Considerations regarding appropriate sample size for conducting ferret transmission experiments.
Belser JA, Maines TR, Katz JM, Tumpey TM.
Future Microbiol. 2013 Aug;8(8):961-5. doi: 10.2217/fmb.13.64.
This article addresses the concerns raised by Nishiuria et al. (2013).
PLoS ONE 8(1): e55358. doi:10.1371/journal.pone.0055358
“We agree with these authors that the generally small sample size inherent in conducting ferret transmission research limits some statistical power, and concur that an n = 3 when comparing two groups, in the absence of other supporting data, is often not sufficient to demonstrate significant or meaningful differences in two-sample comparisons.”
“We agree with the premise of the study conducted by Nishiura et al. that sample size is one of many important determinants in achieving an experimental design that yields accurate, repeatable and meaningful results. As differences in experimental protocols, equipment and reagents between laboratories limit our ability to directly compare differences in virus transmissibility between published studies, it is critical for individual studies to contain sufficient data (including but not limited to the number of animals utilized) to reach scientifically sound conclusions. However, by using fewer ferrets to test one particular virus or condition, this allows researchers to perform a greater total number of experiments in a given study. As a result, generalizations can be drawn between fundamental virological properties (such as virus subtypes, or viruses that exhibit antiviral sensitivity versus resistance) with greater confidence that the results are not inadvertently biased due to unintended strain-specific differences or other variances.”
We learn that is it very difficult to “compare differences in virus transmissibility between published studies”. By using fewer animals per experiment “generalizations can be drawn between fundamental virological properties (such as virus subtypes, or viruses that exhibit antiviral sensitivity versus resistance)”. This is far from the initial promises of GOF influenza research agenda.
