Year of Publication


Degree Name

Doctor of Philosophy (PhD)

Document Type

Doctoral Dissertation




Veterinary Science

First Advisor

Dr. Peter J. Timoney


Formalized epidemiological field investigations were conducted to compare and contrast the transmission dynamics of EHV-1 neurological disease among horses stabled at Churchill Downs Racetrack, Louisville, Kentucky and of EHV-1 respiratory illness among horses stabled in the student barn at Murray State University. Differences were assessed by means of statistical and mathematical modeling techniques applied to survey and biological data collected over the course of the respective disease events.

Regression methods applied to survey data enabled the construction of a statistical model to predict a date of onset of illness for horses within each equine cohort. Comparisons of the epidemic curves revealed that the Murray State University outbreak was 4.5 times longer (9 weeks versus 14 days) than the Churchill Downs Racetrack event.

Survival analysis was used to explore the relationship between time to infection for each equine cohort. Horses stabled in the affected barn at Churchill Downs racetrack had a 3.02 times greater daily risk (p < 0.001) for contracting EHV-1 infection relative to horses stabled in the student barn at Murray State University.

Estimates of the basic R0 number, calculated using mathematical formulae that incorporated the duration of the infectious period for neuropathogenic and nonneuropathogenic strains of EHV-1, were 10.25 and 2.94 for the Churchill Downs racetrack and Murray State University outbreaks, respectively. The generation time for the Churchill Downs outbreak was 6.1 times shorter (0.39 days versus 2.38 days) than for the Murray State University event. An assessment of the temporal occurrence of symptomatic infection is similar for each event and suggests that the appearance of clinical illness is constant over the course of an outbreak.

A Reed-Frost model was constructed for each EHV-1 event where values of the transmission parameters (q, p and k) were estimated by fitting a model that most closely matched the observed profile of EHV-1 cases. The value of prophylactic vaccination on the spread of EHV-1 was assessed by making adjustments to these fitted models for varying levels of herd immunity. The results indicate that the prevention of EHV-1 neurological illness requires a higher level of herd immunity than EHV-1 respiratory illness.