Objective: To assess the incidence of local and systemic infection cau
sed by central venous catheters in a general hospital population. Sett
ing: Concord Repatriation General Hospital, Sydney, April to August 19
91 inclusive. Design: A prospective survey of all patients with in sit
u central venous catheters. Systemic catheter-related infection was de
tected by blood and routine catheter tip culture, and local infection
by clinical observation of the catheter exit site. Outcome measures: L
ocal and systemic infection and complications. Results: 479 central ve
nous catheters were surveyed in 311 patients. Local infection develope
d in association with 54 catheters (11%) and systemic infection with 3
2 (6.7%). Local infection was predictive of systemic infection, but it
s absence did not exclude systemic infection. Haemodialysis catheters
were responsible for a higher systemic infection rate than other cathe
ter types, the most common organism responsible being methicillin-resi
stant Staphylococcus aureus. Twenty per cent of all bacteraemias (33/1
60) detected in the hospital occurred in patients with a central venou
s catheter and 24 of these (73%) were definitely or probably due to th
e catheter. Staphylococci were the predominant isolates and 40% of the
methicillin-resistant S. aureus bacteraemias detected were due to cat
heter-related infection. Infection complications were few: three patie
nts developed local abscesses; one endocarditis; and two died. Conclus
ions: Central venous catheter-related infection is common in general h
ospital populations. Staphylococcal bacteraemia and local infection in
a patient with a central venous catheter are strongly suggestive of c
atheter-related systemic infection. Empirical antibiotic treatment sho
uld include at least antistaphylococcal cover.