A comprehensive and impressive cluster randomised crossover study published in Lancet ID examines whether it makes sense to use single rooms (as compared with multi-bed bays) to apply contact precautions for patients known to be carrying ESBL-Enterobacteriaceae. I need to be careful what I say, because fellow bloggers Marc and Andreas are co-authors. However, the gist seems to be: don’t bother with single rooms for ESBL-E carriers – but many hospitals don’t have capacity to do that anyway, so this may not be a practice-changing finding in many parts of the world!
The large study was a cluster randomised crossover study on medical and surgical wards in 16 Dutch hospitals. Each hospital did a block of ‘isolating’ patients with an ESBL-E positive clinical culture by applying contact precautions in single rooms or in multi-occupancy bays in a randomised sequence. The primary outcome was the acquisition of ESBL-E of a matching type to the index patient’s isolate. 693 index patients and 9527 ward-contacts were enrolled and 463 index patients and 7093 ward-contacts were included in the per-protocol population (who received the “correct” isolation strategy). ESBL-E transmission was identified in 11 (4%) of 275 index patients in single rooms, and 14 (7%) of index patients in multi-occupancy bays. The conclusion was that delivery of contact precautions for ESBL-E in multi-occupancy bays was non-inferior to the delivery of contact precautions for ESBL-E in single rooms.
A couple of reflections:
- This study isn’t about “let’s ditch contact precaution and see what happens”. Contact precautions were applied in both single rooms and multi-occupancy bays.
- “ESBL-E” is a very broad street. We know that the transmission dynamics of E. coli (which accounted for around 75% of the ESBL-E in this study) are fundamentally different to K. pneumonoiae (which accounted for around 14% of ESBL-E in this study). Indeed, if we delve into the Appendix, we find that the effect of single rooms was more pronounced for K. pneumoniae than for E. coli. In a perfect world, I’d love to see this study repeated but solely focussed on K. pneumoniae; I have a feeling that multi-occupancy bays wouldn’t be non-inferior to single rooms for this organism (but accept fully that it matters little what I feel in the face of a cluster randomised RCT!). In support of this, in previous work by some of the same authors, the hospital transmission rate of non-E. coli ESBLs (most of which were K. pneumoniae) was more than double the rate of hospital transmission of E. coli ESBL.
- Practically speaking, how many hospitals have the single room capacity to isolate ESBL-producing coli (or ESBL-producing K. pneumoniae for that matter)?
This is a huge, scientifically sound, and interesting study. It certainly confirms for me that there’s not much value in applying contact precautions for E. coli ESBL in single rooms (even if some are available). But, for me, the jury remains out as to whether we should pursue contact precautions in single rooms for K. pneumoniae ESBL (recognising that they will get “bumped” by higher priority pathogens for single rooms much of the time)!