An alternative framework to conduct inferential statistics for low microbial counts in foods: the poisson-gamma regression
Conference Paper
Overview
Research
Additional Document Info
View All
Overview
abstract
The objective of this article was to compare four Poisson-gamma regression models to assess
the effect of chilling on the concentration of coliforms from beef carcasses. A total of 600
carcasses were sampled before and after chilling at eight large Irish abattoirs, and the total
coliforms were determined. With a coded variable (pre-chill/post-chill) as treatment, and
extracting the variability of batches nested in abattoirs, random-effects models confirmed that
chilling had a decreasing effect on the overall recovery of coliforms. Furthermore, the
expected coliforms concentrations on pre-chill and post-chill carcasses were estimated on a
CFU/cm2 scale, as well as their between-batch variability. This study introduced an
alternative conceptual framework that can find interesting applications in stochastic risk
assessment and in the design of more efficient sampling plans.
