Retrospective cohort study to assess if different patterns of Candida colonization determine different risks of progression to invasive fungal infection (IFI) in preterm neonates in NICU. Weekly surveillance cultures from all neonates weighing at birth <1500 g admitted over a 6-year period were reviewed. Infants with available results from at least 3 cultures/week and from at least 4 different sites were enrolled and identified by the number of sites involved [1-2 (low-grade), 3 or more (high-grade)] and type (low-risk, if colonization was recovered from skin, stool, ear canal swab, gastric aspirate, nasopharynx secretions, endotracheal tube; high-risk, from urine, catheter tip, drains, surgical devices). Progression rates from colonization to IFI were calculated for each subgroup. Univariate analysis was performed looking for significant associations between IFI and a number of risk factors, including the different subgroups of colonization. Multiple logistic regression assessed all significantly (P<0.05) associated risk factors. In the 405 eligible infants, overall colonization rate was 42.9%, IFI rate 9.9%, overall progression rate to IFI 0.23, the latter being significantly higher in high-grade or high-risk than in low-grade or low-risk colonized infants (0.59 vs. 0.18, P=0.001; 0.44 vs. 0.11, P<0.001, respectively). Infants with concomitant high-grade + high-risk colonization had 4-fold higher risk of progression than any other colonized infant, and 7-fold higher risk than infants concomitantly low-grade + low-risk colonized (P<0.001). At multivariate analysis, high-grade and high-risk colonization (P=0.001 for both), birth weight (P=0.02) and presence of central venous line (P=0.04) remained independent predictors of IFI. Density and severity of fungal colonization condition the progression to IFI in preterm infants in NICU, and certain patterns of colonization are independent predictors of IFI. Increased culture surveillance and prophylactic measures should be addressed to preterm colonized infants in NICU featuring the most risky colonization patterns.
Read full abstract