The key question remains: given an optimally functioning horizontal program (i.e., near perfect compliance with hand hygiene and chlorhexidine bathing), what is the incremental benefit of a superimposed vertical strategy?Voila! A new study in Lancet Infectious Diseases by Marc Bonten et al addresses our question. In this 3-year study in 13 European ICUs, involving over 8,500 patients, there was a baseline data collection phase (6 months), followed by a hygiene improvement phase (hand hygiene campaign + universal chlorhexidine bathing for 6 months), followed by cluster randomization to a rapid screening group + contact precautions for carriers or a conventional screening group + contact precautions for carriers. Both screening groups used chromogenic agar to detect MRSA, VRE and ESBL; the rapid screening group also used PCR testing for MRSA and VRE. Primary endpoints were the acquisition of MRSA, VRE or MDR-GNR.
The study showed:
- Hand hygiene increased from 52% in phase 1, to 69% in phase 2, and 77% in phase 3.
- Chlorhexidine bathing was 0% in phase 1, and 100% in phases 2 and 3.
- Improved hand hygiene + chlorhexidine bathing (phase 2) resulted in a significant decrease in MRSA acquisition, with no change for VRE or MDR-GNR.
- Neither search-and-destroy strategy resulted in any further reduction for any of the targeted pathogens.
- There was no change in the prevalence of chlorhexidine resistance genes in MRSA isolates in phase 1 vs. phase 3 (13% vs 14%).
So we now have another study demonstrating the lack of need for active detection and isolation to control multidrug resistant pathogens. Will the search-and-destroyers finally pack it up and go home?
Photo: Jeff Swensen, New York Times

Can we say that CHG (and hand hygiene) then had no effect on VRE and MDR-GNRs then? If that's the case, I'll keep wearing gloves when I see patients.
ReplyDeleteSo Mr Author of this article, what are you saying?? do you know this ONLY JUST ONE study out of like a hundred thousand that says the opposite??
ReplyDeleteInteresting read.
ReplyDelete