Using herd frailty estimates from survival models in a mortality-based syndromic surveillance system.

Area of Science:

• Veterinary Epidemiology
• Animal Health Surveillance
• Disease Eradication Programs

Background:

• Syndromic surveillance offers a complementary approach to diagnostic testing for comprehensive disease monitoring.
• Bovine viral diarrhoea (BVD) eradication programs, like Ireland's initiated in 2013, require robust strategies for detecting potential re-emergence.
• Mortality data, when consistently recorded and traceable, can serve as valuable indicators in syndromic surveillance systems for cattle.

Purpose of the Study:

• To develop and evaluate a syndromic surveillance model utilizing calf mortality data for early detection of BVD re-emergence in Ireland.
• To assess the utility of historical BVD eradication program data for building a mortality-based syndromic surveillance system.
• To determine the effectiveness of calf mortality patterns in identifying herds at risk of BVD re-emergence.

Main Methods:

• Mixed-effects Cox proportional hazards models were constructed using calf mortality data (up to 100 days of age) from 2014 to 2023.
• Herd-level frailty estimates were extracted annually and clustered to identify distinct temporal mortality hazard patterns.
• Four thresholds were applied to flag herds with elevated calf mortality, and their specificity and sensitivity for BVD status were evaluated.

Main Results:

• The developed syndromic surveillance model demonstrated high specificity (86-92%) but low sensitivity (11-22%) in detecting herd-level BVD status.
• Clustering of herd-level mortality hazard revealed distinct temporal patterns across different herd subgroups.
• The model's low sensitivity indicates that mortality-based syndromic surveillance alone is insufficient for reliably detecting BVD re-emergence.

Conclusions:

• Mortality-based syndromic surveillance, while having limitations in sensitivity, can potentially support Ireland's efforts to achieve and maintain BVD-free status.
• This methodology provides valuable insights for broader animal health surveillance initiatives and is adaptable to other species and diseases.
• Complementary surveillance strategies are necessary to reliably detect BVD re-emergence, alongside the evaluated mortality-based model.