Introduction

Enterobacter cloaceae (E. cloacae), a saprophytic microorganism of the normal digestive flora in humans, is one of the most frequently isolated clinical species from septic patients’ blood samples (1,2). Outbreaks due to exogenous E. cloacae infection have been reported in various clinical settings including neonatal intensive care units (NICUs). E. cloacae can be responsible for urinary tract infection, bloodstream infection and pneumonia in hospitalized neonates (3-5).

An important source of infection is contaminated parenteral solutions (6-10). The contamination of parenteral nutrition solution may occur in several ways including the use of components contaminated during the manufacturing process, preparation, storage and administration of the solution (6,11-13).

The aim of this paper is to address the clinical features of a sepsis outbreak in an NICU resulting from the infusion of contaminated parenteral nutrition solution.

Materials and Methods

Tepecik Training and Research Hospital, Neonatology Clinic is a tertiary NICU and consists of 35 incubators, 15 baby cots and 16 for mechanical ventilation. Standard infection control measures are implemented in the care of all patients. Parenteral nutrition solutions are prepared in the parenteral nutrition clean room with the use of aseptic techniques under laminar flow by a single staff nurse and stored in the refrigerator on +4 0C.

On the day of the outbreak, 45 neonates were hospitalized in the clinic and 20 patients received total parenteral nutrition. Thirteen of those 20 high-risk neonates in the NICU developed signs and symptoms of septic shock after the initiation of parenteral nutrition solutions. No other patients had received this solution on that day. During an 18-hour period, 13 patients died despite prompt therapy with vancomycin and meropenem. Time of death were noted in the patients’ records. Blood and parenteral nutrition solutions were cultured from all newborns. DNA analysis was also performed using gel electrophoresis to identify the source. This study was approved by Local Ethic Committee.

Results

The laboratory examination revealed thrombocytopenia and elevated C-reactive protein in the 13 septic newborns that died in the outbreak. Clinical data are shown in Table I. Four of the 13 newborns were female. Gestational age ranged from 26 to 33 weeks (median, 28 weeks), birth weight ranged from 860 to 1.450 g (median, 1.100 g), and postnatal age ranged from 2 to 24 days (median, 9 days).

E. cloacae was identified in the blood cultures of 5 patients and in 11 samples of the parenteral solutions. The identification was positive in the amino acid and dextrose components of the solutions but not the lipid component. The isolates were sensitive to aztreonam, cefuroxime, ceftazidime, ceftriaxone, cefepime, piperacillin-tazobactam, amikacin, gentamicin, tobramycin, trimethroprim-sulfamethoxazole, ciprofloxacin, levofloxacin, imipenem, meropene but resistant to ampicillin, nitrofurantoin, ampicillin-sulbactam and cefazoline. DNA analysis by pulsed-field gel electrophoresis revealed the same profile for the isolates of all E. cloacae (Figure 1). Swabs taken from laminar flow hoods and from the surfaces of compounding equipment were negative for E. cloacae. The hand swabs from the staff during the outbreak and from the nurse who prepared the parenteral nutrition solutions on the following day were also negative. Postmortem examinations confirmed the diagnosis of E. cloacae sepsis in all 13 patients. Table II shows that the patients were critically ill suffering from antenatal problems and complications of preterm birth and prematurity.

Discussion

Enterobacter species have been recognized as increasingly frequent causes of nosocomial infections (14-16). Because a similar organism with the same pattern of antibiotic sensitivity was isolated, it is important to identify the source of contamination.

Some outbreaks caused by intrinsic contamination of the infusion solutions have been reported (9,17). Such contamination may be detected in several hospitals at the same time when a common compound is used in the parenteral nutrition solutions resulting in an intrinsic contamination from the manufacturer. It is unlikely to conclude a similar contamination route in the present outbreak.

The contamination of parenteral nutrition solution may occur in several ways including the use of components contaminated during the manufacturing process, inadequate aseptic technique during the preparation of the solution, sterilization failures, contamination of the multidose lipid emulsion or dextrose solution, contamination of the solution during storage, technical problems during the administration of the solution or ascending method of infusions (6,11-13).

An outbreak of clinical sepsis in a newborn nursery in Brazil was associated with contaminated parenteral medications. Resulting investigation concluded that the locally produced IV solutions were the source of the contamination (18). Similarly the source of the outbreak in our clinic was the parental nutritional solution contaminated by E. cloacae. The results of investigations strongly suggested that the nosocomial sepsis and the molecular typing method were helpful in clarifying the genomic evidence of the E. cloacae strains and confirming the common source of the contamination.

The parental nutrition solutions were prepared in a local setting, therefore contamination during transportation can be eliminated. We think that solutions were likely to be contaminated during preparation, because according to the electronic records of the preparation process, the parental nutrition had been administrated consecutively in these 13 patients after changing the vials of fluids. A hand swab from the staff nurse could not have been taken on site because her shift was over before the outbreak. Therefore the evidence of nosocomial contamination can not be confirmed.

Conclusion

Although the exact source of contamination could not be identified, we think that the source of the sepsis outbreak in our clinic is more likely to be the E. cloacae contaminated parental nutrition solution. The lesson taken from our experience is to establish an action plan in every NICU for the systematic and strategic approach on managing the risk and crisis of a sepsis outbreak.

Ethics

Ethics Committee Approval: It was taken, Informed Consent: It was taken.

Peer-review: External and Internal peer-reviewed.

Authorship Contributions

Concept: Şükran Köse, Design: Şükran Köse, Esra Özer, Data Collection or Processing: Gülgün Akkoçlu, Halide Tokgöz, Neval Ağuş, Analysis or Interpretation: Zeynep Gülay, Recep Öztürk, Önder Ergönül, Writing: Esra Özer.

Conflict of Interest: No conflict of interest was declared by the authors.

Financial Disclosure: The authors declared that this study received no financial support.