PURPOSE: Intraocular infection caused by Candida species can have devastating visual consequences. With the emergence of Candida as a major nosocomial pathogen, the authors investigated the prevalence of ocular lesions in patients with candidemia and evaluated risk factors for eye involvement. METHODS: This study is a prospective, multicentered, observational design. One hundred eighteen patients with candidemia were evaluated by the infectious disease service and received indirect ophthalmologic examination within 72 hours of a reported positive blood culture. Ocular findings were classified on the basis of objective, pre-determined criteria. Candida chorioretinitis was defined as the presence of focal, white, infiltrative chorioretinal lesions without vitreal involvement. Candida endophthalmitis was defined as chorioretinitis with extension into the vitreous or intravitreal "fluff balls." RESULTS: In contrast to previous studies of patients with candidemia citing prevalence rates of endophthalmitis approaching 40%, no patients were shown to have endophthalmitis. Candida chorioretinitis was seen in 9% of the patients, all of whom received antifungal agents. The observation that chorioretinitis never progressed to endophthalmitis suggests that systemic antifungal agents provided adequate ocular therapy. Risk factors for Candida chorioretinitis include fungemia with Candida albicans (versus nonalbicans species), multiple positive blood cultures, visual symptoms, and immunosuppression. Twenty percent of patients had nonspecific ocular lesions not directly related to infection. CONCLUSION: Patients with candidemia who have the risk factors noted above warrant formal ophthalmologic examination.
PURPOSE:Intraocular infection caused by Candida species can have devastating visual consequences. With the emergence of Candida as a major nosocomial pathogen, the authors investigated the prevalence of ocular lesions in patients with candidemia and evaluated risk factors for eye involvement. METHODS: This study is a prospective, multicentered, observational design. One hundred eighteen patients with candidemia were evaluated by the infectious disease service and received indirect ophthalmologic examination within 72 hours of a reported positive blood culture. Ocular findings were classified on the basis of objective, pre-determined criteria. Candida chorioretinitis was defined as the presence of focal, white, infiltrative chorioretinal lesions without vitreal involvement. Candida endophthalmitis was defined as chorioretinitis with extension into the vitreous or intravitreal "fluff balls." RESULTS: In contrast to previous studies of patients with candidemia citing prevalence rates of endophthalmitis approaching 40%, no patients were shown to have endophthalmitis. Candida chorioretinitis was seen in 9% of the patients, all of whom received antifungal agents. The observation that chorioretinitis never progressed to endophthalmitis suggests that systemic antifungal agents provided adequate ocular therapy. Risk factors for Candida chorioretinitis include fungemia with Candida albicans (versus nonalbicans species), multiple positive blood cultures, visual symptoms, and immunosuppression. Twenty percent of patients had nonspecific ocular lesions not directly related to infection. CONCLUSION:Patients with candidemia who have the risk factors noted above warrant formal ophthalmologic examination.