Non-horsey people try the Horse Grimace Scale: Here are the results

The Horse Grimace Scale, developed seven years ago. Image: Dalla Costa, et al. PLOS ONE,
The Horse Grimace Scale, developed seven years ago. Image: Dalla Costa, et al. PLOS ONE,

The Horse Grimace Scale, introduced to the world in 2013, has been put to the test among a group of non-horsey people to see if they could come to grips with it after 30 minutes of training.

The researchers had wanted to find out if they could be taught to successfully apply it by an experienced user in a single session using theory and practical examples.

Its developers had been encouraged by research published in 2018 which revealed that veterinary students were reliably able to apply the Horse Grimace Scale, a system designed to discern pain in equines, even without specific training prior to its use.

Researchers now report in the journal Animals on a fresh study in which 206 undergraduate students from five institutions tried to apply the Horse Grimace Scale after receiving a 30-minute face-to-face training session from an expert in the scoring system.

None of the students had direct experience with horses and all were unfamiliar with the Horse Grimace Scale.

Before training, the students received a brief lecture on the definition of pain and its effect on facial expressions in different species (for example, mice, rats, rabbits) but not horses.

They were then asked to score 10 high-quality images of horses suffering pain from acute laminitis without any knowledge of the Horse Grimace Scale.

They then received their face-to-face training in a class setting on the scale. They were shown example pictures and were told how the scale worked. They were encouraged to interact with the trainer, ask questions and actively discuss the method and the scoring of example pictures.

The system relies on observers identifying the presence or absence of what are described as six facial action units, as being responses to pain:

  • Stiffly backward ears;
  • Orbital tightening;
  • Tension above the eye area;
  • Strained chewing muscles;
  • Strained mouth and pronounced chin;
  • Strained nostrils, and flattening of the profile.

Observers score zero if they consider the facial action unit is not present, 1 if they consider it moderately present, and 2 if they consider it is obviously present. This gives any one horse a maximum rating of 12 on the scale. The higher the score, the greater the level of discomfort being experienced by the horse.

They were then shown 10 further good quality images — again of horses in pain due to laminitis — and were asked to apply the Horse Grimace Scale.

The students grasped important elements of the system, but the researchers concluded that the 30-minute face-to-face training session was not sufficient to allow observers without horse experience to effectively apply the scale.

“However, this standardised training program could represent a starting point for a more comprehensive training program for those without horse experience in order to increase their reliability in applying the Horse Grimace Scale,” Emanuela Dalla Costa and her fellow researchers said.

Discussing their findings, the study team said the results showed a high variability of agreement between naïve observers and the expert for the different facial action units comprising the scale.

“Only stiffly backwards ears and orbital tightening reached a substantial agreement before training, while all other facial action units only showed slight agreement or fair agreement.

“Following training, the agreement for stiffly backwards ears and orbital tightening significantly increased … indicating near-perfect agreement.”

The agreement for prominent strained chewing muscles significantly increased, but indicated only a fair agreement. For the other facial action units, no significant changes were observed from before and after training.

The researchers said more needed to be done to design a training protocol for the Horse Grimace Scale, which could be applied to prepare new assessors without horse experience to ensure reliable assessment of horses and pain.

A dedicated picture collection composed of high-quality and uniform pictures, and a more extensive training program involving a lower number of observers per trainer, may be necessary, they suggest.

“Finally, a session in which observers can practice scoring live animals seems fundamental for improving the accuracy of in-field pain evaluation.”

The study team comprised Dalla Costa, Michela Minero and Francesca Dai, with the University of Milan in Italy; Mattew Leach, with Newcastle University in England; and Amelia Mari MacRae, with the University of British Columbia.

Dai, F.; Leach, M.; MacRae, A.M.; Minero, M.; Costa, E.D. Does Thirty-Minute Standardised Training Improve the Inter-Observer Reliability of the Horse Grimace Scale (HGS)? A Case Study. Animals 2020, 10, 781.

The study, published under a Creative Commons License, can be read here

The original PLOS ONE paper on the Horse Grimace Scale can be read here

Leave a Reply

Your email address will not be published. Required fields are marked *