A new study in JHI reports the findings of longitudinal environmental sampling in a paediatric BMT unit. Perhaps unsurprisingly, sites associated with frequent contact by staff (especially computer equipment) had the highest bacterial counts. These sites also had the highest bacterial diversity suggesting they are accumulating contamination from multiple sources, and acting as a bacterial possible interchange. But computer equipment is diligently cleaned and disinfected after each use, right…?

This study was conducted on a paediatric bone marrow transplant (BMT) unit at Great Ormond Street in London. So, needless to say, this is a very high risk setting with the most immunocompromised patients you could find. The longitudinal study comprised collecting 193 samples over 9 weeks from a range of environmental surfaces in the unit. Samples were plated onto non-selective agar, and most were speciated using MALDI-TOF. 16S rDNA bacterial profiling was done on a small number of the colonies (n=30).

The most contaminated sites identified were computer keyboards, with keyboards accounting for 5/6 sites with the highest mean colony counts. However, even these ‘high’ colony counts (11 cfu / 100 cm^2) were rather low in the context of other studies – some have argued for standards of either 2.5 or 5 cfu per cm^2 – and the highest colony counts in this study were 0.1 cfu per cm^2. 16S rDNA profiling found that computer equipment also had the highest bacterial diversity. This isn’t surprising given the number of hands that are touching these items.

The study team ran some analysis to categorise the environmental sites by various factors to see whether any trends emerged. Grouping sites by perceived contamination risk or possible transmission route to patients (ingestion or contact) were not good predictors of actual contamination. On the the other hand, grouping sites by activity, sample type, or predominant contact group was a better predictor of contamination, with sites associated with ward management activity, computer devices, and sites predominantly touched by clinical staff having the higher levels of contamination.

A couple of limitations to consider. The study did not include any sampling for antibiotic-resistant bacteria. Whilst it is useful to see the distinction between “environmental bacteria” and “human microflora”, a more impactful distinction is whether or not antibiotic-resistant potential human pathogens are present or not. Also, I found it difficult to interpret the significance of differences between contamination levels by sample grouping.

A few people have asked me recently whether we should consider routine environmental sampling. I tend to agree with the current advice in most guidelines, that routine environmental sampling is useful for research purposes, and to search for a point source during outbreaks – but has little value outside of these scenarios. And I don’t think the kind of data that this study has provided would be useful for ongoing risk surveillance.

The major implication of the research is to highlight the challenges associated with contaminated computer equipment (especially keyboards) in hospitals. I’ve been involved in a couple of studies about keyboard contamination – including the inimitably named “How dirty is your QWERTY” study*”! I challenge you to go out and have a look at some of your keyboards. They are often visibility contaminated, and a quick hands-on inspection will tell you that they are difficult to clean and disinfect. Some designs are better than others – but all present challenges. Another issue is responsibility for cleaning and disinfecting them. Another recently published study found that many high-touch items of equipment lack clarity for cleaning and disinfection responsibility (and process for that matter)!

Finally, a reminder that this study will be the subject of Journal Club on Wednesday this week (details and registration here).

* Whilst I was a co-author, I can’t claim to have come up with the name!