Secondary Logo

Journal Logo


Risk of resistant infections with Enterobacteriaceae in hospitalized neonates


Author Information
The Pediatric Infectious Disease Journal: November 2002 - Volume 21 - Issue 11 - p 1029-1033
  • Free



Resistance to broad spectrum antimicrobial agents is an emerging problem in intensive care settings worldwide. 1–4 Premature or low birth weight infants in neonatal intensive care units (NICUs) often receive broad spectrum antimicrobial agents for suspected health care-associated infections. 5 The amount of antimicrobial agent administration correlates with the emergence of resistant microorganisms in adult intensive care units. 4 The CDC-National Nosocomial Infection Surveillance System recently reported substantial increases of third generation cephalosporin-resistant enteric Gram-negative bacilli causing nosocomial infections in intensive care units. 1, 6

Antimicrobial-resistant microorganisms frequently are acquired by NICU patients; colonization has been detected as early as on the first day of hospitalization. 7–9 Colonization may follow transmission from other patients in the unit or may evolve endogenously from the selective pressures afforded by antimicrobial therapy. Very early colonization may result from maternal antimicrobial use. Since the 1990s administration of intrapartum antimicrobial agents to high risk women for group B Streptococcus (GBS) prophylaxis has been a standard practice. CDC figures estimate that 18 to 27% of women delivering babies are exposed to antimicrobial agents. 10 Administration of intrapartum antimicrobials for GBS prophylaxis has yielded conflicting results in incidence of non-GBS early onset sepsis. 11, 12 In one study in which the rate of non-GBS sepsis from 1990 to 1992 (period with no GBS prophylaxis) was compared with the rate of non-GBS sepsis from 1993 to 1996 (period with GBS prophylaxis), it showed no increase in incidence of non-GBS early onset sepsis. In the other study the rate of early onset non-GBS sepsis showed no significant increase in non-GBS sepsis rates between 1996 and 1999; however, an increase in proportion of ampicillin resistant Escherichia coli isolates was seen from 1996 to 1998.

Risk factors associated with progression from colonization to infection caused by third generation cephalosporin-nonsusceptible enteric Gram-negative bacteria have not been elucidated for critically ill neonates. Colonized infants are at risk of infections and serve as reservoirs for transmission from patient to patient and from patient to staff via hands of health care workers. The objective of this study was to identify the risk factors for ANE infection in critically ill, ANE-colonized neonates.


Study design

Hospital epidemiology personnel initiated active surveillance of all patients admitted to the NICU at Children’s National Medical Center (CNMC) in 1998 after a cluster of four cases of ANE infection were identified. Surveillance identified additional cases of colonization and infection with health care-associated ANE. The NICU is an outborn 40-bed unit that provides care for critically ill patients within the first year of life. All patients are transported to our NICU because there is no obstetrical service at CNMC. The NICU has 6 bays and each bay houses 6 infants. At the entrance to each bay and within each bay are 2 sinks equipped for chlorhexidine hand scrubbing. There are 4 single bed rooms with negative pressure ventilation that are reserved for infants requiring airborne or contact isolation, and segregation from other infants is judged to be beneficial. Patient care is delivered by 2 teams consisting of attending physicians, neonatology fellows, pediatric residents and nurse practitioners. Each patient is also assigned to a nursing team. Nursing assignment is based on the severity of illness of the patient. Hospital epidemiology personnel make daily rounds with the clinical team and collect data prospectively on all patients. The data presented here extend for a 3-year period (1998 to 2000). Data were collected on a standardized form in the NICU. Information gathered from each patient included age, gender, birth weight and histories of previous NICU hospitalizations, surgery, necrotizing enterocolitis, administration of parenteral nutrition, premature rupture of membranes, antimicrobial agent administration and placement of invasive devices. Administration of ampicillin, cefotaxime, gentamicin, vancomycin and other commonly prescribed antimicrobial agents during the entire hospitalization was analyzed as an independent variable. Antimicrobial days refer to the total number of days that antimicrobial agents were administered before ANE colonization or infection was detected. Invasive device use was defined as placement of a central venous catheter/umbilical catheter or dependence on mechanical ventilation before surveillance specimen collection. Rectal and/or stool surveillance cultures were performed on patients at the time of admission to the NICU and at weekly intervals thereafter.

