Reader Comments

Post a new comment on this article

Positive control for manikin system?

Posted by lkobzik on 13 Sep 2014 at 10:45 GMT

A very nice article, and refreshingly good news in the influenza field.
I wonder if the authors considered some kind of positive control for their manikin system, i.e., what level of aerosolized virus, generated artificially rather than by the more relevant human volunteers, can be detected via the manikin? The worst case scenario would be to discover that the system failed to detect aerosols with log orders more virus than delivered by the volunteers, and this would raise questions about the findings. This outcome is unlikely but it would be good to exclude.

No competing interests declared.

RE: Positive control for manikin system?

JWTang49 replied to lkobzik on 29 Sep 2014 at 20:21 GMT

We did think about this initially, but there are many studies on detectable aerosols produced, artificially, from various types of mechanical nebulisers - often with mono-dispersed droplet aerosols (i.e. droplets of one size at a time) using simulated saliva with vaccine viruses - in which case one could also argue that these are not realistic positive controls for human sources with naturally occurring influenza viruses, either.

We can consider doing this, but then the question becomes, which virus should we use? Vaccine or naturally occurring? What solution composition should we use to carry the virus? How can we make it of the same viscosity as saliva - and include the local variability in viscosity that I mention in the Discussion section of the paper related to effects of antiviral salivary mucins? This may well affect the viral load expelled in the aerosol droplets produced by these human volunteers that may not be easily reproduced in an artificial nebulizer system.

And what viral concentration should we use in the solution - bearing in mind now that we suspect (and this is also borne out by some other studies including Milton et al. 2013's study cited in the paper: http://www.plospathogens.... and Nishimura et al. 2014: http://www.ncbi.nlm.nih.g...) the droplets expelled in the aerosol do not contain a corresponding amount of virus in that droplet volume that would be expected given the diagnostic viral load from the swab? The how should we aerosolize this artificial saliva solution to mimic a human cough or breath realistically?

It is becoming clearer that these questions may affect the overall realism and value of the 'positive' control - in fact, the positive control, if not set up properly, might actually be misleading in that it may produce expectations that may not be realistic when using naturally-infected human volunteers performing natural human respiratory activities.

I am not trying to avoid the issue here, but given the subtleties of the experimental findings reported here and in the Milton et al. 2013 and Nishimura et al. 2014, I would be happy to hear this reader's suggestions for how to set up a realistic positive control for this study?

Julian W Tang (Corresponding author)

No competing interests declared.