Document Server@UHasselt >
Research >
Research publications >

Please use this identifier to cite or link to this item: http://hdl.handle.net/1942/9531

Title: Salmonella serosurveillance: Different statistical methods to categorise pig herds based on serological data
Ogunsanya, V.
Van der Stede, Y.
Issue Date: 2009
Publisher: Elsevier B.V.
Citation: PREVENTIVE VETERINARY MEDICINE, 89(1-2). p. 59-66
Abstract: This study proposes three different statistical methods that can be applied in order to categorise pig herds into two groups (high seroreactors vs. low seroreactors) based on serological test results for Salmonella-specific antibodies in pigs. All proposed statistical methods were restricted to allocate about 10% of the herds into the group defined by each of the statistical approaches as high seroreactors. Previously, semi-parametric quantile regression has been used for this purpose, and here we compare it with two other alternatives: a naive method (based on the mean values) and another based on activity region finder methodology in combination with random forest regression models. The serological response values (the sample-to-positive ratio (S/P ratio)) of 13 649 pigs from 314 Belgian pig herds were used for this comparison. Nearly 14% of these herds were assigned to the high-seroreactor-herd group by at least one of these three methods. The corrected level of agreement was calculated together with the pair-wise agreement among all three methods in order to classify herds as high- or low-level seroreactors, resulting in an agreement level greater than 92%. The results obtained from a fourth method, which is adopted by the Belgian Federal Agency for the Safety of the Food Chain (FASFC), were also compared to the previous three methods. The methods were compared in terms of their agreement as well as their advantages and disadvantages. Recommendations for each applied method are presented in relation to the objectives and the requisite policy for classifying pig herds based on serological data.
URI: http://hdl.handle.net/1942/9531
DOI: 10.1016/j.prevetmed.2009.01.009
ISI #: 000265570800008
ISSN: 0167-5877
Category: A1
Type: Journal Contribution
Validation: ecoom, 2010
Appears in Collections: Research publications

Files in This Item:

Description SizeFormat
N/A329.85 kBAdobe PDF

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.