Selective decontamination of the digestive tract: time to implement it in all UK intensive care units? Maybe not yet.

In this issue, a survey of UK intensive care units (ICUs) is described and the uptake of selective decontamination of the digestive tract (SDD) within these units is reported. Impressively, the authors achieved a 100% response from all the units contacted, but out of the total of 250, only three reported using a full SDD protocol including i.v. antibiotics, 10 more (13 in total) reported using SDD protocols without the i.v. component. The SDD protocols when used were applied variably; some units used them in all ventilated patients (n1⁄47), others reserved SDD for trauma patients, and in one unit just for liver patients. Of the 250 units approached, 205 contributed data to the case mix programme of the Intensive Care National Audit & Research Centre (ICNARC) and these were used to compare unit infection rates. Despite only three units using the i.v. component, acquired blood infections were fewer in these units. First, it is briefly worth considering the survey response rates. 100% is an astounding response to a survey. Physician surveys are normally poorly responded to; a 50–70% response would normally be considered the best that could be expected, so why did this one succeed so well? There are a few tactics which may improve response rates. First, the clinical directors were the point of contact. Choosing a person with authority within a department, who is easily identifiable and likely to know the answers to the questions, circumvents problems associated with contact, confidence, and knowledge. Secondly, the questionnaire was very simple with most questions requiring a tick rather than precise data or detailed comments, and thirdly, if the first post did not work, this was followed up with a reminder and then a phone call with the questionnaire filled in over the phone; there was no escape. It is also likely that as a small community responding to a well-known chief investigator, there was a willingness to contribute to the survey. Finally and perhaps as important as any other reason, the subject was considered important. SDD evokes strong opinions and raises very many questions. There is no one standard form of SDD; different antibiotics may be applied in different ways and some regimens include parenteral antibiotic application and others do not. A typical regimen will include non-absorbable agents such as polymixins B and E (colistin), tobramycin, and amphotericin applied as a paste to the oropharynx and in suspension via a nasogastric tube to the gastrointestinal tract. Cefotaxime is often used as the parenteral antibiotic of choice. A distinction is made between selective oropharyngeal decontamination (SOD) and SDD dependent on whether the antibiotics are applied beyond the oropharynx or not and a further distinction based on whether SDD include parenteral antibiotics or not. Does SDD work? The answer would appear on the evidence available that it does. ACochrane analysis in 2009, a multicentre cross-over trial of SDD, SOD, and standard care involving 5939 patients, and a very recent meta-analysis have all shown reductions in infection rates in critically ill patients and improved outcome. The recent meta-analysis from Price and colleagues suggest a mortality benefit from the use of both SDD and SOD with odds ratio of 0.73 (95% confidence interval 0.64–0.84) and 0.85 (0.74–0.97), respectively. In a speciality where results of interventions are rarely this effective, data such as these would normally set the standard in every unit, so why not in this case? The correspondence pages are where the clues to this may be found. – 11 The major doubt is over antibiotic resistance. The term counterintuitive has been used to justify not using SDD and when the current literature is studied carefully, this begins to look a little more justified than the headline results might suggest. de Smet and colleagues followed up their original multicentre trial with an analysis of antibiotic resistance and found that resistant organisms were fewer in the SDD group; 15% patients acquired respiratory tract colonization with highly resistant microorganisms during standard care, compared with 8% during SDD and 10% during SOD. However, this was criticized observing that the overall rate of resistant organisms developing was unusually low, the observed period was too short to assess resistance fully, and that the potential resistant pathogens selected for study were not covered by SDD prophylaxis. These criticism have been widely echoed and a recent meta-analysis of the occurrence of resistance emerging as a consequence of SDD has concluded that while no evidence of resistance was found, the data do not yet exist for confidence in the long-term safety of SDD. The emergence of resistant organisms is complex. In 1968, methicillin-resistant Staphylococcus aureus (MRSA) was being described as a British problem having been identified in Plymouth. In 1972, screening of patients in Philadelphia revealed zero cases of MRSA, but now in the USA, MRSA is endemic and increasing numbers of strains of vancomycinEditorial BJA