Skip to main content

Table 3 Final multivariable logistic regression models describing the likelihood of a teat having a roughened or thickened teat-end callosity ring, respectively [22]

From: Milk-flow data collected routinely in an automatic milking system: an alternative to milking-time testing in the management of teat-end condition?

Outcome and BIC

Variable

Odds ratio

P

95% CI

Lower

Upper

ROUGHNESS

DIM

1.016

0.032

1.001

1.030

BIC = 290.59

Parity 1 (reference)

Parity 2

0.291

0.201

0.044

1.928

Parity ≥ 3

2.943

0.073

0.903

9.593

AVGFLOW, kg/min

0.020

0.001

0.003

0.160

THICKNESS

DIM

1.024

0.006

1.007

1.041

BIC = 281.38

Parity 1 (reference)

Parity 2

0.386

0.381

0.046

3.249

Parity ≥ 3

3.969

0.056

0.963

16.362

AVGFLOW, kg/min

0.019

0.001

0.002

0.181

  1. The models were selected based on having the lowest Bayesian information criterion (BIC) among other models for the same outcome
  2. DIM, days in milk; AVGFLOW, quarter average milk flow rate
  3. ROUGHNESS, dichotomized outcome variable where smooth teat-end callosity rings form the comparison group and teat-ends with a roughened callosity ring is considered abnormal
  4. THICKNESS, dichotomized outcome variable where teat-ends having a thin or not visible teat-end callosity ring form the comparison group, and medium, thick or extreme are considered abnormal