A powerful study from Korea demonstrates the value of chlorhexidine gluconate (CHG) daily bathing of ICU patients to reduce the transmission of carbapenem-resistant Acinetobacter baumannii (CRAB). Where enhanced screening, contact precautions and environmental disinfection failed, the introduction of CHG daily bathing made an impressive reduction in the transmission of CRAB.
The study was performed in a 16 bed ICU, with a 14 month control period, and a 12 month intervention period. The authors evaluated how many patients were identified as carrying CRAB within the first 48 hours of their admission to the ICU ('prevalent' cases) and those that were first identified as carrying CRAB after 48 hours on the ICU ('incident' cases). Universal admission screening was in place throughout the study, meaning that the incident cases are presumed acquisitions. In the control period, 21% of 593 eligible admissions acquired CRAB compared with 10% of 554 admissions in the intervention period (incidence density reduced from 44 to 21 per 1000 at-risk patient days, p<0.001) (see figure below). A time series analysis confirmed a significant reduction in the incidence density of CRAB associated with the introduction of CHG.
Figure legend: Rate of acquisition and environmental contamination in the control and intervention periods
The authors also compared environmental contamination at one timepoint in the control period, with four timepoints in the intervention period. The proportion of sites contaminated with CRAB in the intervention period was significantly lower than in the control period (30% of 127 vs. 10% of 540) (see figure above). It is not clear whether this is due to reduced shedding of CRAB from affected cases, or due to the fact that there were less cases around due to reduced transmission; it seems likely that both contributed to the reduction in CRAB environmental contamination.
Meanwhile there was no significant change in the prevalence of CRAB on the unit, compliance with hand hygiene or contact precautions, providing firm evidence that the reduction was not due to reduced importation of CRAB into the ICU, or improvements in infection control.
Due to the potential risk for CRAB developing reduced susceptibility to CHG, the authors tested the MIC and MBC of prevalent and incident cases identified during the intervention period. (It was a shame they didn't compare cases from the control period with the intervention period - but the isolates from the control period were not saved.) They found no significant different between the MIC/MBC of prevalent and incident cases, suggesting that the daily use of CHG did not result in reduced susceptibility. It's worth noting that the median MIC/MBC was 32 mg/L, and the concentration of CHG applied to the skin was in the region of 1000x more concentrated than this, making clinically significant reduced susceptibility a distant possibility.
One of the most striking aspects of the study is the extremely high levels of CRAB on the unit. Combining prevalent and incident cases suggests that a whopping 47% of 593 eligible patients were affected by CRAB at some point during their admission. These very high levels of CRAB probably limit the generalisbility of the findings to units with a lower admission prevalence. Notwithstanding this, there was a clear step-change in the rate of CRAB acquisition associated with the introduction of CHG daily bathing, suggesting that this intervention should be performed to reduce the transmission of CRAB.
A review in a very high impact journal (Annals of Internal Medicine) has addressed cleaning and disinfection strategies to prevent HCAI. Whilst it is great to see this issue addressed in a general, high impact medical journal, there are a number of important limitations to the review.
Firstly, it only includes data on MRSA, C. difficile and VRE. This is a shame, because it means that emerging data on the role of the environment in the transmission of Gram-negative bacteria - most crucially Acinetobacter baumannii - is excluded. It also means that there is no data on viruses such as norovirus, which limits the impact of the review. Secondly, it is very US centric, only including products or process available or under evaluation in the US. Thirdly, the scope of the literature review seemed to exclude some key literature, not least the statement that: 'No conference abstracts within the past 2 years were identified for inclusion', which certainly does not match our experience of recent conferences! Finally, there is poor resolution between the various automated room disinfection systems that are available, lumping together all hydrogen peroxide technology together, and doing the same for UV. This is disappointing, since we now know that there is a vast difference in the impact of UVC vs. pulsed-xenon UV, and between hydrogen peroxide vapour vs. aerosol systems.
The figures in the review are useful, providing a visual summary of the relative weight of evidence behind the various approaches to improving cleaning and disinfection in hospitals. The 'evidence hierarchy' adapted from the McDonald & Arduino version is also useful, providing a different perspective on how far up the hierarchy the various approaches have made it. However, it's important to note that the interpretation of these figures should be tempered by the fact that the review only included data from the US!
The key evidence gap identified is that there are few good quality studies with a clinical outcome (i.e. reduced infection or colonisation), which is a function of how difficult it is to perform these studies. However, a number of these are in progress, so this should improve over the coming years.
It seems clear that there is compelling evidence behind a number of usually complimentary but occasionally conflicting approaches, suggesting that several interventions to address environmental hygiene could be used simultaneously. Specifically, conventional cleaning and disinfection can be improved through implementing better disinfectants and processes for delivery by using wipes that are demonstrably effective against the target pathogens. It is also vital that monitoring processes are undertaken more rigorously and can provide organisations with data that demonstrate safe preparation of areas for patient occupancy. In addition, automated room decontamination systems can provide additional assurance that rooms are clean and safe for the next occupant.
If you were to put chlorhexidine and CRE in the ring, who would win? On the face of it, you'd back CRE, since this is carried in the gut and applying chlorhexidine to the skin won't have a hope of decolonising the gut. True, but what is the mechanism of transmission of a gut-dwelling organism? Probably, contamination is spread from the gut to the patient's skin, then picked up on staff hands or equipment resulting in a risk for onward transmitted. So, although chlorhexidine isn't going to decolonise a CRE carrier, it may reduce their burden of skin coloinisation and hence reduce the transmission risk. An additional challenge for chlorhexidine when dealing with CRE is that Gram-negative bacteria are inherently less susceptible to chlorhexidine than Gram-positive bacteria (such as MRSA) due to the structure of their cell wall.
A study (click here for the full article) pitting chlorhexidine against CRE was performed in long-term acute care hospital patients in Chicago recently. Five skin sites were sampled before and after CHG application from 62 patients treated daily with chlorhexidine. Just over half of these patients (56%) had skin colonisation with CRE at the start of the study (despite already being bathed daily with chlorhexidine). After chlorhexidine bathing, only 32% of the patients had skin colonisation with CRE, with a significant reduction in the proportion of all sites combined, and 4/5 sites individually noted.
The concentration of chlorhexidine on the skin was measured before and after the application of chlorhexidine. Unsurprisingly, the concentration on the skin was higher after application of chlorhexidine than before (312.5 vs. 78.1 mg/L). However, worth noting that the concentration of chlorhexidine measured on the skin, even immediately after application, was considerably lower than the concentration of chlorhexidine applied. Since the MIC90 of chlorhexidine against a collection of CRE was 128 mg/L, the concentration one day after application may have been too low to be effective. In support of this, a skin concentration of <128 mg/L was a risk factor for skin colonisation with CRE. Perhaps this argues for more frequent application of chlorhexidine?
This study does not demonstrate that using chlorhexidine daily bathing reduces the transmission of CRE. But it does demonstrate neatly that the transmission risk is reduced.