Logistic Regression Analysis of Some Factors Influencing Incidence of Retained Placenta in a Holstein Dairy Herd

To investigate the effects of certain factors on the rate of retained placenta, 2844 calving records from 1288 Holstein cows in a herd were used. These cows calved during year period of 2001 to 2007. A generalized statistical linear model was applied to analyze the data. Logistic regression model was applied as the statistical model. In the model, fixed effects of year, season (warm or cold) and sex and birth type (single or twin) of calf and parity and gestation length of dam were included. In data file, retained placenta incidence was introduced as binary codes: zero for healthy cows and one for cows with retained placenta and probability of code one was analyzed. Incidence of retained placenta was significantly affected by all the fixed effects included in the model. Odds ratio estimation for cold season compared to warm seasons was 0.619, which means probability of occurrence of this disease in cold season, is 39% less than in the warm season. Odds ratio estimation for female calves in comparison to male calves was 0.6, which implies that the former are 40% less probable to cause retained placenta. Probability of this disease in twin calving was 5.9 times more than singles. By increasing the parity number of the dam, the probability of experiencing retained placenta in cow increased. It could be concluded that the incidence of retained placenta may be reduced by management modifications.