Management of an outbreak of Enterobacter cloacae in a neonatal unit using simple preventive measures

Journal of Hospital Infection (2002) 51: 21±26 doi:10.1053/jhin.2002.1186, available online at http://www.idealibrary.com on

Management of an outbreak of Enterobacter cloacae in a neonatal unit using simple preventive measures Y. v. Dijk*, E. M. Biky, S. Hochstenbach-Vernooijy, G. J. v. d. Vlistz P. H. M. Savelkoulx, J. A. Kaank and R. J. A. Dieperslootk *Department of Infection Control, Diakonessen Hospital, Utrecht, yDepartment of Microbiology and Immunology, St. Antonius Hospital, Nieuwegein, zDepartment of Paediatrics, Diakonessen Hospital, Utrecht, xDepartment of Microbiology and Infection Prevention, VU Medical Center, Amsterdam, and kDepartment of Microbiology and Immunology, Diakonessen Hospital, Utrecht, The Netherlands Summary: Enterobacter cloacae is becoming an increasingly important nosocomial pathogen. Outbreaks of E. cloacae in intensive care units and burns units have been described frequently. In December 1999, a neonate with line sepsis was transferred from a university hospital to the neonatal unit of the Diakonessen Hospital. Blood culture yielded E. cloacae. An outbreak of E. cloacae was occurring in the university hospital at that time. In February 2000, a second neonate in our hospital developed line sepsis caused by E. cloacae. Direct measures taken included cohorting of infected children, disinfection of incubators, thermometers and wards, and screening patients. Of nine neonates, seven were colonized with E. cloacae. Despite these measures, the outbreak continued. Forty-one patients were screened; 15 were colonized. Environmental searches yielded E. cloacae in a sink and on two thermometers. Sixteen isolates were typed by arbitrarily primed PCR using four primers. All the patient isolates and the two isolates from thermometers were identical. The strain isolated from the sink was unrelated. Amplified fragment length polymorphism typing showed that the outbreak clone was identical to that in the university hospital. After the introduction of disposable thermometer covers, E. cloacae colonization slowly decreased. & 2002 The Hospital Infection Society

Keywords: Enterobacter cloacae; neonatal unit; thermometer; transmission; genotyping.

Introduction Enterobacter species, particularly Enterobacter cloacae, are commensals of the gastrointestinal tract in healthy adults, and rarely cause primary human disease. However, E. cloacae is becoming an increasingly important nosocomial pathogen.1 It

Received 20 June 2001; revised manuscript accepted 15 January 2002. Author for correspondence: Dr Y. v. Dijk, Diakonessen Hospital Utrecht, Department of Infection Control, Bosboomstraat 1, 3582 KE Utrecht, The Netherlands. E-mail: [email protected]

0195-6701/02/010021 1 06 $35.00/0

is frequently cultured form body substances of hospitalized patients, particularly those treated with antibiotics.2 Colonization has been associated with blood, wound, respiratory and urinary tract infections.3 It has been suggested that colonization and infection may arise endogenously and that cross-infection is relatively uncommon.1 Nevertheless, outbreaks of infections have been reported in neonatal intensive care units (NICUs),4±6 surgical wards7 and burns units,8 mediated by factors such as contaminated pharmaceutical products,9,10 crosscontamination by healthcare workers' hands,11 and medical equipment.12 Outbreaks often continue for & 2002 The Hospital Infection Society

22

Y. v. Dijk et al.

months and require radical measures, such as ward closure or changing antibiotic policy.13 Here we report an outbreak of cross-colonization and infection by E. cloacae with resistance to third-generation cephalosporins, among patients in a neonatal care unit, and describe the ability of relatively simple infection control procedures to stop the outbreak.

bacterial isolates were identified to species level using standard methods, and tested for susceptibility to amoxicillin, amoxillin/clavulanate, gentamicin, amikacin, cotrimoxazole, ceftazidime, ceftriaxone, ciprofloxacin and imipenem. All the isolates were stored at ÿ70 C, and were tested for the presence of extended-spectrum b-lactamase using the E-test system (AB-Biodisk, Solna, Sweden).

