The silent Mycobacterium chimaera epidemic


There has been much discussion about the risk of Mycobacterium chimaera infections associated with contaminated heater-cooler units (HCUs) used in cardiothoracic surgery. A study published recently in CID explores the risk in the UK, and provides further evidence to link these tricky-to-treat infections to contaminated HCUs.

The study includes a 4-year national look-back exercise to detect potential cases. This identified 18 probable cases (and quite a few more possible cases – 54). Whilst this represents a very rare outcome from cardiothoracic surgery (around 1 per 10,000 person-years), the problem is widespread with a quarter of cardiothoracic centres in the UK reporting a case, the case-fatality was high (50%), and the risk appears to be increasing (from <0.2 to 1.6 per 10,000 person-years between 2010 and 2013). There was a considerable time lag between the surgery and the detection of the infection: of the 18 probable cases identified, more than half (61%) presented >1 year after surgery, with one case presenting >5 years after surgery!

The team also undertook an aerobiological / environmental investigation to support the clinical findings. HCUs taken out of the clinical environment were filled with sterile water, and fairly promptly began to spew out M. chimaera and other organisms, especially from the rear of the machine. Presumably this is due to biofilm in the water tanks / pipes. The team then performed a field sampling investigation of 35 HCUs in the UK. Mycobacteria grew in water from 27/35 (77%) HCUs; M. chimaera grew in water from 17/35 (48%), and from air samples close to 6/25 (24%) of HCUs in operation. Mycobacteria were not recovered from the air around HCUs from a different manufacturer, but were found in water in these machines. Does this mean that some models are less risky than others?

This study raises a number of important questions:

  • Is this the surgical outcome to focus on? Just to provide a bit of context, of 10,000 patients undergoing such procedures, 300-400 patients would be expected to develop endocarditis, 120 would be expected to develop an SSI, and only one of these would be due to chimaera. I certainly don’t want to downplay the importance – and preventability – of this issue, but is there as much discussion on the other 119 causes of SSI?
  • Are the HCUs fit for purpose? The authors talk about ‘systematic decontamination failure’ as a root cause of this issue. However, these machines are being deep cleaned and decontaminated within an inch of their lives with all kinds of chemicals, and yet still they continue to spew mycobacteria. I am beginning to think that the root cause is that these machines are simply not fit for purpose and need a fundamental redesign. Couldn’t we have a closed HCU a la an overgrown PCR block (or some other approach to heating and cooling)?!
  • What do we do about it? There is some useful guidance from PHE about ways to mitigate the risk from these devices, including changing the decontamination approach, and locating the HCUs as far from the patient as possible.
  • Do we need to simplify the microbiology somehow? At a first read, I found the mycological details of this paper confusing (and this is no criticism of the authors by the way): Mycobacterium avium complex, M. intracellulare, and M. chimaera were used more or less interchangeably on a couple of occasions. I suspect that a general reader would struggle to grasp the nuance of the terminology here!
  • Is this the tip of the iceberg? Sticking with the microbiological complexity, this is very likely to be the tip of the iceberg. Many labs don’t have appropriate methods in place to detect mycobacteria efficiently, much less speciate chimaera. The massive time delay between the surgical procedure and the presentation of the case certainly doesn’t help matters.
  • Why is it getting worse? The risk is increasing year on year. The authors attribute this to an aging fleet of HCUs in operation. But I wonder whether this is simply improved case detection?

It seems to me that the best way to solve this problem once and for all (while we wait for a more suitably designed HCU to be developed and tested) is to move the HCUs out of the operating theatre altogether, which has been achieved in some centres, but may not be feasible elsewhere.

Image: FDA.


3 thoughts on “The silent Mycobacterium chimaera epidemic

  1. Spectacular write-up and very timely and relevant Jon. I love your way of thinking–Forcing us to challenge our assumptions and current status quo. Concerning and alarming to me is the continued surge in healthcare surfaces and devices that can’t be adequately disinfected. This problem is further worsened by the tenacious and complex microbes that we now have to manage in our healthcare facilities. There needs to be more solutions aimed at root cause or else endoscopes, HCU, mobile devices, etc is just a tip of the iceberg in un-disinfectable surfaces(not a real word)


  2. Thanks for asking the questions that government agencies can’t. I guess Healthcare as in many other sectors likes to focus on the new and unusual (e.g. Ebola) rather than addressing the regular SSIs that require system and behavior change.


  3. Excellent article, Jon. The HCU issue is likely the tip of the iceberg wrt bioaerosols and high humidity / saturated surfaces in healthcare facilities. Air handling units, ice machines, soiled utility rooms, sink and shower drains, all have similar issues with mold and bacteria growth and known disease transmission. All of these require appropriate engineering measures to ensure patient safety. Technologies such as UV, ozonated water, copper, and aqueous and gaseous reactive oxygen species should be deployed. We need the participation of engineers as well as national and international standards to eliminate biohazards of devices, air, water and surfaces in healthcare facilities.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s