Case patients were those infected with ANE. Health care-associated (nosocomial) infections were defined in accordance with the CDC-National Nosocomial Infection Surveillance System criteria. 13 Health care-associated infections, in the context of our study, were localized or systemic infections caused by ANE in patients who had no evidence of infection at time of admission to the NICU. Colonized patients had ANE in surveillance cultures but did not display signs of infection caused by these microorganisms. The purpose of the surveillance cultures was explained to the NICU staff and is a part of standard care. Parents were informed of the need for contact isolation when their child was found to be positive for ANE, and informed consent was not obtained.


Surveillance cultures were performed according to a method developed by the clinical microbiology laboratory at CNMC. Rectal swab specimens were inoculated onto MacConkey agar, and both a ceftazidime disk (30 μg) and a cefpodoxime disk (10 μg) were applied to the streaked area of the culture plates. After overnight incubation at 35°C, cultures exhibiting inhibitory zone diameters <22 mm around either or both disks were selected for further workup. Growth from the inner edge of the zones of inhibition was subcultured and streaked for isolation onto fresh MacConkey agar plates. Pure culture isolates were identified, and their antimicrobial susceptibility patterns were determined after overnight incubation of broth microdilution panels (Dade MicroScan, West Sacramento, CA). The triple disk approximation test using ceftazidime (30 μg), amoxicillin/clavulanate (20/10 μg) and cefpodoxime (10 μg) disks was performed on isolates exhibiting confirmed ceftazidime nonsusceptibility to determine whether they produced ESBL. This test was as sensitive as if not slightly more sensitive than the broth dilution test described in the current NCCLS standard. 14–16 ANE were defined as members of the Enterobacteriaceae family exhibiting nonsusceptibility to ceftazidime (MIC ≥ 16 μg/ml) or evidence of ESBL production.

Statistical analysis

The Kruskal-Wallis test for continuous variables and Fisher’s exact test for nominal and categoric variables were used for univariate comparisons of variables between cases of infection and colonization. 17, 18 A significance level of <0.05 was chosen. Multiple logistic regression analysis was performed on variables that were statistically significant during univariate analysis of colonized patients. Statistical analysis was performed with SAS/STAT software (SAS Institute, Inc., Cary, NC).


During the 3-year surveillance period 1710 patients were admitted to the NICU. Of the 1710 patients 300 (18%) were excluded from the risk factor analysis [138 (8%) were admitted with preexisting rectal colonization with ANE, 46 (3%) had no surveillance cultures performed and 116 (7%) were admitted to the NICU more than once]. Forty-six (33%) of the 138 patients with preexisting ANE colonization weighed <1000 g. The incidence of colonization was 10 to 13 per 1000 patient days for the 3 consecutive years, and there was no significant variation from month to month.

Of the 1410 remaining neonates the incidence of health care-associated ANE colonization was 17% (240 of 1410 patients). Thirty-four (14%) of the colonized patients developed ANE infection. Of the 34 patients 27 had documented colonization on weekly surveillance cultures before being diagnosed with ANE infection. The mean time interval from identification of colonization to infection was 21 days. Colonizing ANE, in order of frequency, were Enterobacter spp. (72%), Citrobacter spp. (13%) , Klebsiella spp. (5%) and others (10%), whereas those causing infections, in order of frequency, were Enterobacter spp. (71%), Klebsiella spp. (14%) Citrobacter spp. (6%) and others (9%). The 34 infected patients experienced 36 health care-associated ANE infections. The lower respiratory tract (n = 13), urinary tract (n = 11), blood stream (n = 9) and other locations (n = 3) were the most common sites of infection.

The 206 patients who did not develop ANE infections yielded growth of 221 ANE isolates; 138 patients yielded at least 2 ANE isolates of the same species on different dates. The elapsed time between collection dates for ANE culture isolates was 2 weeks or less for 80 of the 138 patients (58%). Of the remaining 58 patients who were colonized or infected with the same ANE species for >2 weeks, 18 patients were culture positive for >8 weeks.

Sixty of the 206 colonized patients (29%) and 14 of the 34 infected patients (41%) harbored ESBL-producing isolates. Overall 35% of the ANE were ESBL producers. (Enterobacter aerogenes 73%, Enterobacter cloacae 1%, Klebsiella 91%, Citrobacter 10% and others 29%). Overall 35% of the ANE were also resistant to gentamicin (E. aerogenes 72%, E. cloacae 4%, Klebsiella 64%, Citrobacter 10% and others 26%).