Patients and methods

Epidemiological typing of bacterial isolates

Characteristics of the ward The outbreak occurred in the Diakonessen Hospital in Utrecht, The Netherlands, a teaching hospital of 370 beds. The neonatal unit is a 12-bed unit divided into a large room and one smaller room, and has 250 admissions per year. Premature neonates <30 gestational weeks or with severe organ dysfunction are transferred to the intensive care unit of a university hospital for life-support treatment and are returned to our unit when stabilized. Nurses assigned to the unit work in all the rooms. Nurses, staff and visitors wear gowns whenever children are taken out of the incubator. Nurses and (para-) medical personnel are instructed to disinfect their hands prior to each contact. The children are not routinely screened for bacterial colonization. First-line therapy for suspected sepsis is amoxicillin in combination with gentamicin. Case definition A case was defined as a child admitted to the neonatal unit, who had E. cloacae identified in an isolate from any site. Criteria for infection were as defined by the Centers for Disease Control.14 Isolates from specimens not associated with infection were classified as colonization. Bacteriological cultures During the outbreak, specimens for surveillance culture were obtained once a week from the nasopharynx and rectum from all children in the neonatal unit. Samples were also obtained from environmental sites, including sinks, hand basins, bedrails, ventilators, thermometers, oil tissues, solutions, and incubators. Specimens were cultured on 5% sheep blood agar and Colombia CNA agar (bioMeÂrieux, Marcy-l'Etoile, France) and incubated at 37 C. The

To examine the relatedness between different E. cloacae isolates, two molecular typing methods were used. First, 13 patient isolates and three environmental strains (two from thermometers and one from a hand basin) were typed using arbitrarily primed PCR (AP-PCR). Chromosomal DNA was extracted from bacterial cultures on sheep agar plates using the QIAamp DNA mini kit (QIAgen, Westburg, The Netherlands). AP-PCR was performed in 25 mL reaction volumes using primers ERIC1, ERIC2, RW3A and AP1.15±17 Each reaction mixture contained 10 mM Tris-HCl (pH 9.0), 50 mM KCl, 2.5 mM MgCl, 0.1% Triton X-100, 0.01% (w/v) gelatin, 200 mM of each of the four deoxynucleotide triphosphates, 40 pmol primer, 0.75 U of SuperTaq polymerase (HT Biotechnology Ltd., Cambridge, UK) and 50 ng of chromosomal DNA. A Perkin Elmer PE9600 thermocycler was used for amplification. Cycling conditions consisted of the following steps: 5 min at 95 C, 35 cycles of 1 min at 95 C, 1 min at 25 C, and 2 min at 72 C, followed by a 7-min extension at 72 C. PCR products were separated by electrophoresis on a 1% agarose gel, stained with ethidium bromide and visualized by UV light. Secondly, isolates from different hospitals were compared by amplified fragment length polymorphism (AFLP), using restriction enzymes EcoR1 and Mse1 as described previously.18 Control strains from unrelated patients were included in both typing techniques. Infection control measures At the start of the outbreak (February 2000), the infection control measures on the unit were reviewed and personnel were urged to be careful in their use, and incubators and materials were disinfected (hypochlorite 300 ppm or 70% ethyl alcohol). All infected and colonized children were cohorted. Cohorting of the personnel was impossible because of shortage of nurses. Gloves were worn to handle

E. cloacae in neonatal unit

23

colonized children. Contaminated objects and surfaces were routinely disinfected with hypochlorite 300 ppm or 70% ethyl alcohol. A three-day observational study was conducted to monitor the infection prevention practice of healthcare workers and parents. The results of the interventions were discussed during weekly meetings with representatives of the neonatal care unit and the Department of Infection Prevention to watch over the progress.

Measures In our weekly meetings with the neonatal unit, we discussed our first interventions. All sinks and hand basins were disinfected daily with hypochlorite (300 ppm). Disinfection of thermometers changed from a quick rinse with 70% ethyl alcohol to 10 min immersion in 70% ethyl alcohol. Other measures included cohorting of the colonized and infected children, disinfection of the incubator beds and wards, reinforcement of handwashing practice, and weekly bacterial screening of all patients. Observation of the healthcare workers' handwashing practice indicated good compliance. The parents turned out to regularly take new universal oil tissues with contaminated hands. Although no micro-organisms were isolated from tissues, it was decided to give each child their own tissues. During the observation, no other possible route of transmission was noticed. Despite these preventive measures, the outbreak continued. A total of 41 patients were screened, of whom 15 were colonized with E. cloacae (Figure 1). No other child developed an infection. Despite the extended disinfection time, Gramnegative micro-organisms were still isolated from the thermometers. Therefore, all thermometers were replaced and disposable covers were introduced. After this and the introduction of individual oil tissues, the number of new cases of E. cloacae slowly decreased. One month and six months after the outbreak, all neonates were screened again. In July 2000, no E. cloacae was found. In October 2000, E. cloacae

