Relationships between uterine microbiology, cytology and pregnancy rates in Thoroughbred mares 2006-2013

Abstract

Endometrial bacteriology and cytology are used to diagnose endometritis in broodmares. Recent studies have investigated the diagnostic value of both methods (Riddle et al, 2007; Davies Morel et al, 2013). The aim of this retrospective study was to investigate the predictive value of endometrial microbiology and cytology by relating results to per-cycle pregnancy rates in a UK Thoroughbred population over eight consequent breeding seasons. We also aimed to identify any changes in microbial prevalence over this speriod. We first examined the relationship between bacteriological and cytological classification. A multiple logistic regression was then performed to investigate the predictive value of these indices of endometritis for pregnancy. The influence of mare age and status (barren, maiden, foaling), stud farm, year in which the results were obtained, and treatment of presumed infections on subsequent endometrial status and pregnancy were also investigated. Endometrial swabs (n=8752) were collected via the external cervical os from 2398 Thoroughbred mares during the breeding seasons of 2006-2013. Status, age, intra-uterine treatment and establishment of pregnancy were recorded for all mares for each swab. 25.8% of all swabs were positive for microbial culture, whereas only 8.2% of mares were positive for inflammatory cells via cytology. 5.7% of mares were positive on both culture and cytology. Beta-haemolytic Streptococcus, Escherichia coli and Staphylococcus aureus were the most frequently cultured organisms. Relative organism prevalence did not change significantly over the years examined (P>0.05). Culture and cytological results were well related according to the Chi squared test (P<0.05), but not according to McNemar’s test (P<0.05). Chi-Squared tests were performed to examine the influence of each factor on pregnancy rates; a factor was entered into the logistic regression model if P<0.10. All factors (year, stud farm, mare age, mare status, bacteriology result, cytology result, and treatment) were included. In the logistic regression model, treatment was of the least influence on conception rates, followed by cytological status. All other factors had a significant relationship with pregnancy rates. However, type of pathogen and grade of cytology were not accounted for in this model. In conclusion, this study shows that simple positive or negative microbiology or cytology on an endometrial swab are not enough to determine the chances of a mare conceiving. Grade of growth and inflammation, type of pathogen and mare age and status must be taken into account, as well as clinical findings and stallion and management factors.