In cardiothoracic surgery, the costs of surgical-site infection (SSI)
arise from additional postoperative procedures (approximately US $5000
per patient) and prolonged hospital stay (approximately $11 500 per p
atient). Application of nasal mupirocin reduced SSIs by 63% compared w
ith historical controls. This would have resulted in savings provided
that the attributable cost of an SSI was more than $245. Mupirocin was
estimated to reduce the risk of bacteraemia in haemodialysis patients
by 84% compared with historical controls. A model using data on Medic
are payments for haemodialysis admissions was used to estimate the imp
act on hospital costs. The conclusion was that mupirocin would have be
en cost-saving but the model did not provide sufficient detail about h
ospital costing to allow assessment of its relevance in other settings
. In a prospective, randomized, placebo-controlled trial in continuous
ambulatory peritoneal dialysis (CAPD) patients, mupirocin reduced the
risk of staphylococcal exit-site infection (ESI) from 0.42 to 0.14 pe
r patient-year. However, as in a previous comparison with historical c
ontrols, there was an increase in the rates of ESIs caused by Gram-neg
ative bacteria in patients who received mupirocin, bringing the rate o
f total ESIs up to that observed in the placebo group. There was some
evidence that infections caused by Gram-negative bacteria had less sev
ere consequences than staphylococcal infections. It is concluded that
application of nasal mupirocin to nasal carriers of Staphylococcus aur
eus may be cost-saving in patients undergoing cardiac surgery or haemo
dialysis but, if the analysis is restricted to the cost of management
of ESIs, it may not be cost-saving in CAPD. However, reducing the risk
of staphylococcal ESI may reduce the risk of catheter loss and subseq
uent transfer to haemodialysis and this merits further study.