Eh. Kossoff et al., CANDIDEMIA IN A NEONATAL INTENSIVE-CARE UNIT - TRENDS DURING 15 YEARSAND CLINICAL-FEATURES OF 111 CASES, The Pediatric infectious disease journal, 17(6), 1998, pp. 504-508
Objective. To determine changes in the incidence of candidemia in a ne
onatal intensive care unit (NICU) during a 15-year period (1981 to 199
5) and to compare the prevalence and case fatality rates of Candida al
bicans and Candida parapsilosis infections. Methods. A retrospective s
tudy was conducted of candidemia occurring in infants in a NICU betwee
n January 1, 1981, and December 31, 1995. Cases were identified throug
h computerized searching of a microbiology blood culture database. Can
didemia was considered contributory to mortality if death occurred wit
hin 3 days of positive blood cultures or if there was autopsy evidence
of disseminated candidiasis. Results. One hundred eleven cases of can
didemia occurred in 107 infants, representing 1% of all NICU patients
during the study period. The rate of candidemia in the NICU increased
from 2.5 cases per 1000 admissions in 1981 to 1985, to 4.6 per 1000 ad
missions in 1986 to 1990 and to 28.5 per 1000 in 1991 to 1995 (P = 0.0
01). C. albicans was the predominant cause of candidemia between 1981
and 1990, C. parapsilosis was the most prevalent species between 1991
and 1995, causing 53 of 89 cases (60%), The mortality from C. albicans
, 13 of 50 cases (26%), was significantly higher than the mortality fr
om C. parapsilosis, 2 of 54 (4%) (P = 0.002; relative risk, 7; 95% con
fidence interval, 1.7 to 30). Conclusions. The rate of candidemia in o
ur neonatal intensive care unit increased >11-fold in the 15 years fro
m 1981 to 1995; the prevalent Candida species shifted from C. albicans
to C. parapsilosis; and candidemia associated with C. albicans has si
gnificantly higher mortality than with C. parapsilosis.