Selective decontamination of the digestive tract (SDD), a strategy designed
to prevent or minimize the impact of infection by potentially pathogenic m
icro-organisms in critically ill patients requiring long-term mechanical ve
ntilation, comprises four component protocols, aiming to control the three
types of infection occurring ill such cases: (i) a parenteral antibiotic, c
efotaxime, administered fur a few days to prevent primary endogenous infect
ions typically occurring 'early'; (ii) the topical antimicrobials polymyxin
E, tobramycin and amphotericin B employed throughout the stay in the inten
sive care unit to prevent secondary endogenous infections tending to develo
p 'late'; (iii) a high standard of hygiene to prevent exogenous infections
that may occur throughout the stay in the intensive care unit; (iv) surveil
lance samples of throat and rectum to distinguish between these three types
of infection, to monitor the compliance and the efficacy of the treatment,
and to detect the emergence of resistance at an early stage. A recent, rig
orous, meta-analysis examining 33 randomized SDD trials involving 5727 pati
ents demonstrated a significant reduction in overall mortality (20%) and in
the incidence of respiratory tract infections (65%); conclusive evidence t
hat SDD saves the lives of critically ill patients and confirmation that SD
D is now an evidence based medicine manoeuvre. This same meta-analysis foun
d no instance of the emergence of resistance or of associated superinfectio
ns and/or outbreaks in any of the 33 studies during a period extending upwa
rds of 10 years. By the criterion of cost-per-survivor, four recent randomi
zed trials showed that patient survival is improved more cheaply by employi
ng SDD than by the traditional approaches. (C) 2000 The Hospital infection
Society.