Should all patients with candidaemia have an ophthalmic examination to rule out ocular candidiasis?

Purpose To determine the incidence of ocular candidiasis and length of opht halmic follow-up required to rule out ocular candidiasis in candidaemic pat ients. Methods We prospectively studied patients with candidaemia at our instituti on. Eligibility criteria included a dilated ophthalmological examination wi thin 72 h of fungaemia. Patients without ocular candidiasis on initial exam ination had follow-up dilated ophthalmoscopy performed at 1, 2, 4, 12 and 2 4 weeks. Results Between May 1996 and h larch 1997 a total of 50 patients with funga emia were identified of whom 31 were included in the study; 15 excluded pat ients died before an initial examination was performed. The overall inciden ce of ocular candidiasis was 26% (8/31 patients), all manifested as chorior etinitis. Five patients (16%) had ocular candidiasis on their initial exami nation. One of 21 patients (5%) without ocular candidiasis on initial exami nation developed ocular candidiasis within 1 week. Two of 16 patients (13%) without ocular candidiasis on initial examination or at 1 week follow-up d eveloped ocular candidiasis within 2 Reeks. No evidence of ocular candidias is occurred in the 12 patients with follow-up at 4 weeks, the 8 patients wi th follow-up at 12 weeks and the 4 patients with follow-up at 24 weeks. Conclusion The incidence of ocular candidiasis among hospitalised patients is clinically significant We recommend ophthalmological follow-up for devel opment of ocular candidiasis for at least 2 weeks after an initial negative eye examination.