The present study is a description and evaluation of two control measures. The first was applied for an index case in a diabetology ward and included a screening based on culture technique. The second used a PCR assay and was applied for a secondary case hospitalized in a Nephrology ward.

We adapted French national recommendations based on the pragmatic following rules: (i) if no new GRE case was identified by initial cross-sectional screening of contact patients, patient transfer to other wards or HCFs were allowed, as well as the admission of new patients. Contact patients transferred to another ward were to be placed in single room with contact precautions in the downstream ward until the result of two subsequent screening tests was available. (ii) If another GRE-positive case was identified by cross-sectional screening, national guidelines were to be followed scrupulously.

Two simultaneous culture techniques were used: (i) One rectal swab inoculated in enrichment broth (AES VRE) incubated 24 hours, and then subcultured onto the chromogenic medium (Oxoid Brillance VRE) incubated aerobically at 37°C and read after 24 and 48 h; (ii) a second swab was directly plated on the same type of medium. Enterococcus faecium were identified using a mass spectrometry assay (MALDI-TOF-MS system). Strains were suspected as GRE in case of minimum inhibitory concentrations (MICs) >8 mg/L for vancomycin and/or teicoplanin using E-test strips (BioRad). Vancomycin-resistance genotypes were identified using a DNA strip assay (GenoType Enterococcus; Hain Lifescience GmbH).

For the molecular diagnosis, Xpert vanA PCR assay was performed by the manufacturer’s instructions (Cepheid) using rectal swabs. Considering the high negative predictive value of the test, the patient was considered at low risk to be colonized if the result was negative for the vanA and vanB genes. Otherwise, conventional culturing was performed to confirm or disprove the presence of GRE. Cultures were also performed in case of amplification inhibitors in the sample.

Time required for each step of the microbiological analysis was collected prospectively. The cost of screening was computed based on the use of material resources needed and personnel costs on the basis of the hourly salary of a senior staff member.

We estimated the costs attributable to the decreased occupancy by multiplying the number of missed patient-days (difference between admission capacity and the number of admitted patients when a GRE-positive patient was identified) by the mean cost billed per hospital day (total amount billed in 2011 based on French Diagnosis-Related Groups divided by the number of patient-days in 2011) 7 .

On 2 February, 2012, a 68-year-old man was admitted to the diabetology ward for the management of complicated type-2 diabetes.

On 27 February, an E. faecium strain highly resistant to vancomycin and teicoplanin was cultured from a wound in the right foot PCR confirmed the presence of the vanA gene. The diabetology ward has 32 beds in 16 double rooms. All 31 patients hospitalized in the ward were considered contact patients of the first case. At this time, the Xpert™ vanA/vanB PCR had been recently introduced in our laboratory and was used on an exceptional basis to screen the two patients who had shared the room of the first case patient, one of whom was PCR-positive. The investigation of this secondary case will be described in the corresponding paragraph.

On 28 February, rectal swabs were obtained from the 31 contact patients of the first case and cultured according to standard techniques. Transfers to other units or HCFs and admissions were stopped pending the results of the rectal-swab cultures. However, 17 of the 31 contact patients were sampled and discharged home over the next two days.

On 29 February, the two colonized patients (the initial case and the secondary case identified from the same room by PCR) and 13 contact patients were cohorted in a separate area of the ward and cared for by dedicated staff to prevent cross transmissions.

As shown in Table 1, the median turnaround times (TAT) for culture techniques was 70.5 hours. Of the 31 screened patients, one was found colonized with GRE corresponding to the secondary case previously identified by PCR (see above). None of the 17 patients discharged home was finally found to be colonized.

PCR, polymerase chain reaction; GRE, Glycopeptide-Resistant Enterococci ; Q1, First quartile; Q3, Third quartile.

^aReal-time PCR assay performed on an exceptional basis to screen the two patients who shared the room of the first patient in the diabetology ward.

^bPositive culture with identification of E. faecium by mass spectrometry assay, MICs >8 mg/L for vancomycin and/or teicoplanin on E-test strips, antibiotic susceptibility testing by disk diffusion in solid media for clinical purpose, and detection of the vancomycin resistance genotype by DNA strip assay.

^cPositive culture with identification of Enterococcus faecium by mass spectrometry assay with MICs ≤8 mg/L for vancomycin and/or teicoplanin on E-test strips.

^dPositive culture with identification of E. faecium by mass spectrometry assay, MICs >8 mg/L for vancomycin and/or teicoplanin on E-test strips, and antibiotic susceptibility testing by disk diffusion in solid media for clinical purpose.

^eEstimated costs of inpatient care based on reimbursement rates of the diagnosis-related group. In France, the diagnosis-related group price is calculated by multiplying standard amounts for operating and capital expenses found in yearly surveys by a national “weight” associated with the DRG for each hospitalisation. The weighting takes in account variations due to geographic area and atypical observations.

^fNumber of missed patient-days due to the interruption of patients’ transfers and admissions.

^gCosts estimation assuming the strict implementation of French national guidelines with three weekly screening of patients.

On 1 March, new admissions were allowed in a different area of the ward. Two additional weekly screening of contact patients performed using the same culture method identified no additional cases.

This local GRE control protocol resulted in a 72-hour period without admissions or transfers, with 41 missed patient-days and €13,968 of lost income (Table 1). Following the national guidelines would have resulted in 15 days without admissions or transfers, with 250 patient-days of loss of activity and €85,175 of lost income. The cost of microbiological testing using the culture method was €333.50. The global estimated costs were therefore €14,302 and €86,175 with the local and national guidelines, respectively.

As described previously, a secondary case was rapidly identified using the Xpert™ vanA/vanB PCR. This patient stayed in the nephrology ward from 1 January to 20 February, 2012. The nephrology ward has 28 beds with 12 double and 4 single rooms. On 28 February, the 22 patients hospitalized during the same period and still present in the nephrology ward were considered to be contact patients and were screened for GRE by rectal swabbing. We decided to use the GeneXpert™ test for this purpose and, given the TAT with this test, to continue transfers and admissions as usual unless another GRE-positive patient was identified. The median time to results was 4.6 hours after sampling (Table 1). However, because only a four-site GeneXpert™ system was available, the results for all 22 patients were obtained 9.5 hours after sampling, and decision about the GRE-control strategy was therefore made at the end of the day. None of the contact patients had vanA-positive strains. Consequently, transfers and admissions were continued. The overall cost of PCR testing for vanA-positive GRE was €870.40.