Fig 1.
Descriptive statistics for enrolled farms.
(A) Dry cows’ severe lameness prevalence (visual locomotion score > 3; SEVLAME); dry cows’ lameness prevalence (visual locomotion score > 2; LAMEDRY); claw horn disruption lesion incidence in the subsequent lactation to our visit (CHDL). (B) Mean (with standard error) mature equivalent 305-d milk yield in kg (ME305MILK); dry cows’ mean (with standard error) body condition score (BCS).
Fig 2.
Lameness, severe lameness at dry-off and claw horn disruption lesion (CHDL) by different stall bedding type.
Mean farm lameness (visual locomotion score (VLS) > 2) and severe lameness (VLS > 3) prevalence at drying off and mean farm previous and subsequent lactation claw horn disruption lesion (CHDL) incidence by different stall bedding type for dry (A) and lactating cows (B). Error bars represent standard error of the mean.
Fig 3.
Lameness, severe lameness at dry-off and claw horn disruption lesion (CHDL) by rubber floor use.
Average farm lameness (visual locomotion score (VLS) > 2) and severe lameness (VLS > 3) prevalence at drying off and average farm previous and subsequent lactation claw horn disruption lesion (CHDL) incidence for farms that were or were not using rubber flooring in the dry cows pens and for farms that were or were not using rubber flooring in the lactating cows pens. Error bars represent standard error of the mean.
Fig 4.
Linear regression scatter plots.
Linear regression scatter plots (fitted lines with confidence intervals in grey) indicating (A) the positive association of the farm level incidence of claw horn disruption lesions (CHDL) in the previous lactation (Y axis) and the farm level incidence of CHDL in the subsequent lactation (X axis); (B) the positive association of the prevalence of lameness (visual locomotion score > 2) at drying off (Y axis) and the farm level incidence of CHDL in the subsequent lactation (X axis); (C) the negative association of the prevalence of lameness (visual locomotion score > 2) at drying off (Y axis) and the body condition score at drying off (X axis).
Table 1.
Explanatory categorical variables retained in the logistic regression model with dry cows’ lameness prevalence (0 = visual locomotion score ≤ 2, 1 = visual locomotion score > 2) as an outcome variable.
Least squares means with 95% confidence intervals (CI) and P values are presented.
Table 2.
Explanatory categorical variables significantly associated with claw horn disruption lesion in the subsequent lactation.
Least squares means with 95% confidence intervals (CI) and P values are presented.
Fig 5.
Claw horn disruption lesion (CHDL) incidence in the subsequent lactation by CHDL prediction quartile.
The bars represent the mean CHDL incidence in the subsequent lactation (light gray) and mean CHDL incidence restricted to the first 100 days of the subsequent lactation (dark gray). Error bars indicate the 95% confidence interval.
Fig 6.
Receiver operating characteristic curve.
ROC curve that identified the optimal cut off point for the predicted probability values obtained by the logistic regression model that had subsequent lactation claw horn disruption lesion (CHDL) as an outcome. The gold standard in this analysis was CHDL incidence.
Fig 7.
Specificity, Sensitivity, Positive Predictive Value and Negative Predictive Value.
Bars for each farm, using the predicted probability threshold suggested by ROC analysis and claw horn disruption lesion incidence in the subsequent lactation as the gold standard.
Fig 8.
Percentage of cows identified as being at risk and percentage of cows that developed CHDL in the subsequent lactation.
Percentage of cows identified as being at risk to develop claw horn disruption lesions (CHDL) in the subsequent lactation by the logistic regression model (using the cut off value derived by ROC analysis) and percentage of cows that developed CHDL in the subsequent lactation that were characterized as being at risk by the logistic regression model (using the cut off value derived by ROC analysis).