Publication
Modelling the effect of chilling on the occurrence of Salmonella on pig carcasses at study, abattoir and batch levels by meta-analysis
| dc.contributor.author | Gonzales-Barron, Ursula | |
| dc.contributor.author | Cadavez, Vasco | |
| dc.contributor.author | Sheridan, James | |
| dc.contributor.author | Butler, Francis | |
| dc.date.accessioned | 2014-10-31T17:23:07Z | |
| dc.date.available | 2014-10-31T17:23:07Z | |
| dc.date.issued | 2013 | |
| dc.description.abstract | The effect of chilling on the occurrence of Salmonella on pig carcasses was investigated at study, abattoir and batch level by meta-analysis. Both the fixed-effects and random-effects model confirmed (p b 0.05) the significant effect of chilling in decreasing Salmonella occurrence on pig carcasses; although the random-effects solution was preferred to account for the significant variability in effect size (p b 0.001) estimated from the 13 primary studies considered, the 32 abattoirs surveyed, and the 51 sampled batches. Conservatively, it can be said that chilling reduces the Salmonella incidence on pig carcasses by a mean ratio of ~1.6 (95% CI: 1.0–2.6). Multilevel meta-analysismodels investigating study characteristics that could explain the heterogeneity (τ2) in the true effect size among primary studies (τ2 = 0.578), among surveyed abattoirs (τ2 = 0.431), and among sampled batches (τ2 = 0.373), revealed that study size (represented by the moderating variables of ‘total sample size’ and ‘number of batches sampled in an abattoir’) and ‘carcass swabbed area’ have a significant impact (p b 0.05) on the measured effect size of chilling. The fact that swabbed area explained between 56 and 62% and total sample size between 23 and 38% of the total heterogeneity in the chilling true effect size, indicates that differences in experimental design greatly affect our substantive conclusion about the effect of chilling on Salmonella occurrence. This inconsistency to elucidate the effect of chilling arises because of themany factors influencing both the performance of the chilling operation and themeasurement itself.Meta-analysis was not only instrumental to showthat small-size studies (i.e., only one batch sampled per abattoir, total number of sampled carcasses per batch b 50) and small swabbed areas (b100 cm2) lead to imprecise and even conflicting conclusions, but most importantly, enabled definition of the characteristics of a well-designed study having aminimumstatistical power to produce precise results. A sound experimental design derived by multilevel meta-analysis consists of swabbing carcass areas of at least 500 cm2 from 25 pre-chill and 25 post-chill carcasses from a single production batch, with a minimum of two batches sampled per surveyed abattoir. If the survey were to be conducted in more than one abattoir, the total sample size should not be less than 400. Two methods to test for publication bias, a common problem in meta-analysis, suggested that whilst the presence of unpublished small-size studies is probable, it is not likely that this would significantly bias the overall chilling effect estimated in this study. | por |
| dc.identifier.citation | Gonzales-Barron, Ursula A.; Cadavez, Vasco; Sheridan, James; Butler, Francis (2013). Modelling the effect of chilling on the occurrence of Salmonella on pig carcasses at study, abattoir and batch levels by meta-analysis. International Journal of Food Microbiology. ISSN 0168-1605. 163:2-3, p. 101-113 | por |
| dc.identifier.doi | 10.1016/j.ijfoodmicro.2013.03.006 | |
| dc.identifier.issn | 0168-1605 | |
| dc.identifier.uri | http://hdl.handle.net/10198/11307 | |
| dc.language.iso | eng | por |
| dc.peerreviewed | yes | por |
| dc.publisher | Elsevier | por |
| dc.subject | Meta-analysis | por |
| dc.subject | Multilevel model | por |
| dc.subject | Chilling | por |
| dc.subject | Risk | por |
| dc.subject | Pig | por |
| dc.subject | Slaughterhouse | por |
| dc.title | Modelling the effect of chilling on the occurrence of Salmonella on pig carcasses at study, abattoir and batch levels by meta-analysis | por |
| dc.type | journal article | |
| dspace.entity.type | Publication | |
| oaire.citation.endPage | 113 | por |
| oaire.citation.startPage | 101 | por |
| oaire.citation.title | International Journal of Food Microbiology | por |
| oaire.citation.volume | 163(2-3) | por |
| person.familyName | Gonzales-Barron | |
| person.familyName | Cadavez | |
| person.givenName | Ursula | |
| person.givenName | Vasco | |
| person.identifier | R-000-HDG | |
| person.identifier.ciencia-id | 0813-C319-B62A | |
| person.identifier.ciencia-id | 441B-01AB-A12E | |
| person.identifier.orcid | 0000-0002-8462-9775 | |
| person.identifier.orcid | 0000-0002-3077-7414 | |
| person.identifier.rid | A-3958-2010 | |
| person.identifier.scopus-author-id | 9435483700 | |
| person.identifier.scopus-author-id | 9039121900 | |
| rcaap.rights | openAccess | por |
| rcaap.type | article | por |
| relation.isAuthorOfPublication | 17c6b98f-4fb5-41d3-839a-6f77ec70021a | |
| relation.isAuthorOfPublication | 57b410e9-f6b7-42ff-ab3d-b526278715eb | |
| relation.isAuthorOfPublication.latestForDiscovery | 57b410e9-f6b7-42ff-ab3d-b526278715eb |