In the current climate of cost containment and quality control, nosoco
mial infection is a worrisome adverse event in hospital care. Hospital
ised patients require care for increasingly severe illnesses, and are
therefore more susceptible to infection, especially by opportunistic m
icro-organisms. It is thus necessary to accurately assess and adjust f
or the severity of the underlying illness in studies of risk factors i
nvolved in nosocomial infections. The appearance of new diagnostic and
therapeutic techniques provides novel opportunities for infection con
trol and represents a constant challenge to hospital systems. The cont
inuous selection of resistant flora, together with the identification
of new pathogens, calls for a reconsideration of hospital policies reg
arding the dispensation of antibiotics. Epidemiological surveillance c
ontinues to be the most important aspect of attempts to monitor infect
ion control programmes, and to identify changes in risk factors that m
ay increase the infection rate. Among the major challenges now facing
the infection control practitioner is the use of nosocomial infection
rates as an indicator of quality of care. Awareness of infection stati
stics would serve as a stimulus to the prevention and control of infec
tion, but would be useless if not accompanied by adequate systems to g
uarantee the comparability of data from different studies and centres.
Suitably sensitive and specific surveillance systems should be develo
ped, and the use of site-specific and procedure-specific infection rat
es adjusted for the patient's intrinsic risk should be encouraged.