Candidemia results in a mortality of >50% among adults, but data on ch
ildren with candidemia are limited. We reviewed 70 episodes of pediatr
ic candidemia that occurred between January 1988 and October 1992. Of
these episodes, 53% were caused by Candida albicans, 24% were caused b
y Candida parapsilosis, 16% were caused by Candida tropicalis, and 3%
were caused by Candida krusei. Twenty-five percent of the patients wer
e premature infants. Other underlying conditions included malignancy (
15%); cardiac disease (14%); and short-gut syndrome (14%). A central v
enous catheter was in place during 61 (87%) of 70 episodes. Candiduria
preceded candidemia in only 4 (8%) of 52 patients. The overall mortal
ity rate was 19%; 36% of those with intravenous catheters that were no
t removed within 3 days died, whereas none of the patients from whom c
atheters were removed within 3 days died (P < .0001). Only two survivo
rs had complications. Therapy with amphotericin B (with or without flu
cytosine) was administered to 74% of these patients. Seventeen patient
s were not treated medically; all were immunocompetent and survived. O
f these patients, 15 were >2 months of age; 14 had candidemia for less
than or equal to 2 days; and 15 had an intravenous catheter removed w
ithin 2 days of the onset of candidemia. No patient stopped receiving
amphotericin B because of side effects. The results of this study sugg
est the following: that mortality associated with candidemia is lower
among children than among adults; that failure to remove the indwellin
g intravenous catheter usually results in a poor outcome; that candidu
ria rarely precedes candidemia in children; and that amphotericin B is
well tolerated by children.