426 - Candida Bloodstream/Meningitis Infections Among Neonatal Intensive Care Unit Patients

Publication Number: 2414.426

Nora Elhaissouni, Johns Hopkins University School of Medicine, Irving, TX, United States; Erica C.. Prochaska, Johns Hopkins University School of Medicine, Baltimore, MD, United States; David Kaufman, University of Virginia School of Medicine, Charlottesville, VA, United States; Elizabeth Colantuoni, Bloomberg School of Public Health, Johns Hopkins University, Baltimore, MD, United States; Dustin D. Flannery, Children's Hospital of Philadelphia, Wynnewood, PA, United States; Sagori Mukhopadhyay, Childrens Hospital of Philadelphia, Philadelphia, PA, United States; Veeral Tolia, Pediatrix, Dallas, TX, United States; Aaron Milstone, Johns Hopkins Children's Center, Baltimore, MD, United States

Background: Candida infections have been a substantial cause of morbidity and mortality in the neonatal intensive care unit (NICU). There is little data on the current incidence and attributable mortality of Candida Bloodstream/Meningitis (BSI/M) infections in infants admitted to NICUs.

Objective: Estimate the incidence and attributable mortality of Candida BSI/M infection among hospitalized infants in the US.

Design/Methods: We conducted a retrospective cohort study and emulated trial using infants admitted to 315 NICUs in the Pediatrix Medical Group between 2016-2021. A Candida BSI/M was defined as a blood/CSF culture that grew any Candida species during hospitalization. For calculating incidence, infants at all postnatal ages were included. Attributable mortality was determined using an emulated trial for infants with Candida BSI/M diagnosed 4-21 days after birth and defined as mortality within 7 days of infection. Infants with Candida BSI/M were matched to those without Candida BSI/M on the same day of life and the following variables: birthweight, gestational age, NICU acuity, presence and intensity of a central line. Emulated trials were conducted in 3-day age bins, and the difference in 7-day mortality between cases and controls in each trial were pooled to calculate the cumulative difference in absolute mortality.

Results: Of 540,024 eligible infants, 288 infants had Candida BSI/M, for an overall incidence of 5.3 (95% CI, 4.74-5.99) per 10,000 infants. Most infants with infection were  < 28 weeks' gestational age (79.9%), very low birthweight (VLBW < 1,500 grams, 87.5%). Incidence for infants < 750g was more than 3 times higher relative to infants born weighing 751-999g (132.8 [95% CI, 113.8-154.0] vs 36.9 [95% CI, 27.6-48.4] infections per 10,000 infants, respectively). 70 of 74 deaths following Candida BSI/M occurred among VLBW infants (90.6%). 72.9% of infants with Candida BSI/M were 4-21 days old at time of infection and were included in the emulated trials. Compared to matched infants without Candida BSI/M, infected infants had a cumulative absolute difference in mortality of 9.9% (95%CI: 5.49, 14.35).

Conclusion(s): This study found VLBW infants are high risk for Candida BSI/M and attributable mortality in the NICU. Most events occurred in VLBW infants and were highly consequential, resulting in an almost 10% increase in absolute mortality within 7 days from infection as compared to infants without Candida BSI/M who were matched on established mortality risk factors. Future studies should focus on infection prevention and control measures to reduce mortality from Candida BSI/M in this vulnerable population.

Age Distribution of Candida BSI/M Infections Among Hospitalized Infants
Infants with a Candida BSI/M infection are represented once for their first infection event. The x-axis is truncated to postnatal age 120, which captures 287 of the 288 infected infants (99.6%).

Emulated Trial for the Estimation of Candida BSI/M Attributable Mortality Among Hospitalized Infants
A, The estimated 7-day all-cause mortality for hospitalized infants anchored by 3-day postnatal age bins (4-21 days) at the time of the first positive culture and the 7-day all-cause mortality for matched hospitalized infants without a Candida BSI/M infection at the same postnatal age. B, The absolute difference in 7-day all-cause mortality between infants with Candida BSI/M infection (cases) and matched infants without Candida BSI/M infection (controls) anchored by 3-day postnatal age bins at the time of the first positive culture. Infants were matched by individual characteristics (ie, postnatal age, birth weight ±100 g, gestational age bins, the presence of a central line within the preceding 48 hours, and, if applicable, central line intensity) and by site-level characteristics measured in percentages (mortality, outborn admissions, very low birth weight admissions). Error bars indicate 95% CIs.

Demographics and Birth Characteristics of Hospitalized Infants Stratified by Candida BSI/M Infection