Differentiating characteristics of infected vs. colonized patients were lower mean birth weight (1403 g), older mean age (53 days) and longer mean length of stay (94 days) (Table 1). Univariate analysis revealed that infected patients also received a larger number of antimicrobial agents and a longer course of antimicrobial therapy (mean, 36 days). These neonates had longer mean durations of central venous access (26 days) and longer mean use of mechanical ventilation (20 days), and they were more likely to have had gastrointestinal surgery for management of necrotizing enterocolitis before infection (P < 0.05;Table 2). Many of these risk factors can be markers of prolonged, complicated hospitalizations, and their significance was further elucidated by multiple logistic regression analysis. The latter analysis revealed that hospitalized neonates with very low birth weights (<1000 g), and those who had received prolonged exposures to antimicrobial agents were at greater risk of ANE infection (odds ratio, 2.5 and 1.8, respectively) (Table 3).

Patient characteristics of colonization vs. infection (N = 240)
Univariate analysis of patients with infection or colonization (N = 240)
Multiple logistic regression analysis of patients with infection or colonization (N = 240)*


Our study identified two groups of hospitalized neonates who were at significant risk of ANE infection: (1) those of very low birth weight (< 1000 g); and (2) those who have received prolonged antimicrobial agent therapy. Cumulative exposure to antimicrobial agents was a greater contributing factor than the specific agent(s) used. This finding is consistent with results we reported earlier on nosocomial enterococcal infections. 19

Colonization with resistant Gram-negative bacilli was related to empiric use of amoxicillin and cefotaxime in a European study. 20 That crossover study found that neonates who were given intravenous amoxicillin-cefotaxime therapy were 18 times more likely to harbor resistant Gram-negative bacilli than neonates who were given penicillin and tobramycin. Gentamicin and cefotaxime are often used in the NICU setting in combination with ampicillin or vancomycin. The combination of amoxicillin and cefotaxime may exert selective pressure encouraging emergence of ESBL-producing bacteria. Cefotaxime may also select for derepressed amp C mutants among the colonizing Enterobacteriaceae. The concerns about the appearance of ESBL-producing ANE are 3-fold: (1) such isolates are clinically resistant to all penicillins, cephalosporins and monobactam agents, making therapeutic intervention difficult; (2) detection of such isolates is problematic for the clinical microbiology laboratory. Special procedures, such as the triple disk approximation test or the methods described by the National Committee on Clinical Laboratory Standards must be used by the laboratory to recognize ESBL producers 14, 16; (3) the spectrum of Enterobacteriaceae capable of ESBL production is broadening. In particular E. cloacae, E aerogenes, Serratia spp. and Citrobacter spp. have been found capable of ESBL production. 20–22

Very low birth weight has been recognized as a factor that predicts neonatal mortality with good discrimination (area under the curve, 0.869). 23 As documented by our study such patients are at heightened risk of colonization and infection by ANE (area under the curve, 0.737). We have also shown that neonates frequently arrive already colonized with ANE when transferred from community hospitals to Level III NICUs. These patients are smaller, are sicker and are often transferred to undergo operative procedures that are not available at the referring hospitals. Monitoring ANE colonization by weekly surveillance cultures has proved beneficial in our setting.

Preterm and low birth weight neonates are more likely to have been exposed to antimicrobial agents for preterm-premature rupture of membranes (P-PROM). Consensus has not yet been reached on recommended antimicrobial agent regimens or on the length of antimicrobial treatment for women with P-PROM. The role of antimicrobial agent therapy for P-PROM was examined during a randomized clinical trial conducted by the National Institute of Child Health and Human Development Maternal-Fetal Medicine Units Network. 24 A regimen consisting initially of intravenous ampicillin and erythromycin for 48 h followed by oral amoxicillin and erythromycin for 7 days was shown to prolong pregnancy and reduce infant morbidity. Most treatment regimens target a wide range of organisms including aerobic/anaerobic bacteria and Mycoplasma. However, they did not look for ANE colonization and infection of neonates during this clinical trial.

