- Andrew P Walden, consultant in intensive care medicine1,
- Marc J Bonten, professor of molecular epidemiology of infectious diseases2,
- Matt P Wise, consultant in critical care medicine, NISCHR AHSC fellow3
- 1Intensive Care Medicine, Royal Berkshire Hospital, Wantage OX12 8HG, UK
- 2University Medical Centre, Utrecht, Netherlands
- 3University Hospital of Wales, Cardiff, UK
- Correspondence to: A P Walden
- Accepted 14 August 2012
Healthcare associated infection represents a major burden for critically ill patients; a recent point prevalence survey by the Health Protection Agency observed that 23.4% of patients in intensive care units had evidence of a healthcare associated infection.1 Ventilator associated pneumonia remains the leading cause of nosocomial infection in this population, and, although recent estimates of attributable mortality (5-10%) are lower than previously thought, length of stay and treatment costs are substantially increased.2 3 Colonisation of the oropharynx with enteric bacteria is considered a key step in the development of ventilator associated pneumonia and offers a potential site for intervention with oropharyngeal decontamination.
Selective digestive decontamination involves the administration of topical, non-absorbable antibiotics to the oropharynx and stomach via a nasogastric tube in combination with parenteral antimicrobials to reduce the burden of potentially pathogenic bacteria in the aerodigestive tract. Some studies have focused on decontamination strategies limited to the oropharynx alone (selective oral decontamination), avoiding enteral and intravenous antibiotics. Selective digestive decontamination was first used for immunocompromised haematology patients, but this intervention has been extensively studied in intensive care units over the past three decades. However, many clinicians remain sceptical as to whether this evidence is applicable to different healthcare systems, which vary according to environment and antibiotic resistance rates, and their own clinical practice.
What is the evidence of uncertainty?
A search of PubMed, the Cochrane Library, and Embase identified nine published meta-analyses on the topic of selective digestive decontamination in intensive care patients. The most recent was an updated Cochrane review in 2009, which identified 36 randomised clinical trials involving 6914 patients.4 However the largest study of selective digestive decontamination5 was not included on the basis that the cluster design prevented individual patient randomisation.4 The odds ratio for death was 0.75 (95% confidence interval 0.65 to 0.87) with a …