One of my research interests is engineering control technology, as it applies to controlling air pollution, both indoors and outdoors.  This includes technology that uses filtration, adsorption, or ultraviolet germicidal irradiation (UVGI).  Filtration is used to remove particulate matter from an air stream, adsorption is used to remove gases from an air stream, and UVGI is used to inactivate or damage microorganisms.  Each of these approaches has pluses and minuses, and would be used depending on the control requirements and source of pollution.  A main reason to use control technology is to reduce exposure to air pollution (the other reasons are to reduce visibility and material damage).  We know that controls work. For example when we added controls to cars (i.e. catalytic converters), we significantly reduced emissions of nitrogen oxides (NOx), hydrocarbons (HC), and carbon monoxide (CO), while continuing to increase the number of cars on the road.  Air cleaners in homes are commonly used by sensitive individuals (i.e. asthmatics) to reduce levels of particles and gases that could trigger a disease episode.  A recent paper shows that use of air cleaners in homes decreased indoor PM2.5 and improved lung function.  Currently, we are pushing the diesel engine controls for particulate matter since these engines emit so many fine particles, which we know causes major health effects.

I also get asked by friends and colleagues (and people who just find me via the internet) to recommend a technology that they could use in their home to improve their indoor air quality.  Routinely I recommend technology that has been proven time and time again to work effectively for the application.  I stick with what the research shows as effective.  Which gets to a main point of this blog.  There are many companies that have in-house testing reports that they will send you on request or that you can download off of the internet that shows their technology works.  The problem with relying on these reports is that, while they may have been done by an independent lab, they most likely have not been peer-reviewed by additional experts in the field.  Peer review is important for many reasons.  For example, helpful suggestions can be given to the researchers to improve their analysis and reporting.  Also, reviewers will look at the experimental design critically and can provide feedback on whether enough replications were completed, or ask about the error analysis, or inquire about how an assay or instruments were used to collect the data.  If the research is not up to the accepted standards of the journal and/or research community, the publication can be rejected.

A while back one company hired me as a consultant to review an in-house report on their filtration technology.  While the data seemed to show that there was indeed a substantial reduction in particulate concentrations, in fact the time over which the reduction occurred was so long that it really was not a meaningful technology to use in any space except a really small closet.  Or, another technology we studied for removing particles ended up generating ozone, which is definitely a problem in a building, especially if it accumulates and individuals are exposed.

It can cost alot to fund an independent study of a control technology.  For example, I’ve done independent research funded by both government agencies and companies on products such as UVGI light fixtures and filtration units for both in-room and ceiling-mounted applications.  These studies took 2-3 years and resulted in a published paper that showed how effective these devices were for specific applications.  But in the end this investment pays off because these studies were peer reviewed and the results are available for everyone to assess and use.  Similarly, other research groups publish their work – and then the compilation of all the papers lends weight to the control technology and helps us understand how to best apply it.

If a technology has not been published in peer-reviewed literature, I usually do not use or recommend until it has.  You probably shouldn’t either.