Summary

Introduction: Vancomycin-resistance in enterococci (VRE) poses a major threat in hospitals. It is well known that patients in the neonatal intensive care unit (NICU) is especially at risk of nosocomial VRE acquisition. In this study, it was aimed to present the VRE outbreak analysis identified in NICU.

Materials and Methods: Patients with VRE colonization and infection were included in the study between 17.10.2016-30.03.2017 in NICU.

Results: There are a total of 30 beds in the NICU, including five beds in the level 1 ICU, eight beds in the level 2 ICU and 17 beds in the level 3 ICU. Newborns in the NICU are hospitalised in the same ICU (room) in a mixed form. Enterococci were detected in blood culture samples of four patients in the NICU. Three of these cases were VRE and the other was VSE (vancomycin-sensitive enterococci). The most important problem identified in the NICU was the inability to use the interior architecture effectively. In addition, inadequate nurse both in number and level of education and lack of night observations were identified as other problems.

Conclusion: As a result, the outbreak in NICU was controlled thanks to educated nurse support, frequent visits and change in interior architecture.

Introduction

Prolonged hospital stays result in intestinal colonization by VRE, thereby making the patient a VRE reservoir. Patient-topatient transmission of VRE occurs via the hands of healthcare personnel, contaminated surfaces, and contaminated medical instruments. VRE colonization increases the rate of infection 5 to 10 fold[4, 5].

According to 2016 data from the Turkish National Hospital Infections Surveillance Network, the overall VRE weighted general average in Turkey was 13.33% (50th-75th percentile) and the VRE weighted general average in the ICUs of university hospitals was 13.76% (50th-75th percentile)[6].

Due to their immature immunity, neonates are a defenseless population. In newborns, infectious diseases usually occur as a result of contact with healthcare personnel, parents, other patients, medical devices, and the hospital environment. This contact may lead to nosocomial outbreaks as well as colonization or infection with severe morbidity and mortality[7]. The aim of this study was to present an analysis of a VRE outbreak detected in a neonatal intensive care unit (NICU) and the precautions taken.

Methods

The study design was descriptive and retrospective. Approval for the study was obtained from an Local Ethics Committee (date: 02.28.2018, case number: 03).

Perirectal swab samples obtained for surveillance culture were sent to the Medical Microbiology Laboratory. As recommended by EUCAST the samples were inoculated in 5% sheep blood agar and incubated at 35±1 ºC for 24 hours[5]. Colonies with morphologic and Gram-staining characteristics consistent with enterococci were passaged onto Mueller-Hinton agar and a vancomycin disc (5 μg) was applied to the inoculated agar. Isolates with zones of inhibition having sharp borders and a diameter >=12 mm were considered susceptible; for isolates showing indistinct zone borders, colonies growing within the zone of inhibition, and zone diameter <12 mm, vancomycin resistance was confirmed using BD Phoenix 100 (Becton Dickinson Diagnostic Systems, Sparks) automated identification and antibiotic sensitivity system.

By decision of the hospital Infection Control Committee (ICC), rectal swab cultures for VRE were obtained weekly in the NICU, and rectal swab cultures were obtained routinely from all patients upon NICU admission.

Statistical Analysis
The age and gender of the patients included in the study, type of sample showing growth or colonization (blood, rectal swab, etc.), date of NICU admission, diagnosis, date of positive culture, and the antibiotic therapy administered were recorded from the patients" medical records. In addition, physical conditions and operations were analyzed based on the NICU floor plan. The NICU was closely and frequently monitored. The obtained data were recorded. Statistical analysis of the study data was done using SPSS v.17.0 software package (SPSS Inc, Chicago, Illinois, USA). Categorical data were summarized as numbers and percentages.

Results

VRE was identified in the blood culture and rectal swab of one patient on November 3, 2016; this patient was considered to be an index case and was isolated. The following day, VRE was identified in the blood culture and rectal swab of another patient (case 2), and an emergency meeting was held on November 4, 2016 with the ICC and the NICU head physician and nurse. The patients with VRE-positive rectal swab (two infected, six colonized patients), staff, and medical tools and equipment were cohorted. The physical environment was thoroughly cleaned twice using chlorine tablets diluted 1/100 (1000 ppm). However, on November 15, 2016, VRE was identified in venous blood and catheter cultures of a third patient (three infected, 15 colonized patients). The characteristics of the patients are presented in Table 1. The diagnosis of catheter-related bloodstream infection prompted an emergency meeting of the ICC. It was decided to stop new NICU admissions until the outbreak was controlled. During that period, a point-prevalence study was conducted in the NICU to evaluate antibiotic use. The rate of antibiotic use was calculated as 61%. Blood culture antibiogram results of the patients are listed in Table 2.

Table 1: Characteristics of the patients

Table 2: Blood culture antibiotic susceptibility results

One of the important findings was educational disparities among the nurses. An analysis of the nurses" education levels showed that 67.9% (n=19) were high school graduates and 32.1% (n=9)

were university graduates. By ICC recommendation, this ratio was adjusted by increasing the total number of nurses to 34 [17 high school graduates (50%) and 17 university graduates (50%)].

Furthermore, ICC frequently visited and performed observations in the NICU in order to inform all personnel about VRE infection control and protection measures. Neonatal intensive care unit staff compliance with hand hygiene was also investigated and determined to be 60.2% in 2015, 85.6% in 2016, and 89.8% in 2017. Changes implemented in the NICU in line with the decisions are summarized in Table 3.

