The article reports the frequency of resistance phenotypes in chicken collected at retail level. The title of the article is appropriate.

The abstract is clearly written although the consequences of the findings are not clearly referred to. The findings are not really surprising as production methods are bound to impact on the level of resistance among bacteria such as E. coli. Also, I think the reference being made to consumers may be misleading as it may imply consumer risk. It would be helpful to provide more details on specific husbandry practices used for kosher chicken as readers may not be familiar with them.

Regarding the statistical method; I am not clear what the dependent variable for the ANOVA was. It should typically be a numerical not categorical variable, so I assume it was percentage? I understand that analysis was done by brand. So we have a multi-level clustering here (sample-brand-production system). Data should therefore be analysed in this way.

I assume that resistance was established as a binary variable. Note that it is recommended to move towards more quantitative measurement of resistance. The description of the statistical analysis is too superficial to conclude on validity of results. The number of samples was low.

It has been demonstrated that the extent and type of antimicrobial usage is hugely variable between farms even within one production type (e.g. among conventional producers). It is therefore recommended to use data that allow for linking of resistance status in the product to the true exposure of the animal, i.e. to link retail back to pre-harvest. I know that this is difficult, but else evidence will remain weak.