167 Pathogen-associated mortality following nosocomial bloodstream infections in very low birth weight infants

Friday, March 19, 2010
Grand Hall (Hyatt Regency Atlanta)
Christine Geffers, MD , Charite, Institute for Hygiene, Berlin, Germany
Brar Piening , Charite, Institute for Hygiene, Berlin, Germany
Petra Gastmeier , Charite, Institute for Hygiene, Berlin, Germany
Frank Schwab , Charite, Institute for Hygiene, Berlin, Germany
Background:

In the very low birth infant (VLBW), nosocomial bloodstream infections (BSI) are the most common nosocomial infection and the mortality of infected patients is high. Only little information exists about differences in the mortality of BSI with various organisms.

Objective:

To describe mortality for different pathogens associated to nosocomial primary bloodstream infections within a national network of neonatal intensive care units (NICU) reporting data to the German Surveillance System for Nosocomial Infections in VLBW (NEO-KISS).

Methods:

All patients between 2000 and 2008 of participating neonatal departments, who had a birth weight <1500g and a minimum length of stay of 3 days were included. Cases of nosocomial laboratory confirmed primary BSI were determined using modified CDC definitions (a recognized pathogen or coagulase negative staphylococci (CoNS) in combination with clinical signs are cultured from blood and the organism is not related to an infection at another site). Pathogen-associated mortality rates were calculated as deceased patients per 100 patients with BSI with a specific organism and Cox models adjusting to other variables like birth weight, gestational age, type of delivery etc were used to determine pathogens as risk factors for death after BSI.

Results:

199 NICUs submitted data of 22,477 patients with 871,160 patient-days. 2,429 culture positive BSIs with 2,586 organism occurred in 2,286 VLBW. The BSI incidence was 10.8 (BSI/100 patients). Overall mortality rate for patients with laboratory confirmed BSI was 7%. One third of the deceased patients died within the first 7 days after BSI. Pathogen-related crude mortality rates for the 8 most common pathogens, representing 88% of all pathogens, were shown in table 1. Using CoNS as referent in the regression analysis Enterococci (Hazard ratio (HR) 2.29; 95% confidence interval (CI)= 1.03-5.07), Candida albicans (HR 2.90; 95% CI= 1.53-5.52), E. coli (HR 4.90; 95% CI=2.82-8.51), Enterobacter spp. (HR 5.26; 95% CI=3.15-8.79), Klebsiella spp. (HR 5.35; 95% CI=3.02-9.46) and Serratia spp. (HR 8.54; 95% CI=4.11-17.71) were confirmed as independent risk factor for death after BSI. S. aureus (HR 1.08; 95% CI= 0.46-2.53) shows no significant independent influence to death for patients with BSI when CoNS is the reference.

Table 1 Crude pathogen-associated mortality rates for the most common organism of BSI

Number of BSI with the organism

deceased patients

Mortality rate

CoNS

1,399

58

4.1%

S. aureus

244

9

4.0%

Enterbacter spp.

156

26

16.7%

Enterococci

139

15

10.8%

E. coli

139

19

13.7%

Klebsiella spp.

111

17

15.3%

Candida albicans

80

18

22.5%

Serratia spp.

35

10

28.6%

Conclusions:

Nosocomial BSI with various organisms shows wide differences in mortality rates. Also after adjusting to other variables the type of pathogen of BSI is an independent risk factor for death for VLBW with BSI.