As a young and angry PhD student I was confronted with the concept of Selective Digestive Decontamination (SDD). That was in the early nineties (previous millennium). Coming from the field of microbiology I expected that SDD would increase antibiotic resistance. It were intensivists that told me not to worry: “antibiotics in SDD kill bacteria and dead bacteria cannot mutate”. They may have been right, suggests a new study.

SDD, in short, is daily application of topical antibiotics (colistin and tobramycin) in the oropharynx and gut of ICU patients. The bad bugs disappear from the conventional surveillance cultures and in settings with low antibiotic resistance (such as Dutch ICUs) it’s associated with lower incidence of ICU-acquired bacteremia and MDR-GNB carriage and improved ICU- and hospital survival. But what happens under the detection limit of conventional culture techniques? Is resistance building up in the anaerobic flora, ready to spread to Gram-negatives after the antibiotics are stopped?

Elena Buelow studied fecal samples from 10 ICU patients during and after ICU stay (all received SDD in ICU) and 10 healthy subjects, see. The ICU patients were considered healthy before ICU admission, thus there gut composition at the time of admission should not yet be disturbed to much by illness and antibiotics. Microbiota profiling demonstrated marked differences between ICU patients and healthy subjects, and that composition changed during and after ICU-stay. The Gram-negative flora (such as E. coli) declined to astonishingly low levels, whereas enterococci increased.

The juice is in the resistome analysis; the bacterial DNA encoding for resistance (antibiotic resitance genes, ARG) . The number of detected ARG per sample ranged from 6 to 38, and 11 ARG were detected in >80% of healthy subjects and critically ill patients. Yet, only 2 patients had ESBL and 1 had KPC (in the 1st sample only); Dutch patients! Four ARG were more prevalent in ICU patients: 1 encoding low-level aminoglycoside resistance in enterococci, the other for resistance to macrolides, tetracycline and chlorhexidine (difficult to link to SDD). In clinical studies SDD was not associated with an increased incidence of ICU-acquired enterococcal bacteremia, making the selection of more enterococci and more ARG encoding for low-level aminoglycoside resistance not that relevant.

This study adds another (little) piece to our understanding of SDD. Although not compared to a control population in ICU (would have required SDD abstinence against guideline recommendation) it confirms that Gram-negative bacteria almost disappear without evidence of resistance capacity building for Gram-negative bacteria in the remaining flora. I’m sure that Jon Otter still will not be reassured, so more research is needed. As it is selection rather than mutations that fuels our current resistance problem, the intensivist was wrong: it should have been “dead bacteria cannot be selected”.