Objectives: To evaluate the long term neurodevelopmental outcome of premature infants exposed to either gram- negative sepsis (GNS) or neonatal Candida sepsis (NCS), and to compare their outcome with premature infants without sepsis.
Methods: Historical cohort study in a population of infants born at <30 weeks gestation and admitted to the Neonatal Intensive Care Unit (NICU) of the Academic Medical Center in Amsterdam during the period 1997-2007. Outcome of infants exposed to GNS or NCS and 120 randomly chosen uncomplicated controls (UC) from the same NICU were compared. Clinical data during hospitalization and neurodevelopmental outcome data (clinical neurological status; Bayley-test results and vision/hearing test results) at the corrected age of 24 months were collected. An association model with sepsis as the central determinant of either good or adverse outcome (death or severe developmental delay) was made, corrected for confounders using multiple logistic regression analysis.
Results: Of 1362 patients, 55 suffered from GNS and 29 suffered from NCS; cumulative incidence 4.2% and 2.2%, respectively. During the follow-up period the mortality rate was 34% for both GNS and NCS and 5% for UC. The adjusted Odds Ratio (OR) [95% CI] for adverse outcome in the GNS group compared to the NCS group was 1.4 [0.4-4.9]. The adjusted ORs [95% CI] for adverse outcome in the GNS and NCS groups compared to the UC group were 4.8 [1.5-15.9] and 3.2 [0.7-14.7], respectively.
Conclusions: We found no statistically significant difference in outcome at the corrected age of 24 months between neonatal GNS and NCS cases. Suffering from either gram-negative or Candida sepsis increased the odds for adverse outcome compared with an uncomplicated neonatal period.