Evaluation of the thoraco-laryngeal reflex (‘slap test’) as an aid to the diagnosis of cervical spinal cord and brainstem disease in horses.

This research evaluates the ‘slap test’, a diagnostic method for identifying cervical spinal cord and brainstem diseases in horses, by using endoscopic evaluation and comparing it with conventional neurological examination. The study was conducted over a year and had 15 ataxic horses assessed, along with 13 healthy horses for control. Despite statistical significance, the slap test showed low sensitivity but high specificity. The conventional neurological examination was more effective.

Study Design and Procedure

• The study was conducted over a year, with a test group of 15 ataxic horses and a control group of 13 healthy horses.
• Test horses were subjected to the ‘slap test’ — a diagnostic method believed to identify neurological issues in horses.
• Post-mortem examination of the horses was then conducted to confirm the presence of histopathological lesions.
• Responses of all horses were recorded and subsequently reviewed by three independent assessors.

Study Results

• Out of the 15 test horses, 12 were found to have histopathological lesions confirming their clinical signs.
• The correlation between the diagnosis of cervical spinal cord and/or brainstem disease and the response to the ‘slap test’ was statistically equivalent according to the McNemar chi-Square test.
• However, the ‘slap test’ showed low sensitivity (50% for the left side and 58% for the right side).
• The specificity of the ‘slap test’ was higher (around 70% on average).

Comparative Analysis

• Conventional neurological testing of horses showed 100% sensitivity and 81% specificity, proving to be more effective than the ‘slap test’ in the detection of lesions.
• The consistency between each assessor’s own slap test scores on different days was good (kappa values 0.59 to 0.85).
• However, agreement among different assessors on slap test scores was poor (kappa value 0.35), indicating that the slap test might not provide uniform results across different evaluators.