Table 3: Changes made in the neonatal intensive care unit

Risk factors identified in patients with VRE included use of multiple antibiotics (cephalosporin, penicillin, etc.), urinary or venous catheterization, and admission to the ICU (Table 1).

After implementing these changes and precautions, no new infections were detected but the number of colonized patients increased. In the second outbreak, E. faecalis was isolated in a blood culture from a patient identified as the index case. During this period, we increased the frequency of training and inspections, which resulted in a decline in the number of colonized patients. For the following six months, no new cases were identified. The four cases identified during the outbreak were not patients transferred from other centers. The first case was initially colonized and later infected. Findings of the colonized patients and index cases between January 2016 and April 2017 are presented in Figure 1. One of the main problems identified was the ineffective layout of the NICU. Therefore, some changes were made in the organization of the NICU.

Figure 1: The distribution of vancomycin-resistance in enterococci cases by month

Discussion

Antibiotic resistance in enterococci is an important problem. Resistance to glycopeptide in particular has made the treatment of enterococcus infections difficult[16]. In this analysis of a NICU outbreak, VRE was found in the blood cultures of three patients and VSE was found in one; rectal swab samples also revealed cases of VRE colonization.

Penicillin and aminoglycoside antibiotic groups are also used against enterococci. However, high-level gentamicin resistance (HLGR) and high-level streptomycin resistance (HLSR) must also be considered. In a study by El-Kersh et al.[17], HLGR and HLSR were detected in enterococci at rates of 25% and 11%, respectively. In a study by Ertürk et al.[18], these rates were found to be 25% and 50%, respectively. In yet another study, 67.9% of enterococci showed HLGR and 67.9% showed HLSR[19]. In the present study, three patients exhibited both gentamicin and streptomycin resistance, but one was not analyzed for aminoglycoside sensitivity.

Risk factors for enterococcus-related bacteremia in newborns include extended hospital stays and the associated prolonged use of the 3rd generation cephalosporin and/or vancomycin, surgical interventions, intrahospital or interhospital transfers, enteral nutrition, urinary and central venous catheterization, and necrotizing enterocolitis[20, 21]. Hospital outbreaks occur more frequently in the NICU, pediatric ICU, and hematology/ oncology wards[22]. The present study analysed a VRE outbreak in the NICU.

Enterococci are responsible for approximately 10% of neonatal sepsis and bacteremia, and this figure is steadily increasing. This is thought to result from increased survival in preterm neonates and dissemination within the hospital. For this reason, all patients are screened for VRE at the time of admission. Cases of VRE are reported by the microbiology department to the ICC. The ICC takes measures such as patient monitoring, isolation, cohorting, environmental monitoring, disinfection, sterilization, personal protective equipment, hand hygiene, healthcare personnel training and monitoring, antibiotic changes, inspection of equipment and protective- and carerelated materials, surveillance, and reducing workloads[23, 24].

Similar actions were used in our hospital to control the outbreak and minimize the number of colonized patients.

The hands of hospital workers are important in the dissemination of nosocomial infections, and attention to hand hygiene is therefore crucial in preventing hospital outbreaks[25]. Studies have shown that increased hand hygiene compliance may decrease hospital infections by about 30-50%. Koşucu et al.[26] reported 58% adherence to hand hygiene among health care personnel in their study. In another study conducted in Switzerland, a 3-year monitoring of adherence to hand hygiene resulted in an increase in adherence from 48% to 66% (p<0.001) [27]. The rate of adherence to hand hygiene in our hospital increased between assessments conducted before and after the NICU outbreak (from 60.2% in 2015 to 85.6% in 2016 and 89.8% in 2017). These hand hygiene compliance rates are much higher than the figures reported in the literature. This may be attributed to increased training frequency, conducting bedside practice, and a 100% participation in educational activities.

The recurrence of VRE cases an outbreak despite the high rate of hand hygiene adherence may be related to the temporary lack of hand hygiene monitoring for night shift workers and reduced adherence to hand hygiene and isolation in the absence of observers.

It is important to employ qualified nurses in ICUs to ensure higher quality healthcare service[28]. Needleman et al.[29] determined that better educated and qualified nurses caring for the same number of patients resulted in reduced hospital stays, preventable deaths, adverse events, and costs. In other studies, higher nurse education level was associated with greater job satisfaction and higher morale[30]. Fesci et al.[31] reported that a low nurse-patient ratio resulted in nurses having difficulty fulfilling their duties, even if they are knowledgeable and experienced. Infection rates in our NICU were reduced after improving the nurses" education level and reducing their workload.

Limitations of our study are that the similarity of the four VRE cases was not investigated via molecular methods, colonized patients were not evaluated by antibiotic susceptibility tests, and environmental cultures and personnel colonization were not assessed during the outbreak.

Conclusion

Ethics
Ethics Committee Approval: Approval for the study was obtained from Local Ethics Committee (date: 02.28.2018, case number: 03).

Informed Consent: Since this was a retrospective outbreak analysis informed consent was not received from patients.

Peer-review: Externally and internally peer-reviewed.

Authorship Contributions
Surgical and Medical Practices: S.Y., S.A.,Concept: S.N., S.A., Design: S.N., Ş.İ., Data Collection or Processing: S.N., Ş.İ., S.Y., Analysis or Interpretation: S.N., Literature Search: S.N., A.R.Ş., Writing: S.N., Ş.İ.

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

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