Journal of Hospital Infection (2005) 61, 213–218
An outbreak of Pantoea spp. in a neonatal intensive care unit secondary to contaminated parenteral nutrition H. Habsaha,*, M. Zeehaidaa, H. Van Rostenbergheb, R. Noraidab, W.I. Wan Pauzib, I. Fatimahb, A.R. Roslizaa, N.Y. Nik Sharimahb, H. Maimunahc a
Department of Medical Microbiology and Parasitology, Kubang Kerian, Kelantan, Malaysia Department of Paediatric, PPSP, USM, Malaysia c Hospital Infection Control and Epidemiology Unit, HUSM, Malaysia b
Received 13 October 2004; accepted 4 January 2005 Available online 20 April 2005
KEYWORDS Outbreaks; Neonatal intensive care; Parenteral nutrition; Pantoea spp.; Contamination
Summary Contaminated parenteral nutrition (PN) is an important source of infection in neonates. Many organisms have been reported to cause contamination that results in outbreaks in intensive care units. The objective of this study was to investigate an outbreak caused by Pantoea spp., which contaminates PN, in a neonatal intensive care unit (NICU). This was a descriptive study of an outbreak of sepsis in an NICU of a tertiary teaching hospital in Malaysia. Pantoea spp. infection was detected in eight patients over a three-day period from 24 to 27 January 2004 following the administration of PN. Seven of the eight patients died due to the infection. Extensive environmental samplings for culture were performed. PN solution from the NICU and the pharmacy were also cultured during the outbreak period. Pantoea spp. was isolated from blood cultures of all infected patients, and the unused PN from the pharmacy and the NICU. All the strains of Pantoea spp. had a similar antibiotic susceptibility pattern and biochemical reaction. From the results, we concluded that PN was the source of the outbreak and the contamination may have occurred during its preparation in the pharmacy. A thorough investigation has been carried out and, where possible, corrective measures have been taken to avoid similar outbreaks in the future. Q 2005 The Hospital Infection Society. Published by Elsevier Ltd. All rights reserved.
* Corresponding author. Tel.: +60 97664607; fax: +60 97652678. E-mail address: [email protected]
0195-6701/$ - see front matter Q 2005 The Hospital Infection Society. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.jhin.2005.01.004
Introduction Neonates are at risk of acquiring infections during their stay in neonatal intensive care units (NICUs) for various reasons, including an immature host immune system, frequent invasive diagnostic procedures, and intravenous administration of medication, parenteral nutrition (PN) or other fluids. Contaminated PN and intravenous fluid have been reported to cause sepsis outbreaks.1–3 The contamination may have resulted from errors during compounding of PN in the pharmacy or during handling of the solutions in the ward. Staphylococcus epidermidis, Staphylococcus aureus, Staphylococcus saprophyticus, Enterobacter cloacae, Klebsiella oxytoca, Serratia marcescens, Acinetobacter calcoaceticus, Stenotrophomonas maltophilia, Pseudomonas aeruginosa, Burkholderia cepacia, Flavobacterium spp. and Candida albicans were found to be the species most likely to contaminate PN during preparation or administration. These micro-organisms have been implicated in more than 95% of all outbreaks and sporadic cases of nosocomial bloodstream infections related to contaminated parenteral admixtures.4 To our knowledge, Pantoea spp. has not been implicated previously in similar outbreaks. In the hospital where this outbreak occurred, the neonatal PN consisted of three standard solutions: solution A consisting of 10% dextrose, calcium and water-soluble vitamins; solution B consisting of amino acids, electrolytes and trace elements; and solution C consisting of fatty acid emulsion and lipid-soluble vitamins. Solution A was prepared in batches and kept for a maximum of one week. Solutions B and C were prepared on a daily basis in the pharmacy.
H. Habsah et al. patterns. All affected babies were on PN. None of the enterally fed neonates had signs of septicaemia during that period. In this outbreak, six of the eight babies had been born prematurely and two were term, as shown in Table I. All the affected infants except for one term neonate died. For the purpose of investigating this outbreak, a case was defined as an infant warded in the NICU during the outbreak, with clinical signs of sepsis or a positive blood culture for Pantoea spp. between 24 and 29 January 2004. The blood cultures from all infected babies grew Pantoea spp. The initial identification by conventional biochemical reaction showed that the micro-organisms belonged to the family Enterobacteriacae. The analytical profile index (API) showed no reaction with lysine decarboxylase, arginine dihydrolase and ornithine decarboxylase, pointing towards P. agglomerans.5 However, the full profile of API only gave 40.7% identification of P. agglomerans. The amylase was positive and the gelatinase was negative, which is against P. agglomerans. The DNase was negative which excludes the possibility of Serratia. With these profiles, the micro-organism was identified as Pantoea spp. alone. The isolates were sensitive to amikacin, cefoperazone, ceftazidime, cefuroxime, ciprofloxacin, co-trimoxazole, gentamicin, imipenem, netilmicin, piperacillin, piperacilin/tazobactam, cefepime and cefoperazone/sulbactam, but resistant to ampicillin. In all cases, the isolates showed similar biochemical reactions and antibiotic susceptibility patterns, indicating that they were phenotypically the same.