Strategies implemented to prevent colonization and infection by antimicrobial-resistant bacteria and control their spread have met with limited success. 25–27 During our 3-year study, the incidence of health care-associated ANE colonization and infection remained unchanged despite the initiation of control measures that included barrier precautions for colonized and infected patients. The compliance with hand washing and the use of barrier precautions were not assessed during our study. Others have shown that health care professionals routinely wash their hands less often and for shorter durations than recommended. 28

In Europe a program designed to promote hand hygiene by providing disinfectant solutions at the bedside and active surveillance led to a reduction in the incidence of nosocomial infections from 17% to 10% and reduced rates of methicillin-resistant Staphylococcus aureus transmission. 29 Control of vancomycin-resistant enterococci (VRE) spread in a health care setting was achieved with use of standard interventions like active surveillance and barrier precautions. 30 In this study the prevalence of gastrointestinal colonization with VRE was reduced from 2.2% to 0.5% during a 4-year period. Environmental contamination was an important factor in the spread of VRE during this investigation.

Restricting use of cephalosporins along with provision of educational programs reduced colonization and health care-associated infections with ANE in a Brazilian neonatal unit from 32% to 11%. 31 When an endemic problem with antimicrobial resistance exists, such as in our setting, a combination approach of antimicrobial agent restriction and focused use of barrier precautions is needed. 6, 32

NICU patients infected or colonized with antimicrobial resistant microorganisms are reservoirs for nosocomial spread. Because there is heavy reliance on broad spectrum antimicrobial agents and proximity of incontinent patients, spread of resistant microorganisms among patients and/or staff occurs readily. Infected or colonized patients admitted to NICUs should have contact isolation to reduce nosocomial spread. The duration of ANE colonization was prolonged in some of our study patients. Previous recognition of colonized patients is helpful in making therapeutic decisions when patients develop signs of systemic infection. The information derived from this study should facilitate recognition of hospitalized neonates who are at greatest risk of progressing from ANE colonization to infection.

We thank the NICU staff and Lucy Couto of the Department of the Neonatology for assisting with surveillance study and implementation of control measures.


