Fig 1.
Schematic diagram of an endurance event.
The competition is divided into phases of 30–40 km that are followed by 40- to 50-minute rest periods. Horses are checked by veterinarians at the vet gates (VGs) between each phase. The HR must be below 65 bpm when the horse is examined at the vet gate. The cardiac recovery time (CRT) is defined as time interval between the end of the phase and presentation at the vet gate; it counts towards the total running time.
Table 1.
Distribution of horses according to the outcome of the event and the reason for elimination.
Table 2.
The AS, CRT and HR as a function of the outcome of the event, for different distance categories.
Fig 2.
Distribution of the eliminated horses as a function of the cause of elimination, the vet gate and the event distance.
1*: 80–119 km, 2*: 120–139 km, 3*: 140–160 km, LA: lameness, ME: metabolic reasons, RET: retirement, OR: other reasons. # indicates a significantly greater elimination rate for all distance categories.
Table 3.
Results of the logistic regressions used to estimate the probability of elimination at each vet gate (based on variables recorded at the previous vet gate).
Fig 3.
ROC curves of the logistic regressions for vet gates 2 and 3.
The AUROC values (0.72 and 0.74, respectively) were acceptable and were associated with correct classification rates of 64% and 68%, respectively (AUROC = area under the ROC curve, Q = qualified, EL = eliminated).
Table 4.
Validation of the predictive logistic regressions by applying an independent dataset (including 80 horses having competed in endurance events in 2014).
By applying a 70% probability of elimination in the regression calculations, the correct prediction rate was 75%. There were relatively few false negative predictions (8.70%), whereas the false positive rate was higher (16.30%).
Fig 4.
Three-dimensional scatter plots showing the probability of elimination at vet gates 2, 3, 4 and 5, according to the corresponding logistic regressions with a fixed HR of 64 and the AS and CRT measured at the previous vet gate (n-1).
Red corresponds to a probability of elimination of 60 to 80%, whereas brown (the darkest areas) corresponds to a probability of 80 to 100%. The white line corresponds to a probability of elimination of 70% (the threshold chosen to compute the probability of elimination in an independent data set used for validation).