Relationships between cytology, bacteriology and vaginal discharge scores and reproductive performance in dairy cattle

The objective was to compare three diagnostic approaches for intrauterine infection and inflammation: scoring of vaginal contents; quantification of percentage of nucleated cells that were polymorphonuclear leukocytes (PMN) following endometrial cytology; and intra-uterine bacteriology. Dairy cows (n = 303) were examined twice, Days 28 (D28) and 42 (D42), where Day 0 = day of calving. Associations between gross vaginal inflammation scores, uterine cytology, and bacteriology, and subsequent reproductive performance were examined using multivariable models. There was fair agreement at D28 (Kappa = 0.29), but only slight agreement at D42 (Kappa < 0.15), between PMN% and gross vaginal inflammation score. Cows were categorized as having PMN% in the highest quartile (H), or not (L), at both D28 and D42; therefore, cows were categorized as PMNLL, PMNLH, PMNHL, or PMNHH. Cows in the highest PMN% quartile at both time periods were slower to conceive (P < 0.001) than those in all other quartiles (mean ± SEM 32.2 ± 2.3, 37.0 ± 5.3, 40.8 ± 4.1, and 55.3 ± 7.3 d from start of breeding to conception for PMNLL, PMNLH, PMNHL, and PMNHH PMN% cows, respectively). Milk yield was greater (P = 0.001) in cows in the lower quartiles for PMN% at D28 and D42 (i.e., PMNLL) than those in the PMNHH and PMNHL categories, with PMNLH intermediate (P = 0.001). We concluded that PMN% was a better predictor of reproductive performance than either intra-uterine bacteriology or gross vaginal inflammation score. Cows in the highest quartile for PMN% at both D28 and D42 had lower pregnancy rates, took longer to conceive, and had a lower milk yield than those in the lower PMN% categories.