An evaluation of early intervention protocols based on camera-based autonomous mobility score trends

Objectives

The objective of this project was to determine if changes in camera-based autonomous mobility scores could be used to reduce the duration of lameness.

Materials and methods

Starting in July 2021, a 2300 cow herd was scored continuously using camera-based autonomous mobility scoring (CattleEye). The camera-based mobility scoring (CMS) outputs a score of 0 to 100, where 0 indicates no mobility abnormalities and 100 indicates severe mobility abnormalities. A score above 49 is considered lame. At the same time, the herd’s hoof trimmers were trained in nomenclature for the recording of lesions and their hoof trimming technique was standardized. Starting in December 2021, cows with an increasing CMS trend over the previous 14-day period were enrolled in the study on a weekly basis. This increasing trend was determined based on a linear model of the CMS that gave each cow a trend score. Cows with trend scores above 20 were eligible for enrollment. To ensure only new cases of lameness were enrolled only cows with a CMS below 50 in 14-28 days before enrollment were eligible for enrollment. Additional enrolment criteria included cows being greater than 14 days in milk, not being coded as ‘do not breed’, and not having a concurrent illness. These data were retrieved from the farm’s on farm management software. Eligible cows were scored by trend score from highest to lowest and odd numbered cows were placed on the hoof trimming list (TEST). Even numbered cows were enrolled in the study but not placed on the hoof trimming list and were considered our CONTROL group. Due to hoof trimming capacity, between 5-12 cows with the highest trend scores were placed on the farm’s hoof trimming list and an equal number of CONTROL were enrolled. Our outcome of interest was percentage of days cows spent about their maximum score during the 14-day pre-enrolment window. Due to ongoing enrolment, our initial analysis only included a 14-day post enrolment window. Data was analyzed by comparing means and using a linear model that included, treatment, lactation group, days in milk, trend score and max score at enrollment as confounders. [...]