Investigating the uses of machine learning algorithms to inform risk factor analyses: the example of avian infectious bronchitis virus (IBV) in boiler chickens

Authors: CAMPLER, MAGNUS - The Ohio State University; CHENG, TING-YU - The Ohio State University; Lee, Chang; HOFACRE, CHARLES - Southern Poultry Research, Inc; LOSSIE, GEOFFREY - Southern Poultry Research, Inc; SILVA, GUSTAVO - Purdue University; EL-GAZZAR, MOHAMED - Iowa State University; ARRUDA, ANDREIA - The Ohio State University

Submitted to: Research in Veterinary Science
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 2/24/2024
Publication Date: 2/29/2024
Citation: Campler, M.R., Cheng, T., Lee, C.W., Hofacre, C.L., Lossie, G., Silva, G.S., El-Gazzar, M.M., Arruda, A.G. 2024. Investigating the uses of machine learning algorithms to inform risk factor analyses: the example of avian infectious bronchitis virus (IBV) in boiler chickens. Research in Veterinary Science. 171(2024):105201. https://doi.org/10.1016/j.rvsc.2024.105201.
DOI: https://doi.org/10.1016/j.rvsc.2024.105201

Interpretive Summary: Infectious bronchitis is an economically important contagious disease in poultry caused by coronavirus, infectious bronchitis virus (IBV). Despite the importance of understanding potential geospatial influence in the assessment of sources of infections and transmission of IBV in the field, study has been limited to routine monitoring of IBV antibody levels in birds in relation to vaccination and potential virus infection. The main objective of this study was to retrospectively analyze the effect of geospatial factors on IBV antibody titers in broiler chickens. IBV antibody monitoring data collected between 2016-2021 were obtained from 166 commercial broiler farms. Geospatial data of farms was obtained from publicly available sources and mapped using a Geographic Information System. Geospatial factors were screened and analyzed using multiple algorithm and statistical analysis and significant associations between increased odds of high IBV antibody level and 1) specific years, 2) average age of sampled birds, and 3) distance from commercial layer farms. Although our results suggest a limited impact of geospatial factors on IBV antibody levels in broiler farms, this study demonstrated that publicly available datasets combined with the use of a variety of epidemiological tools can be successfully used with routinely collected animal health-related data for the investigation of disease prevention and timely control.

Technical Abstract: Infectious bronchitis virus (IBV) is a contagious coronavirus causing respiratory and urogenital disease in chickens and is responsible for significant economic losses for both the broiler and table egg layer industries. Despite IBV being regularly monitored using standard epidemiologic surveillance practices, knowledge and evidence of risk factors associated with IBV transmission remain limited. The study objective was to compare risk factor modeling outcomes between a traditional stepwise variable selection approach and a machine learning-based random forest Boruta algorithm using routinely collected IBV antibody titer data from broiler flocks. IBV antibody sampling events (n = 1111) from 166 broiler sites between 2016 and 2021 were accessed. Ninety-two geospatial-related and poultry-density variables were obtained using a geographic information system and data sets from publicly available sources. Seventeen and 27 candidate variables were screened to potentially have an association with elevated IBV antibody titers according to the manual selection and machine learning algorithm, respectively. Selected variables from both methods were further investigated by construction of multivariable generalized mixed logistic regression models. Six variables were shortlisted by both screening methods, which included year, distance to urban areas, main roads, landcover, density of layer sites and year, however, final models for both approaches only shared year as an important predictor. Despite limited significance of clinical outcomes, this work showcases the potential of a novel explorative modeling approach in combination with often unutilized resources such as publicly available geospatial data, surveillance health data and machine learning as potential supplementary tools to investigate risk factors related to infectious diseases.