Population-based epidemiology of Staphylococcus aureus bloodstream infection in Canterbury, New Zealand

Background: Few contemporary reports describe population-based incidence of Staphylcoccus aureus bloodstream infection (SABSI). Aim: To describe longitudinal incidence of SABSI in a region of New Zealand with low MRSA prevalence. Methods: Blood cultures growing S. aureus were identified from hospital laboratories between 1 July 1998 and 30 June 2006. Record linkage was used to combine information from local and national databases into a single patient event record. Information from the New Zealand census was used to determine regional incidence of disease. An address-based measure of deprivation was used to analyse the relationship between incidence and socioeconomic status. Morbidity data were not collected. Results: 779 patients with SABSI were identified (482/779 (62%) male, 297/779 (38%) female). The crude incidence of S. aureus bacteraemia varied between 18.5-27.3/100 000 per annum. Three of 779 isolates (0.4%) were MRSA. Two hundred and seventy-seven of 776 (36%) patients with complete records had hospital-acquired SABSI. One hundred and forty-one of 778 patients (18%) died within 30 days and 235/778 (30%) died within a year. Proportional hazards survival models significantly associated age, male sex and more than 14 days of hospitalization in the year prior to index culture with adverse outcome. Higher socioeconomic status was associated with lower rates of SABSI (adjusted rate ratio 0.74, 95% confidence interval: 0.56-0.98, P = 0.007 after adjustment for age and sex, and comparing the highest and lowest deprivation quintiles). Conclusion: Population factors significantly influence SABSI incidence and survival. Further research is required to determine whether these have the potential to invalidate inter-hospital comparison of SABSI incidence as a measure of health-care quality.