Preparation and sterility testing of PN Description of the outbreak The start of the outbreak, 24 January 2004, coincided with the last of four consecutive days of public holiday in Malaysia. The outbreak occurred in the NICU of a tertiary teaching hospital in Malaysia. The NICU comprised of one unit with 30 beds; 10 beds equipped with ventilators and 20 beds for intermediate care. Low-care neonates were nursed in a separate 30-bed special care nursery. On the first day of the outbreak, six babies with signs and symptoms of sepsis had their blood taken for culture. Two days later, two more babies developed sepsis and were investigated accordingly. All the blood cultures taken grew Pantoea spp. with a similar biochemical reaction and antibiogram
In this hospital, solution A was prepared as a large batch that was kept for a maximum duration of seven days. Upon request from the NICU, the pharmacist would freshly prepare an admixture of solutions A and B that was used within 24 h. Two samples containing 2 mL of solution A and admixture of both solutions A and B were sent to the microbiology laboratory for sterility testing upon preparation of solution A. However, repeated withdrawals of solution A for subsequent usage made it liable to contamination. Contamination may occur during storage or repeated handling during PN preparation. The stock of solution A was not subjected to any more sterility testing until it was discarded at the end of the week.
Pantoea spp. outbreak in an NICU secondary to contaminated PN Table I Case
Characteristics of the eight infants with sepsis caused by Pantoea spp Age at onset of sepsis (days)
1 2 3
11 5 5
4 5 6 7 8
5 4 4 3 180
Gestational age (weeks) 26 26 36 Term 36 Term 32 33
Diagnosis on admission
RDS RDS IUGR, polycythaemia, feeding intolerance VACTERL anomaly Perinatal asphyxia Perinatal asphyxia Premature with RDS Down’s syndrome with CLD and bronchopneumonia
Date of PN commencement
Date of positive blood cultures
14.1.04 21.1.04 20.1.04
24.1.04 (am) 25.1.04 (am) 25.1.04 (am)
Died 25.1.04 Died 30.1.04 Died 27.1.04
21.1.04 21.1.04 21.1.04 22.1.04 24.1.04
25.1.04 25.1.04 24.1.04 25.1.04 24.1.04
Died 26.1.04 Died 29.1.04 Survived Died 26.1.04 Died 30.1.04
(am) (am) (pm) (am) (pm)
RDS, respiratory distress syndrome; IUGR, intra-uterine growth restriction; VACTERL, vertebrae anomaly, imperforated anus, cardiac anomaly, tracheoesophageal fistula, renal anomaly, limb anomaly; CLD, chronic lung disease; PN, parenteral nutrition.
Procedure for sterility testing of PN in the microbiology laboratory The PN solution was processed according to the existing procedure. One millilitre of PN solution was inoculated into trypticase soy broth (TSB) and brain heart infusion (BHI), respectively. They were incubated at 36 8C for 24 h before being cultured on blood agar. The plates were incubated further at 36 8C for 24 h. Any growth noted from the samples was identified and notified to the pharmacy personnel urgently. A negative result was to be reported after 48 h of incubation.
