Staphylococcus aureus bacteraemia and endocarditis : epidemiology, short- and long-term mortality

Abstract: Staphylococcus aureus is a major cause of bloodstream infections and endocarditis. S. aureus bacteraemia (SAB) is associated with substantial morbidity and mortality, and endocarditis is a severe complication. Population-based studies on S. aureus bacteraemia have been sparse, and few large studies exist on S. aureus endocarditis (SAE). The objective of this thesis was to study the epidemiology, characteristics, and short- and long-term outcome of S. aureus bacteraemia and endocarditis in Iceland and Stockholm. In paper I and II we studied SAB in the entire Icelandic adult and paediatric populations. Cases were retrospectively identified at the clinical microbiological laboratories. In adults the incidence was 24.5 /100,000 person-years during 1995-2008 (721 cases), increasing by 28% during the study period (p=0.01). The paediatric incidence was 10.9 /100,000 child-years during 1995-2011 (146 cases), decreasing by 36% during the period (p=0.001). At the same time the average annual frequency of blood cultures from children analysed at the main study site decreased by 27% (p<0.001). SAB incidence was highest in infants (<1 year), 58.8 /100,000. The proportion of adults with nosocomial infections decreased from 56% in 1995-99 to 39% in 2005-08 (p=0.001), while community acquired SAB increased from 29% to 46% (p<0.001). Health-care associated community-onset cases were 15%. Among the paediatric cases 34% were nosocomial, 14% health-care associated, and 51% community acquired. Bone or joint infection was the focus of SAB in 40% of children, followed by intravascular catheters in 30%, and an unknown focus in 10%. The 30-day mortality in adults was 17.1%, and decreased from 22.2% during 1995-99 to 11.4% during 2005-08 (p=0.001). The 1-year mortality was 33.0%, and decreased from 38.9% to 28.2% (p=0.06). In children the SAB-related mortality was 0.7%, 30-day mortality 1.4%, and the 1-year mortality 3.6%. These case fatality ratios are lower than those observed in most previous studies. In paper III we studied SAE in adults in Stockholm, and in paper IV we specifically focused on SAE in people who inject drugs. Individuals treated for SAE at the Department of Infectious Diseases at the Karolinska University Hospital were retrospectively identified by diagnostic codes from medical records. The calculated incidence of SAE in adults in Stockholm County was 1.56 /100,000 person-years during 2004-13 (245 cases), and the incidence of SAE related to intravenous drug use (IVDU) was 0.76 /100,000 person-years (120 cases). This incidence is high in comparison with other regions. The SAE incidence increased by 42% during the study period (p=0.002), and this was largely caused by a change in the incidence of the IVDU-related SAE which increased by 91% (p=0.02). The SAE incidence among people who inject drugs in Stockholm was estimated to be 2.5 (range 1.5-6.5) per 1,000 person-years. Thirty-day, in-hospital, and 1-year mortality rates were 6.1%, 9.0%, and 19.7%, respectively, among all SAE cases. In-hospital and 1-year mortality rates associated with IVDU-related SAE were 2.5% and 8.0%, respectively. The case fatality ratios noted are very low compared to previous reports. Age and female sex were independently associated with in-hospital mortality in a multivariate analysis, and age and left-sided disease with the 1-year mortality. Central nervous system (CNS) involvement was observed in 12% of patients, and valvular surgery was performed during hospitalisation in 15%. In left-sided SAE the strongest predictors for surgery were lower age and not being an intravenous-drug-user, and for CNS involvement lower age. In conclusion, we found an increasing incidence of SAB and SAE in adults, probably related to a change in risk factors both for SAB and SAE, and possibly due to more liberal diagnostics. The decrease noted in SAB incidence in children was probably in part due to lower blood culture frequency and possibly a result of infection control measures introduced. The reason for the favourable short- and long-term outcomes associated with SAB and SAE in Iceland and Stockholm is not clear. It could be related to diagnosis of more early and mild cases, but other factors might also have contributed.