JSTOR

Methods

The Besançon Hospital is a university‐affiliated hospital in France. For this study, data were collected for 5 years (from 2002 through 2006) in 4 units: a 15‐bed medical intensive care unit (ICU), a 15‐bed surgical ICU, a 12‐bed neurosurgery unit, and a 16‐bed general surgery unit. MRSA screening on admission and once a week thereafter (on a given day of the week) of all patients by systematic sampling of the anterior nares was recommended by the infection control department. MRSA isolates were also recovered from clinical samples obtained according to signs of infection. In addition to the screening program, standard infection control measures were used for patients with MRSA colonization or infection; these measures consisted of contact isolation of identified MRSA carriers in single rooms (or cohorts), use of dedicated material (gown, gloves, and if indicated, masks), and reinforcement of hand disinfection with alcohol‐based hand rub. MRSA cases were categorized as imported (involving patients whose screening or clinical sample tested positive for MRSA within 48 hours of hospitalization) or hospital acquired (involving patients whose admission screening had negative results for MRSA but whose subsequent screening or clinical sample had positive results for MRSA). A transmission index, defined as the ratio of the number of hospital‐acquired cases to the number of imported cases, was calculated. For example, a transmission index of 1.0 would indicate that, on average, each imported MRSA case generated one secondary (hospital‐acquired) MRSA case. Among hospital‐acquired MRSA cases, we distinguished 2 categories of patients: those who had positive results for MRSA only from a screening and those who had positive results for MRSA from a clinical sample (regardless of MRSA screening result). For each unit, the MRSA colonization pressure was calculated as the ratio of the number of MRSA‐positive patient‐days to the total number of patient‐days. Patients were considered to have MRSA colonization or infection from the date of recovery of the sample that yielded positive results to the date of hospital discharge.

Results

Of 8,101 patients admitted during the survey, 6,065 patients (74.9%) were screened for MRSA upon admission, representing 92.2% of the days of hospitalization. The rate of screening varied from 61.3% in the surgical unit to 82.7% in the medical ICU. The median number of screenings per patient (on admission and during the stay) was 3.03 (there were 2.9 screenings per patient in both ICUs, 3.1 screenings per patient in the general surgery unit, and 3.55 screenings per patient in the neurosurgical unit).

During the survey, 373 patients had positive test results for MRSA, giving an incidence of 4.31 MRSA cases per 1,000 patient‐days (6.15 MRSA cases per 100 hospitalizations). The distribution of MRSA cases is reported in the Table. Of the 373 patients, 148 (39.7%) acquired MRSA during their stay, whereas 225 (60.3%) were colonized on admission. Of the 148 cases of hospital‐acquired MRSA, 83 cases (56.1% of hospital‐acquired MRSA cases and 22.3 % of all MRSA cases) were detected only with weekly screening samples (excluding screening on admission). This proportion varied according to the departments: 6.1% in the general surgical unit, 23.3% in the medical ICU, 31.6% in the surgical ICU, and 43.5% in the neurosurgical unit. The global transmission index was 0.66 (0.28 in the general surgical unit, 0.71 in the medical ICU, 0.85 in the surgical ICU, and 2.07 in the neurosurgical unit). The total MRSA colonization pressure was 100.3 MRSA‐positive patient‐days per 1,000 patient‐days (54.9% of this colonization pressure was attributable to imported MRSA cases and 45.1% to hospital‐acquired MRSA cases). The 83 cases that were detected only with weekly MRSA screening were responsible for 13.0 MRSA‐positive patient‐days per 1,000 patient‐days (13.0% of total MRSA colonization pressure).

Table 1.  Distribution of Cases of Methicillin‐Resistant Staphylococcus aureus (MRSA) Colonization or Infection, by MRSA Status and Hospital Unit

Open New Window

Discussion

We explored the usefulness of weekly MRSA screening in 4 units of a university hospital: 2 adult ICUs, 1 neurosurgical unit, and 1 general surgery unit. This last unit is a dedicated cohort facility that is mainly dedicated to orthopedic patients with infection and some other surgical patients with infection. We have previously demonstrated that having this dedicated cohort facility is useful in the reduction of MRSA cross‐transmission.⁵ Weekly MRSA screening allowed the identification of 83 MRSA carriers who would have remained undetected by the combination of MRSA screening upon admission and clinical samples obtained according to signs of infection during hospitalization. These 83 MRSA carriers represent 22.3% of all patients with MRSA colonization or infection. This rate increases to 29.3% if we omit the general surgical unit, where most of the patients with MRSA colonization or infection were classified as imported because of the mode of recruitment of this unit (some patients were admitted after discovery of an infection in another surgical unit). This 22.3% of MRSA cases was responsible for 13.0% of the total MRSA colonization pressure. The difference between these 2 rates is linked to the prolonged length of stay associated with MRSA infection, compared with the length of stay associated with simple MRSA colonization (data not shown).

An MRSA control program has been implemented in all high‐risk units of our hospital, including the 4 units that participated in this study. Because the control program was not perfectly implemented, MRSA cross‐transmission occurred. However, the transmission index (0.66) was relatively low: 3 cases of imported MRSA were needed to observe 2 new hospital‐acquired MRSA cases. This transmission index was similar to those reported in facilities that applied a similar design of MRSA control (including weekly screening).^6‐8

Our study has a number of limitations. Only 74.9% of admitted patients were screened. Even if most of the unscreened patients had a limited length of stay (less than 2 days) in the units considered, this represents a potential bias in the calculation of the true MRSA colonization pressure. We did not obtain surveillance cultures from healthcare workers, who may carry a substantial reservoir of MRSA that limits the success of control measures.⁹ We did not perform audits on the application of hygiene procedures, especially hand hygiene observation. Finally, this retrospective study was not designed to determine the efficiency of weekly MRSA screening, compared with the efficiency of simple admission screening, in the reduction of MRSA acquisition.

In conclusion, our results suggest that weekly MRSA screening allows the detection of a significant proportion of MRSA cases that would have remained undetected with screening on admission only. These MRSA cases are likely to facilitate the spread of MRSA and could represent a major impediment to MRSA control.^10,11 The introduction of weekly MRSA screening requires a careful assessment of MRSA epidemiology and may be useful in ICUs and similar units.