Late-onset Bloodstream Infections in Hospitalized Infants
Late-onset Bloodstream Infections in Hospitalized Infants
We included in the analysis a total of 206,019 infants cared for between DOL 4 and 120 (Fig. 1). Of 15,480 (8%) infants who had at least 1 culture done, 3% (534) met our definition for late-onset BSI. We identified 552 episodes of late-onset BSI (18 infants had 2 episodes). The overall incidence of late-onset BSI was 2.7 episodes per 1000 NICU admissions. The mean age at late-onset BSI diagnosis was 11 days (5th, 95th percentiles: 8, 34).
(Enlarge Image)
Figure 1.
Study population.
For both infected and noninfected infants, the mean GA was 39 weeks (37–41), and the mean birth weight was 3275 g (2085–4332) and 3313 g (2343–4285), respectively. Both infected and noninfected groups were mostly male, but a higher number of infants in the noninfected group received antenatal antibiotics (24% vs. 16%; Table 1 ). The mean length of stay for infected infants was 20 days (4–48) and for noninfected infants was 8 days (3–17). Most infants (96%) were admitted on or before DOL 3. The mean day of admission was 4 (0–21) for infected and 1 (0–3) for noninfected infants.
Gram-positive organisms were the most frequent pathogens (64%; Table 2 ). Gram-negative organisms represented 26% of the positive cultures, and Escherichia coli (14%) was the most predominant organism in this group. We identified yeast in 5% of infants with late-onset BSI. Although the overall incidence of late-onset BSI decreased slightly over time, the incidence in 1997 was not statistically lower than that in 2010—3.3 vs. 2.5/1000 admissions; P=0.63 (Fig. 2).
(Enlarge Image)
Figure 2.
Incidence of late-onset sepsis BSI by year. GBS, group B Streptococcus.
On multivariable regression, infants with the following characteristics had a reduced risk of late-onset BSI: small for GA, delivered through Cesarean, exposed to antenatal antibiotics and discharged later ( Table 3 ). Black race was associated with higher risk of late-onset BSI.
Overall mortality was 0.3% (529/197,984). Among infants with late-onset BSI, 4% (20/480) died, compared with 0.3% (509/197,504) of noninfected infants. Mortality by organism group was highest for yeast [13.0% (3/23)], followed by Gram-negative organisms [3.9% (5/128)] and Gram-positive organisms [3.8% (12/319)]. The organisms with highest mortality were Pseudomonas sp. [16.7% (1/6)] in the Gram-negative group and group B Streptococcus [6.7% (2/30)] among Gram-positives ( Table 2 ). On multivariable regression, late-onset BSI was associated with increased risk of mortality [odds ratio 8.43 (95% confidence interval: 4.42–16.07)].
Results
We included in the analysis a total of 206,019 infants cared for between DOL 4 and 120 (Fig. 1). Of 15,480 (8%) infants who had at least 1 culture done, 3% (534) met our definition for late-onset BSI. We identified 552 episodes of late-onset BSI (18 infants had 2 episodes). The overall incidence of late-onset BSI was 2.7 episodes per 1000 NICU admissions. The mean age at late-onset BSI diagnosis was 11 days (5th, 95th percentiles: 8, 34).
(Enlarge Image)
Figure 1.
Study population.
For both infected and noninfected infants, the mean GA was 39 weeks (37–41), and the mean birth weight was 3275 g (2085–4332) and 3313 g (2343–4285), respectively. Both infected and noninfected groups were mostly male, but a higher number of infants in the noninfected group received antenatal antibiotics (24% vs. 16%; Table 1 ). The mean length of stay for infected infants was 20 days (4–48) and for noninfected infants was 8 days (3–17). Most infants (96%) were admitted on or before DOL 3. The mean day of admission was 4 (0–21) for infected and 1 (0–3) for noninfected infants.
Gram-positive organisms were the most frequent pathogens (64%; Table 2 ). Gram-negative organisms represented 26% of the positive cultures, and Escherichia coli (14%) was the most predominant organism in this group. We identified yeast in 5% of infants with late-onset BSI. Although the overall incidence of late-onset BSI decreased slightly over time, the incidence in 1997 was not statistically lower than that in 2010—3.3 vs. 2.5/1000 admissions; P=0.63 (Fig. 2).
(Enlarge Image)
Figure 2.
Incidence of late-onset sepsis BSI by year. GBS, group B Streptococcus.
On multivariable regression, infants with the following characteristics had a reduced risk of late-onset BSI: small for GA, delivered through Cesarean, exposed to antenatal antibiotics and discharged later ( Table 3 ). Black race was associated with higher risk of late-onset BSI.
Overall mortality was 0.3% (529/197,984). Among infants with late-onset BSI, 4% (20/480) died, compared with 0.3% (509/197,504) of noninfected infants. Mortality by organism group was highest for yeast [13.0% (3/23)], followed by Gram-negative organisms [3.9% (5/128)] and Gram-positive organisms [3.8% (12/319)]. The organisms with highest mortality were Pseudomonas sp. [16.7% (1/6)] in the Gram-negative group and group B Streptococcus [6.7% (2/30)] among Gram-positives ( Table 2 ). On multivariable regression, late-onset BSI was associated with increased risk of mortality [odds ratio 8.43 (95% confidence interval: 4.42–16.07)].
