Comparison of mortality associated with bacteraemia caused by meticillin-resistant Staphylococcus aureus on arrival at hospital versus hospital-acquired MRSA bacteraemia: a cohort analysis from the BURDEN study group

Abstract number: P1458

Ansari F., Evans J., Davey P.

Objective: MRSA bacteraemia is generally regarded as a hospital acquired infection so may be less likely to be treated empirically when present on arrival at hospital. We therefore tested the hypothesis that mortality associated with MRSA bacteraemia on arrival at hospital is greater than mortality with hospital acquired MRSA bacteraemia.

Methods: The study population was adults (age 18 or over in 2005) resident in Tayside, Scotland from 1st January 2005 to 30th September 2006 who had a new admission to Ninewells Hospital between 1st July 2005 and 30th June 2006. The cohort enrolled patients from this population who were admitted to wards where at least one case of bacteraemia with MRSA occurred in the study period. We defined bacteraemia on arrival at hospital as a positive blood culture within 48 h of admission and hospital acquired bacteraemia as a positive blood culture 48 h after admission. The comparator group was all patients admitted to the same wards in the same time periods without S. aureus bacteraemia. The outcome was 30 day mortality, ascertained from the national registry of deaths. We used multivariable analysis to estimate the relative risk of mortality associated with MRSA bacteraemia adjusted for age, sex and Charlson co-morbidity index (CCI) calculated from the discharge diagnoses of the index admission and of any hospital admission in the previous year.

Results: The cohort included 4,397 patients; 21 had MRSA bacteraemia on arrival, 34 had hospital acquired MRSA bacteraemia and 56 had MSSA bacteraemia. In comparison with other cohort patients those with MRSA bacteraemia were older (mean age 67 vs 63) and had more comorbidities. Mean CCI was 3.76 for patients with MRSA bacteraemia on arrival, 2.56 for hospital acquired MRSA bacteraemia patients and 1.91 for the comparator patients. In comparison with patients with a CCI of 0 the relative risk of 30 day mortality was 1.98 with CCI 1–2, 2.65 with CCI 3–5 and 4.89 with CCI 6 or greater. After adjustment for age, sex and CCI the relative risk of 30 day mortality with MRSA bacteraemia on arrival at hospital (2.48, 95% CI 0.84–7.27) was higher than with hospital acquired MRSA bacteraemia (1.11, 95% CI 0.38–3.29) and also higher than with MSSA bacteraemia on arrival at hospital (1.29, 95% CI 0.38–4.39).

Conclusions: Our cohort study supports the hypothesis that MRSA bacteraemia on arrival at hospital is associated with a higher 30 day mortality than hospital acquired MRSA bacteraemia.