Informatics and the Electronic Medical Record for Syndromic Surveillance in Companion Animals: Development, Application and Utility
Date
2013-01-23
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
There is growing interest in companion animal surveillance to provide an early warning for emerging environmental health risks including zoonotic diseases and chemical contamination of food, water and air. The objectives of this thesis were to plan, implement and evaluate a companion animal surveillance system capable of using existing clinical data to detect emerging environmental risks to people and/or pets. Twelve companion animal practices participated in a sentinel veterinary practice syndromic surveillance network. Enteric diseases were targeted due to their anticipated prevalence and relevance to animal and public health. Customized data extraction software automatically extracted the required fields from the veterinary practice management software at each practice and exported them to a secure data warehouse (n = 447,388 records from January 1, 2007 to December 31, 2010; extraction 99.4% accurate). The records were in free-text with no diagnostic codes or standardized nomenclature. A categorization dictionary developed using commercially available text-mining software automatically classified and retrieved cases of enteric syndrome (n = 18,832; sensitivity, 87.6% and specificity, 99.3%). Using this data it was possible to identify clusters of enteric disease in space and time, describe these patterns by the host factors and observe patterns of antimicrobial use (AMU). There was often not enough specific information recorded in the medical record to describe the clusters by the proportion of vaccinated cases, the probable etiological agent or by environmental factors that may be predictive of increased risk. The lack of contextual information in the medical records limited the usefulness of this system to provide an early warning of environmental hazards. The system could be used to monitor temporal trends in AMU or conduct analytical studies to explore patterns of AMU and concomitant increases in antimicrobial resistance. The system may be useful for studies that support evidence-based veterinary medicine. Future companion animal surveillance systems should consider imposing structured clinical reports onto the submissions from the participating veterinarians. This would address many of the data and system quality issues encountered in this study, but may have an impact on the willingness of veterinarians to participate in community-based surveillance.
Description
Keywords
Bioinformatics, Veterinary Science, Epidemiology
Citation
Anholt, R. (2013). Informatics and the Electronic Medical Record for Syndromic Surveillance in Companion Animals: Development, Application and Utility (Doctoral thesis, University of Calgary, Calgary, Canada). Retrieved from https://prism.ucalgary.ca. doi:10.11575/PRISM/26525