sensitivity, specificity, and predictive values of symptoms in the diagnosis of antibiotic-induced candidal vaginitis (AICV) among 74 women recruited from three primary care practices. All subjects, who were examined both pre- and post-antibiotic treatment for acute respiratory, urinary tract, or skin infections, were initially free of vaginitis. Twenty-four subjects developed candidal vaginitis (CV), indicated by vaginitis symptoms or signs and a positive candidal culture or KOH preparation; there were no mixed infections. Fifty women did not develop AICV and, of this group, four developed a nonyeast vaginitis. Aggregate symptoms (pruritus and/or discharge) had 87.5% sensitivity, 95.8% specificity, and positive and negative predictive values of 91.3% and 93.9%, respectively. These values are much higher than those reported in studies of CV that excluded women on antibiotics. We conclude that women who develop vaginitis symptoms while on short courses of antibiotics may be treated as AICV without confirmatory examination.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adult
  • Anti-Bacterial Agents / adverse effects*
  • Candidiasis, Vulvovaginal / chemically induced*
  • Erythromycin / pharmacology
  • Female
  • Humans
  • Incidence
  • Physical Examination
  • Vaginitis / chemically induced*


  • Anti-Bacterial Agents
  • Erythromycin