1. Fridkin SK, Gaynes RP. Antimicrobial resistances in intensive care units. Clin Chest Med 1999; 20: 303–16.
2. Hanberger H, Garcia-Rodriguez JA, Gobernado M, et al. Antimicrobial agent susceptibility among aerobic Gram-negative bacilli in intensive care units in 5 European countries: French and Portuguese ICU study groups. JAMA 1999; 281: 67–71.
3. Archibald L, Philips L, Monnet D, et al. Antimicrobial resistance in isolates from inpatients and outpatients in the United States: increasing importance of the intensive care units. Clin Infect Dis 1997; 24: 211–15.
4. Fridkin SK, Steward CD, Edwards JR, et al. Surveillance of antimicrobial use and antimicrobial resistance in United States hospitals: Project ICARE phases 2. Clin Infect Dis 1999; 29: 245–52.
5. Gaynes RP, Edwards JR, Jarvis WR, et al. Nosocomial infections in high-risk nurseries in the United States. Pediatrics 1996; 98: 357–61.
6. Siegel JD, Levine GL, Herchline TE, et al. Prevalence (PRVE) of antimicrobial resistant bacteria (AMRB) in Pediatric Prevention network (PPN) intensive care units [Abstract 311]. Presented at the 39th Annual Meeting of the Infectious Diseases Society of America, October 27 to 28, 2001.
7. Almuneef MA, Baltimore RS, Farrel PA, et al. Molecular typing demonstrating transmission of Gram-negative rods in a neonatal intensive care unit in the absence of recognized epidemic. Clin Infect Dis 2001; 32: 220–7.
8. Toltzis P, Dul MJ, Hoyen C, et al. Molecular epidemiology of antibiotic-resistant Gram-negative bacilli in a neonatal intensive care unit during a nonoutbreak period. Pediatrics 2001; 108: 1143–8.
9. Venezia RA, Scarano FJ, Preston KE, et al. Molecular epidemiology of an SHV-5 extended-spectrum beta lactamase in Enterobacteriaceae isolated from infants in a neonatal intensive care unit. Clin Infect Dis 1995: 915–23.
10. Centers for Disease Control and Prevention. Prevention of perinatal group B streptococcal disease: a public health perspective. MMWR 1996; 45: 1–24.
11. Chen KT, Tumoala RE, Cohen AP, Eichenwald EC, Liberman E., et al. No increase in rates of early-onset neonatal sepsis by non-group B Streptococcus or ampicillin-resistant organisms. Am J Obstet Gynecol 2001: 854–8.
12. Baltimore RS, Huie SM, Meek JI, et al. Early-onset neonatal sepsis in the era of group B streptococcal prevention. Pediatrics 2001; 108: 1094–8.
13. Garner JS, Jarvis WR, Emori TG, Horan, et al. CDC definitions for nosocomial infections (1988). Am J Infect Control 1988; 16: 128–40.
14. Coudron PE, Moland ES, Sanders CC. Occurrence and detection of extended-spectrum beta-lactamases in members of the family Enterobacteriaceae at a veteran’s medical center: seek and you may find. J Clin Microbiol 1997; 35: 2593–7.
15. National Committee for Clinical Laboratory Standards. Performance standards for antimicrobial disk susceptibility tests: approved standard. 7th ed. NCCLS Document M2\NA7. Wayne, PA: National Committee for Clinical Laboratory Standards, 2000.
16. National Committee for Clinical Laboratory Standards. Performance standards for antimicrobial susceptibility testing. 12th informational supplement. NCCLS Document M100–S12. Wayne, PA: National Committee for Clinical Laboratory Standards, 2002.
17. SAS Institute, Inc. SAS/STAT software. Released 6.12. Cary, NC: SAS Institute, Inc., 1997.
18. SAS Institute, Inc., SAS/STAT user’s guide: logistic regression. Version 6. 4th ed. vol 2. Cary, NC: SAS Institute, Inc., 1990: 1071–126.
19. Singh-Naz N, Rakowsky A, Cantwell E, et al. Nosocomial enterococcal infections in children. J Infect 2000; 40: 145–9.
20. de Man P, Verhoeven BAN, Verbrugh HA, et al. An antimicrobial agent policy to prevent emergence of resistant bacilli. Lancet 2000; 355: 973–8.
21. Sanders CC, Barry AL, Washington JA, et al. Detection of extended-spectrum beta-lactamase-producing members of the family Enterobacteriaceae with the Vitek ESBL test. J Clin Microbiol 1996; 34: 2997–3001.
22. Tzelepi E, Giakkoupi P, Sofianou D, et al. Detection of extended-spectrum beta-lactamases in clinical isolates of Enterobacter cloacae and Enterobacter aerogenes. J Clin Microbiol 2000; 38: 542–6.
23. Pollack MM, Koch MA, Bartel DA, et al. A comparison of neonatal mortality risk prediction models in very low birth weight infants. Pediatrics 2000; 105: 1051–7.
24. Mercer BM, Miodovnik M, Thurnau GR, Goldenberg RL. Antimicrobial agent therapy for reduction of infant morbidity after preterm premature rupture of membranes: a randomized controlled trial. JAMA 1997; 278: 989–95.
25. Jarvis WR. Preventing the emergence of multidrug-resistant microorganisms through antimicrobial use controls: the complexity of the problem: from the Fourth International Conference on the Prevention of Infection. Infect Control Hosp Epidemiol 1996; 17: 490–5.
26. Goldmann DA, Weinstein RA, Wenzel RP, et al. Strategies to prevent and control the emergence and spread of antimicrobial-resistant microorganisms in hospitals. JAMA 1996; 275: 234–40.
27. Shlaes DM, Gerding DN, John JF Jr, et al. Society for Healthcare Epidemiology of America and Infectious Disease Society of America Joint Committee on the Prevention of Antimicrobial Resistance: guidelines for the prevention of antimicrobial resistance in hospitals. Infect Control Hosp Epidemiol 1997; 18: 275–91.
28. Handwashing Liaison Group. Handwashing (editorial). BMJ 1999; 318: 686.
29. Pittet D, Hugonnet S, Harbarth S, et al. Effectiveness of a hospital-wide programme to improve compliance with hand hygiene: Infection Control Programme. Lancet 2000; 356: 1290–1.
30. Ostrowsky BE, Trick WE, Sohn AH, et al. Control of vancomycin-resistant enterococcus in health care facilities in a region. N Engl J Med 2001; 344: 1427–33.
31. Calil R, Marba S, Nowakonski A, et al. Reduction in colonization and nosocomial infection by multiresistant bacteria in a neonatal unit after institution of educational measures and restriction in the use of cephalosporins. Infect Control Hosp Epidemiol 2001; 29: 133–8.
32. Schiff GD, Wisniewski M, Bult J, et al. Improving inpatient antibiotic prescribing: Insight from participation in a national collaborative. Jt Comm J Qual Improv 2001; 27: 387–402.

Antimicrobial resistance; nosocomial infections; neonates; critical care; infection control

© 2002 Lippincott Williams & Wilkins, Inc.