Monday, February 20, 2012

Jack Gilbert @gilbertjacka clarifies comments at #AAASMtg re: opening windows, cleanliness & microbes

Just got this to post from my college colleague Jack Gilbert in regard to some comments he made to the press at the AAAS Meeting a few days ago.  He sent this in part in response to some news stories that came out of the press conference we had at the meeting (e.g., see Florence Nightingale approach 'could help fight infection in ...Open hospital windows to stem spread of infections, says ... and others).  And I encouraged him to consider whether or not he needed to clarify some of his comments - and here is his response.

A confession, lesson and retraction
By Jack A. Gilbert

At the AAAS 2012 annual meeting on Friday, I was involved in a press conference to announce the initial results and ideology for the Earth Microbiome Project ( Following the press conference we went to another room, where we were openly discussing these concepts with the reporters. Several reporters asked me to comment on the potential impact of this research in the medical sphere. At which point I started to discuss some excellent research by Jessica Green ( regarding her recent evidence that improved ventilation in hospital wards reduced the airborne abundance of organisms that were related to pathogens. I showed these reporters the paper
( and asked them to discuss this with Jessica Green.

I then proceeded to discuss some current research we are doing at University of Chicago that is looking at the impact of having a natural microbial community on surfaces to reduce the likelihood that pathogens can establish in that environment. Specifically we are exploring whether 'good bacteria' can be used as a barrier to outcompete 'bad bacteria', I suggested that this was testing the hypothesis outlined by Florence Nightingale.

To this end I said, that maybe instead of sterilizing every surface in a hospital we could explore a different strategy. There is however currently only circumstantial evidence to support my claims, and I could have done a much better job in making clear that I was discussing an idea - not something for which there was evidence.

I am sorry for my indiscretion and hyperbole, and hope that I didn't cause any groups or individuals concern or worry about this topic. These were concepts being discussed, specifically that by using the EMP we could explore ecological dynamics that could lay the groundwork to help determine if a community could play the role of a barrier against infection.

I want to stipulate that I believe hospitals should be cleaning, and I believe that surgeons should scrub and use the sterile method. To be clear, I wanted to state that 'good' bacteria could in the future play a role in reducing the instances of hospital borne infection, and that this is something we should investigate. People should wash their hands after the toilet, and wash their hands when they are sick; there is nothing wrong with being clean.


  1. There are good bacteria that present a zero chance of infection. The commensal bacteria that I am working with, autotrophic ammonia oxidizing bacteria are incapable of growth on any media used to isolate pathogens (they are obligate autotrophs).

    They make NO and nitrite, so they could be used to suppress quorum sensing and expression of virulence factors by pathogens (such as biofilm formation).

    I think they are an agent of the hygiene hypothesis and that removing them through modern bathing does us no good at all.

    Opening windows is a little over the top unless you need maggot therapy. But maggot therapy is also good if you use the right kind of maggots, but I think medical-grade is better than wild-type. ;)

  2. I would like to comment on Dr. Gilbert's characterization of the Green study as "...evidence that improved ventilation in hospital wards...." The study did not evaluate ventilation quality by the usual metrics of outdoor air exchange rate or room air distribution performance. Ventilation can be evaluated in several ways, and measuring the composition of indoor air is only one type of metric.

    In fact, what the U of Oregon study did was to compare the microbiology measured in the study's patient rooms supplied with air through "window ventilation" to that found in patient rooms supplied with air through ventilation provided by the hospital's mechanical ventilation system.

    In the present context, it is worth noting that, as a result of the study's methodology for comparing the two types of ventilation, the main difference would be that the air provided by the mechanical system would have also come from outdoors but would have passed through the system's fan, filters, cooling and heating coils, and ductwork.

    There is extensive evidence in the building science literature of contamination inside ventilation system components including those in the air handling units and the air distribution duct work itself. So an appropriate extension of Dr. Gilbert's comments about hospital cleaning would be that it might also prove beneficial to improve the hygiene of hospital ventilation equipment.

    It is evident that mechanical ventilation systems alter the composition of outdoor air used for ventilation. Some of this alteration is most likely beneficial in terms of removal of the larger particles that most mechanical ventilation systems' filters typically do remove. The range of bacteria sizesruns from larger particle sizes (1 - 5 microns or more) that are efficiently removed by mechanical system filters down to very small particle sizes (<1 micron) of which typical filters allow a significant fraction to pass through the filters.

    It should be noted that when (bacterial or fungal) particles accumulate on filter media surfaces, they are not necessarily inactivated or permanently attached. Moisture and temperature conditions will affect their survival and reproductive activity. Stopping and starting system fans is known to cause sudden bursts of particles off of filter surfaces and into the supply air stream.

    There are also reports in the literature of growth of bacteria on mechanical ventilation system cooling coils, and some of these same bacteria have been found on cooling coils in automobile air conditioning units.

    Building scientists and epidemiologists have fairly consistently found higher prevalence rates of "building-related symptoms" (BRS) and sick building syndrome (SBS) for more than 25 years in buildings with mechanical ventilation systems and especially in those with air-conditioning compared to building with natural ventilation. This is surprising because, in most of these studies, outdoor air ventilation rates were higher in the mechanically-ventilated buildings. That finding conflicts with another fairly consistent finding that higher ventilation rates results in lower rates of a variety of adverse health outcomes (Sundell et al, 2011. Indoor Air 21[3] 191–204.)

    What would be interesting as a follow-up to the U of Oregon study would be to compare the microbial communities upstream and downstream of various components of the mechanical ventilation system's equipment, especially the filters and the coils, two locations where very different conditions exist that are likely to affect the microbial populations differently. This might help improve the quality of air and reduce the populations of unwanted bacteria in occupied spaces ventilated by mechanical systems.


Most recent post

Another day to think, to pause, to ponder.

Panorama of Sycamore Park and the memorial to Karim   A bit over 10 years ago I wrote a blog post that I repost all the time. Entitled "...