A retrospective evaluation of postbreeding infusions in dairy cattle.

Health and fertility data were recorded in 32 southern Ontario Holstein herds over a period of two and a half years. During that time 2.6% (80/3123) of breedings were followed by a postbreeding infusion. Breedings were divided into groups according to the breeding number (first to fifth), age group (less than 3 yr, 3 to 5.9 yr, greater than or equal to 6 yr) and disease status of the reproductive tract (four groups). The rate of use of infusions was higher following third to fifth breedings, increased with the age of the cow and was much higher in cows with a diagnosis of uterine infection. For each individual group of breedings the odds of a breeding followed by an infusion resulting in a pregnancy was compared to the odds of a breeding not followed by an infusion resulting in a pregnancy. In general the ratios of these odds were less than one, possibly indicating a detrimental effect due to the infusion. Summary odds ratios were calculated for each breeding number, each age group and each infection category. Again, in general, the odds ratios were less than one although the differences from one were not significant at p = 0.05. Two estimates of the overall summary odds ratio (based on all breedings) indicated that breedings followed by an infusion were only approximately 0.7 times as likely to result in a pregnancy as breedings not followed by an infusion. The two estimates of the summary odds ratio were not significantly different from one at p = 0.05. Several possible sources of bias in this type of trial, and their potential effects on the results are discussed.