Outbreak investigation The routine sterility testing on solution A prepared on 19 January 2004 showed no growth. This batch of solution A was used for that week. Sterility testing of the unused PN solution from the NICU sent on 24 January 2004 grew Pantoea spp. Sterility testing of the new batch of solution A prepared on 26 January 2004 also grew Pantoea spp. The micro-organism isolated from solution A from the pharmacy was phenotypically similar to the micro-organism isolated from unused PN solution from the NICU. This evidence strongly suggests that contamination of PN solution may occur during PN preparation in the pharmacy. In the microbiology laboratory, a repeat sterility test was done on the contaminated solution A from the pharmacy but this was negative. We requested the whole bag of contaminated solution A for a repeat culture using a different method. In the revised method, the bag was incubated for 24 h to allow sufficient time for bacterial multiplication
and growth before inoculating into TSB and BHI. It was then further incubated for 48 h and cultured on MacConkey and blood agar. We gave an instruction for the pharmacy to freshly prepare all PN, and to cease the batching practice for solution A. Aliquots of all compounding solutions such as dextrose, sodium chloride, distilled water, potassium chloride, magnesium sulphate, potassium sulphate, pediatrace calcium gluconate and multi-vitamins were also sent for culture. No growth was detected from any of the samples tested. Environmental sampling was performed in the NICU and the pharmacy. In the pharmacy, the in-use disinfectant (70% alcohol, 70% solution spirit), swabs taken from laminar flow, and re-usable items were cultured. In the NICU, culture was performed on samples from the in-use disinfectant which consisted of desmanolw, viracleanw, povidone iodine and 70% surgical spirit solution. Samples were also taken from ventilator cuffs and tubes. All environmental samples were negative for Pantoea spp. Culture from the hands of the pharmacy and NICU staff, and rectal swabs of all babies present in the NICU during the outbreak period were negative for Pantoea spp.
Intervention of the outbreak The intervention was divided into three phases: immediate, intermediate and long term. Since six babies developed signs of septicaemia at the same time, a common source such as PN was suspected. The administration of PN was discontinued immediately. The usage of PN was stopped temporarily for
216 all neonates. All staff involved in the care of babies, pharmacy staff and the infection control team were alerted to the outbreak. The affected neonates were cohorted, and dedicated instruments were assigned to each neonate. The importance of strict handwashing practice was re-emphasized among staff. Based on the rapid deterioration of the neonates, Gram-negative sepsis was suspected and, therefore, the babies were put on treatment with intravenous imipenem. No other babies in the NICU were infected or colonized by Pantoea spp. after three days of the outbreak’s onset. The intermediate intervention included calling an urgent meeting to alert and investigate the possible source of the outbreak, which involved the paediatricians, the infection control team and the hospital pharmacists. The preparation process of the PN solution was investigated and, where possible, optimized, making further breaches of sterility unlikely. The practice of preparing a sufficient quantity of solution A for one week was stopped. PN solution was freshly prepared each day for immediate use to minimize contamination that may occur during frequent handling of solution A. Sterility testing was also done on a daily basis. There was a modification in the processing of sterility testing in the microbiology laboratory during the outbreak. A larger volume of PN solution was requested from the pharmacy for sterility testing. The samples were incubated for 24 h at 36 8C before inoculation into TSB and BHI, compared with the previous method where 1 mL of aliquot was directly inoculated into TSB and BHI. The first incubation is essential to allow proliferation of micro-organisms to a sufficient number for further inoculation, thus increasing the detection rate. The preparation of PN was evaluated for its microbiological contamination on a daily basis for one week. This procedure was undertaken to ensure sterility of the PN solution. PN administration was recommenced back in the NICU after one week of satisfactory microbiological results. The long-term intervention included the revision of protocols for PN administration, care of the PN lines, and preparation of PN in the pharmacy.
Discussion The NICU has been implicated in many outbreaks which have resulted in a wide range of morbidity and mortality. Sources of outbreaks have included ventilator systems,6 intravenous medication,3 intralipid contamination, 7 infant formula,8
H. Habsah et al. umbilical catheters,9 hands of staff (particularly from artificial nails),10 dextrose saline used for the dilution of an antibiotic,11 hygroscopic bandages,12 and many other sources. The sources were sometimes discovered but were usually not found or were not properly investigated. PN has always been attributed as a potential source of outbreak. Contamination has been shown to occur during its administration because of a suboptimal aseptic procedure during administration and poor care of the PN line. External contamination of the intravenous administration set may also play a role in infusate contamination. A few outbreaks in critical care units worldwide have been tracked to PN contamination.2,4,13 An outbreak of fatal septicaemia caused by Serratia odorifera biotype 1 involved infants at several hospitals and was traced to contaminated PN fluid. The transfusate had been made up in a flexible film isolator system whereby the implicated microorganism was recovered from surfaces inside the isolator, despite routine decontamination procedures.13 An outbreak caused by E. cloacae in an NICU has also been described. E. cloacae was found in the refrigerated aliquots of PN solution, in blood cultures from infected newborns, and from in-use PN solutions.2 This outbreak of Pantoea spp. in an NICU was found to originate during the preparation of PN. The onset of clinical sepsis