Results Description of the outbreak

Patient number

In December 1999, a neonate (patient 1) was transferred from the NICU of a university hospital to our neonatal care unit. The neonate developed lineassociated sepsis the following day. Blood culture yielded E. cloacae (Figure 1). In February 2000, a second neonate (patient 2) developed (lineassociated) sepsis caused by E. cloacae (Figure 1). At the time, an outbreak of E. cloacae involving 32 patients was occurring at the NICU of the university hospital from which the index child was transferred. We immediately screened all neonates by specimen collection from the nasopharynx and rectum. Of the nine children screened, seven were colonized with E. cloacae. All those who had been in the neonatal unit for at least two days were colonized. Environmental cultures yielded E. cloacae from a hand basin and two thermometers. No other environmental samples were found to harbour the organism. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Dec 1999

Jan 2000

Feb

Mar

Apr

May

Jun

Jul Oct

Month Figure 1 Patients colonized with E. cloacae strains (

) admission on the neonatal unit; ( l ) culture with E. cloacae. bfd fl

Nov

24

Y. v. Dijk et al.

was isolated from the rectum of one neonate (patient 16).

Sensitivity and typing All the E. cloacae strains isolated during the outbreak displayed similar antibiograms. All were resistant to amoxicillin, amoxicillin/clavunalate, ceftazidime and ceftriaxone. An extended-spectrum b-lactamase was demonstrated in all isolates. Sixteen E. cloacae strains isolated from patients and environmental sites in the Diakonessen Hospital were typed genetically using AP-PCR with four different primers. The result from the ERIC1 PCR fingerprinting is shown in Figure 2. Four control strains (lanes 1±4) could all be

Control strains

distinguished from each other, confirming the resolution of this technique. All E. cloacae strains isolated from patients 2±15 from February to April 2000, and the two isolates from thermometers showed identical fingerprints with each of the primers, confirming the suspected cluster, and linking it to the thermometers. In contrast, the isolate from the hand basin and the strain from patient 16 which was isolated in October 2000, six months after the outbreak, both displayed unique fingerprint patterns, indicated that their genotypes were unrelated (Figure 2, lanes 19 and 20, respectively). AFLP typing showed that the E. cloacae outbreak strain was identical to the outbreak strain at the university hospital from which the index child was transferred (Figure 3).

Patients

Patient Thermometers Sink 16

M C C C C 2 4 5 6 7 8 9 10 11 12 13 15 T T S 16 N C

1000 bp 500 bp

Figure 2 ERIC1 PCR-fingerprinting of E. cloacae strains. 60

70

80

90

100 Outbreak strain university hospital

Outbreak strains Diakonessenhuis

Control strain 1 Control strain 2

Figure 3 AFLP typing of E. cloacae strains.

E. cloacae in neonatal unit

Discussion In this study, we describe an outbreak of E. cloacae colonization in a neonatal unit. Newborns are known to be at risk for colonization of Enterobacter spp.19 While neonates initially have a sterile gastrointestinal tract, bacterial colonization occurs rapidly in hospital, the result of selective antibiotic pressure and nosocomial spread through inadequate hygiene.20 Interestingly, the NICU that transferred the index case to our hospital used a third-generation cephalosporin, ceftriaxone, as routine therapy for neonatal sepsis. In some hospitals, children become rapidly colonized. In a multicentre study, Fryklund et al. found a colonization rate in various neonatal care units of 0±78%.2 Intestinal colonization appears to be a precursor of the widespread colonization and the invasive infection which occurs in only a small proportion of colonized children.21 A case record study of Acoled et al. showed an association between Enterobacter septicaemia and short gestation, low birthweight, longer period of endotracheal intubation, and length of hospital stay.22 This might explain why only two of the children in our study developed an E. cloacae infection. Both (patients 1 and 2; Figure 1) were premature (32 weeks 4 days and 30 weeks 6/7 days) and had low birthweights, 1065 g and 1140 g, respectively. Before the outbreak, isolation of E. cloacae bacteria on our neonatal ward occurred rarely. The high increase in E. cloacae colonized children during the first months of 2000 raised the suspicion of an outbreak. Indeed molecular fingerprinting using AP-PRC and AFLP showed that all the E. cloacae strains from the 15 colonized patients were genetically identical (Figures 2 and 3). Immediately after the discovery of a large number of colonized children, the factors that may have contributed to this outbreak were investigated. Previous reports documented that transmission of E. cloacae from patient to patient was attributed to hospital personnel, or use of contaminated medical devices.5,12 In our hospital, cross-transmission by hospital personnel was unlikely, since there was a strict handwashing routine. Despite an intensive environmental search, E. cloacae was only found in a hand basin and on two thermometers. Of these, the thermometers were colonized with the outbreak strain, as shown by DNA fingerprinting (Figure 2). This strongly suggested that thermometers were the routes of transmission. Our findings correspond to

25

those from another Dutch hospital in which the involvement of thermometers in an E. cloacae outbreak in a NICU is described.23 Outbreaks involving other micro-organisms related to thermometers have also been published.24 In neonatal units, the body temperature is often measured by rectal thermometers. In contrast with other units, disposable covers were not used, since the edges of the standard covers may harm the anal sphincter of neonates. Instead, before the outbreak, every child received a personal thermometer. After use, the thermometer was wiped with gauze soaked in 70% ethyl alcohol and stored in a personal plastic basket. After discharge of the child, the same disinfection procedure was followed. For twins, frequently only one thermometer was used. The isolation of E. cloacae from two thermometers shows that disinfection was inadequate. Van den Berg et al. also showed that after disinfection with 80% ethyl alcohol, one out of 10 thermometers remained contaminated with E. cloacae.23 Even after prolonged incubation of thermometers in ethyl alcohol, our outbreak continued, indicating that such disinfection is inadequate. Special disposable thermometer covers for neonates without sharp edges are available (Uni-Instrumenten, The Netherlands). When we introduced these on the unit, contamination of the thermometers and the outbreak stopped. The genetic unrelatedness of the strain from patient 16 (isolated six months after outbreak) to the outbreak clone indicates that after April 2000, no further patient-to-patient transmission occurred, and that the use of disposable thermometer covers and the other measures were finally vindicated. Most likely, the E. cloacae found in this patient was coincidental. The E. cloacae outbreak strain was probably introduced into the neonatal care unit of our hospital by the transfer of a colonized child. AFLP typing showed that the E. cloacae outbreak clone was closely related to that of the NICU of the university hospital from which the index case was transferred (Figure 3). To our knowledge, this is the first report on hospital-to-hospital transmission of an E. cloacae strain. The fact that this caused outbreaks in at least two neonatal units suggests that it possesses certain specific virulence factors, which enables it to cause epidemic spread. Not much is known about epidemic E. cloacae, and this finding deserves more investigation. Probably the difference in their predilection

26

for survival in the human or in the environment plays an important role.20,21 In conclusion, an outbreak caused by E. cloacae in a neonatal unit was recognized, and confirmed by molecular fingerprinting. Infected thermometers were found to be the most important vector. When an outbreak of this kind occurs, stopping further admission is an accepted intervention.23 Our results show that a strict regime of simple infection control procedures in combination with regular feedback can be a worthwhile alternative.