in this outbreak was within 24 h of PN administration, as reported previously.2 The fatality rate in this outbreak was very high (seven out of eight affected neonates died). Other studies have also revealed a high fatality rate following an outbreak.1,3,14 In previously reported outbreaks, PN contamination occurred during administration of PN in the ward (traced to the hands of the specialized nutrition nurse),15 during preparation in the pharmacy,2,13 or during preparation by the manufacturer, e.g. contamination of the upper surface of capped rubber stoppers from commercial lipid emulsion bottles used for PN.16 Other studies were unable to confirm the source of contamination.17 In most cases, PN was just a risk factor for sepsis.18 This outbreak was curbed within three days. The control of outbreaks has been well reported. The time for controlling an outbreak ranged from one day2 to two months.1,19,20 Our interventions were fast and effective. The approach to this outbreak was multi-disciplinary, involving neonatologists, microbiologists, pharmacists and the infection control unit of the hospital. We recommend this type of approach for future outbreak management. Maximum vigilance and proper planning during
Pantoea spp. outbreak in an NICU secondary to contaminated PN multiple consecutive holidays play a significant role in the prevention of similar outbreaks. The cause of this outbreak was a suboptimal aseptic technique during preparation and storage of PN in the pharmacy. We demonstrated that contamination occurred before the PN was sent to the ward due to the implicated micro-organism in the unused PN. The risk of outbreak was due to the protocol used for preparation and storage of PN solution A for one week in the pharmacy. This protocol had been adopted in our teaching hospital for many years without any adverse events until the long public holiday, which might be the confounding factor. Every step has been taken to prevent recurrence of the outbreak. The protocol for care of the PN line was revised, ensuring that it was practical to perform, especially with limited human resources. In this protocol revision, dedicated nurses were appointed for care of the PN line. They were given a refresher course on aseptic technique and their practices in the ward were observed by the infection control nurse and sister to ensure that they adhered to the protocol. Sterility testing of PN solutions may not prevent outbreaks but may play a role in early detection of contamination and in monitoring adherence to aseptic techniques during preparation. The pathogenicity of Pantoea has not been well studied. It is a ubiquitous member of the Enterobacteriaceae that is found in plants and in the faeces of humans and animals. In the genus of Pantoea, species other than P. agglomerans rarely cause infection in human subjects. P. agglomerans has been reported to cause infections in a few cases of septic arthritis,21–23 synovitis,24,25 ostitis,26 cholelithiasis,27 and occupational respiratory infections and skin allergy in a group of workers exposed to organic dust.28 The majority of cases were associated with occupation. The morbidity caused by this infection in previous reported cases was mild and no mortality was documented. It can cause catheter-related infection as it can form a biofilm on catheters.29 However, there has been no report of neonatal infection or a hospital outbreak caused by this micro-organism.
Conclusion This is the first reported outbreak of Pantoea spp. in an NICU. The outbreak originated from PN contamination. The fatality rate of the outbreak was 87.5%. A multi-disciplinary approach allowed curbing of the outbreak within three days. To prevent
further outbreaks, the preparation protocol for PN should be optimized. An automated compounder can further minimize the chance of future contamination.
Acknowledgement We would like to thank the Director of Universiti Sains Malaysia Hospital, Kubang Kerian, Kelantan, Malaysia for allowing us to send the report for publication.
References 1. Clinical sepsis and death in a newborn nursery associated with contaminated parenteral medications—Brazil, 1996. MMWR Morb Mortal Wkly Rep 1998;47:610—612. 2. 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. 3. Garrett DO, McDonald LC, Wanderley A, et al. An outbreak of neonatal deaths in Brazil associated with contaminated intravenous fluids. J Infect Dis 2002;186:81—86. 4. Didier M, Fischer S, Maki DG. Total nutrient admixtures appear safer than lipid emulsion alone as regards microbial contamination: growth properties of microbial pathogens at room temperature. J Parenter Enteral Nutr 1998;22: 291—296. 5. Sanders WE, Sanders CC. Enterobacter spp.: pathogens poised to flourish at the turn of the century. Clin Microbiol Rev 1997;10:220—241. 6. Garcia DC, Trevisan AR, Botto L, Cervetto M, Sarubbi MA, Zorzopulos J. An outbreak of multiply resistant Pseudomonas aeruginosa in a neonatal unit: plasmid pattern analysis. J Hosp Infect 1989;14:99—105. 7. de Beaufort AJ, Bernards AT, Dijkshoorn L, van Boven CP. Acinetobacter junii causes life-threatening sepsis in preterm infants. Acta Paediatr 1999;88:772—775. 8. Thurm V, Gericke B. Identification of infant food as a vehicle in a nosocomial outbreak of Citrobacter freundii: epidemiological subtyping by allozyme, whole-cell protein and antibiotic resistance. J Appl Bacteriol 1994;76:553—558. 9. Bouallegue O, Mzoughi R, Weill FX, et al. Outbreak of Pseudomonas putida bacteraemia in a neonatal intensive care unit. J Hosp Infect 2004;57:88—91. 10. Gupta A, Della-Latta P, Todd B, et al. Outbreak of extendedspectrum beta-lactamase-producing Klebsiella pneumoniae in a neonatal intensive care unit linked to artificial nails. Infect Control Hosp Epidemiol 2004;25:210—215. 11. Ganeswire R, Thong KL, Puthucheary SD. Nosocomial outbreak of Enterobacter gergoviae bacteraemia in a neonatal intensive care unit. J Hosp Infect 2003;53:292—296. 12. Melamed R, Greenberg D, Porat N, et al. Successful control of an Acinetobacter baumannii outbreak in a neonatal intensive care unit. J Hosp Infect 2003;53:31—38. 13. Frean JA, Arntzen L, Rosekilly I, et al. Investigation of contaminated parenteral nutrition fluids associated with an outbreak of Serratia odorifera septicaemia. J Hosp Infect 1994;27:263—273.
218 14. Reish O, Ashkenazi S, Naor N, Samra Z, Merlob P. An outbreak of multiresistant Klebsiella in a neonatal intensive care unit. J Hosp Infect 1993;25:287—291. 15. Moro ML, Maffei C, Manso E, et al. Nosocomial outbreak of systemic candidosis associated with parenteral nutrition. Infect Control Hosp Epidemiol 1990;11:27—35. 16. Doit C, Loukil C, Simon AM, et al. Outbreak of Burkholderia cepacia bacteremia in a pediatric hospital due to contamination of lipid emulsion stoppers. J Clin Microbiol 2004;42: 2227—2230. 17. Bruun B, Jensen ET, Lundstrom K, Andersen GE. Flavobacterium meningosepticum infection in a neonatal ward. Eur J Clin Microbiol Infect Dis 1989;8:509—514. 18. Moro ML, De Toni A, Stolfi I, Carrieri MP, Braga M, Zunin C. Risk factors for nosocomial sepsis in newborn intensive and immediate care units. Eur J Pediatr 1996;155:315—322. 19. Royle J, Halasz S, Eagles G, et al. Outbreak of extended spectrum B lactamase producing Klebsiella pneumoniae in a neonatal unit. Arch Dis Child Fetal Neonatal Ed 1999;80: F64—F68. 20. Enterobacter sakazakii infections associated with the use of powdered infant formula—Tennessee, 2001. MMWR 2002;51:298—300. 21. Flatauer FE, Khan MA. Septic arthritis caused by Enterobacter agglomerans. Arch Intern Med 1978;138:788. 22. Stromqvist B, Edlund E, Lidgren L. A case of blackthorn synovitis. Acta Orthop Scand 1985;56:342—343.
H. Habsah et al. 23. Kratz A, Greenberg D, Barki Y, Cohen E, Lifshitz M. Pantoea agglomerans as a cause of septic arthritis after palm tree thorn injury; case report and literature review. Arch Dis Child 2003;88:542—544. 24. Oleginski TP, Bush DC, Harrington TM. Plant thorn synovitis: an uncommon cause of monoarthritis. Semin Arthritis Rheum 1991;21:40—46. 25. De Champs C, Le Seaux S, Dubost JJ, Boisgard S, Sauvezie B, Sirot J. Isolation of Pantoea agglomerans in two cases of septic monoarthritis after plant thorn and wood sliver injuries. J Clin Microbiol 2000;38:460—461. 26. Laporte C, Demachy MC, Thevenin-Lemoine C. Tibial osteitis caused by Pantoea agglomerans after open grade IIIB tibial shaft fracture. Rev Chir Orthop Reparatrice Appar Mot 2002; 88:625—627. 27. Flores C, Maguilnik I, Hadlich E, Goldani LZ. Microbiology of choledochal bile in patients with choledocholithiasis admitted to a tertiary hospital. J Gastroenterol Hepatol 2003;18:333—336. 28. Milanowski J, Dutkiewicz J, Potoczna H, Kus L, Urbanowicz B. Allergic alveolitis among agricultural workers in eastern Poland: a study of twenty cases. Ann Agric Environ Med 1998;5:31—43. 29. Murga R, Miller JM, Donlan RM. Biofilm formation by gramnegative bacteria on central venous catheter connectors: effect of conditioning films in a laboratory model. J Clin Microbiol 2001;39:2294—2297.