References 1. Sanders WE Jr, Sanders CC. Enterobacter spp: pathogens poised to flourish at the turn of the century. Clin Microbiol Rev 1997; 10: 220±241. 2. Fryklund B, Tullus K, Burman LG. Epidemiology of enteric bacteria in neonatal unitsÐinfluence of procedures and patient variables. J Hosp Infect 1991; 18: 15±21. 3. Jarvis WR, Martone WJ. Predominant pathogens in hospital infections. J Antimicrob Chemother 1992; 29(Suppl. A): 19±34. 4. Peters SM, Bryan J, Cole MF. Enterobacterial repetitive intergenic consensus polymerase chain reaction typing of isolates of Enterobacter cloacae from an outbreak of infection in a neonatal intensive care unit. Am J Infect Control 2000; 28: 123±129. 5. Verweij PE, Belkum van A, Melchers WJG et al. Interrepeat fingerprinting of third-generation cephalosporin-resistant Enterobacter cloacae isolated during an outbreak in a neonatal intensive care unit. Infect Control Hosp Epidemiol 1995; 16: 25±29. 6. Nierop van WH, Duse AG, Stewart RG et al. Molecular epidemiology of an outbreak of Enterobacter cloacae in the neonatal intensive care unit of a provincial hospital in Gauteng, South Africa. J Clin Microbiol 1998; 36: 3085±3087. 7. Andersen BM, Sorlie D, Hotvedt R et al. Multiply beta-lactam resistant Enterobacter cloacae infections linked to the environmental flora in a unit for cardiothoracic and vascular surgery. Scand J Infect Dis 1989; 21: 181±191. 8. Markowitz SM, Smith SM, Williams DS. Retrospective analysis of plasmid patterns in a study of burn unit outbreaks of infection due to Enterobacter cloacae. J Infect Dis 1983; 148: 18±23. 9. Archibald LK, Ramos M, Arduino MJ et al. Enterobacter cloacae and Pseudomonas aeruginosa polymicrobial bloodstream infections traced to extrinsic contamination of dextrose multidose vial. J Pediatr 1998; 133: 640±644.

Y. v. Dijk et al. 10. Tresoldi AT, Padoveze MC, Trabasso P et al. Enterobacter cloacae sepsis outbreak in a newborn unit caused by contaminated total parenteral nutrition solution. Am J Infect Control 2000; 28: 258±261. 11. Harbarth S, Sudre P, Dharan S et al. Outbreak of Enterobacter cloacae related to understaffing, overcrowding, and poor hygiene practices. Infect Control Hosp Epidemiol 1999; 20: 598±603. 12. Thomas A, Lalitha MK, Jesudason MV et al. Transducer related Enterobacter cloacae sepsis in postoperative cardiothoracic patients. J Hosp Infect 1993; 25: 211±214. 13. FinnstroÈm O, Isaksson B, Haeggman S et al. Control of an outbreak of highly beta-lactam-resistant Enterobacter cloacae strain in a neonatal special unit. Acta Paediatr 1998; 87: 1070±1074. 14. Garner JS, Javis WR, Emori TC et al. CDC definitions for nosocomial infections. Am J Infect Control 1998; 16: 128±140. 15. DelVecchio VG, Petroziello JM, Gress MJ et al. Molecular genotyping of methicillin-resistant Staphylococcus aureus via fluorophore-enhanced repetitive-sequence PCR. J Clin Microbiol 1995; 33: 2141±2144. 16. Versalovic J, Koeuth T, Lupski JR. Distribution of repetitive DNA sequences in eubacteria and application to fingerprinting of bacterial genomes. Nucl Acid Res 1991; 19: 6823±6831. 17. Belkum van A, Kluytmans J, Leeuwen van W et al. Multicenter evaluation of arbitrary primed PCR for typing of Staphylococcus aureus strains. J Clin Microbiol 1995; 33: 1537±1547. 18. Savelkoul PHM, Aarts HJM, de Haas J et al. Amplified-fragment length polymorphism analysis: the state of an art. J Clin Microbiol 1999; 37: 3083±3091. 19. Goldmann DA. The bacterial flora of neonates in intensive care-monitoring and manipulation. J Hosp Infect 1988; 11(Suppl. A): 340±351. 20. Acolet D, Ahmet Z, Houang E et al. Enterobacter cloacae in a neonatal intensive care unit: account of an outbreak and its relationship to use of third generation cephalosporins. J Hosp Infect 1994; 28: 273±286. 21. Tullus K, Berglund B, Fryklund B et al. Epidemiology of fecal strains of the family Enterobacteriaceae in 22 neonatal wards and influence of antibiotic policy. J Clin Microbiol 1988; 26: 1166±1170. 22. Fok TF, Hon Lee C, Wong EMC et al. Risk factors for Enterobacter septicemia in a neonatal unit: casecontrol study. Clin Infect Dis 1998; 27: 1204±1209. 23. Berg van den RWA, Claahsen HL, Niessen M et al. Enterobacter cloacae outbreak in the NICU related to disinfected thermometers. J Hosp Infect 2000; 45: 29±43. 24. Broek van den PJ, Verbakel-Salomons EMA, Franssen A et al. Thermometers as vehicle of Klebsiella pneumoniae producting extended spectrum betalactamase. Infect Control Hosp Epidemiol 2000; 21: 134.