Factors associated with embryo recovery in mares: A retrospective multicenter analysis and practical overview of embryo flushing protocols

Embryo recovery (ER) is a major determinant of efficiency in equine embryo-transfer programs and may be influenced by biological, management, and technical factors. This retrospective multicenter study aimed to identify factors associated with positive ER and to describe the embryo-flushing (EF) techniques used in three commercial embryo-transfer programs. We hypothesized that factors associated with ER would differ among clinics because of differences in mare populations, semen types, management practices, and flushing systems. The study included 2861 EF cycles: 465 cycles from 196 mares in Clinic 1, 661 cycles from 265 mares in Clinic 2, and 1735 cycles from 277 mares in Clinic 3. Factors associated with positive ER were evaluated separately for each clinic using multivariable generalized linear mixed models, with mare identity included as a random intercept. Additional univariable analyses were used to examine selected factors individually. Overall ER was 55.7%, 51.4%, and 73.5% in Clinics 1, 2, and 3, respectively. In the multivariable analyses, none of the evaluated fixed effects were significantly associated with ER in Clinic 1. In Clinic 2, semen type was the only significant fixed effect, with greater odds of ER using chilled than frozen semen (OR = 2.07, 95% CI: 1.37–3.11; p < 0.001). Clinic 3 used fresh semen exclusively, and donor mare age was the only significant fixed effect, with the odds of ER decreasing by 9% for each additional year of age (OR = 0.91, 95% CI: 0.88–0.94; p < 0.001). Although the overall effect of season phase was not significant in Clinic 3, ER was lower during mid-season than early season (OR = 0.67, 95% CI: 0.46–0.99; p = 0.043). Mare identity had a significant random effect in all three clinics ( p ≤ 0.05), indicating substantial between-mare variability. Repeated EF within the same season did not reduce ER in any clinic, including among mares undergoing four or more cycles. In the additional analyses, multiple ovulation was associated with higher ER in Clinics 1 and 3 but not in Clinic 2. In Clinic 2, experienced operators showed a tendency toward higher ER than less experienced operators (52.4% vs. 37.7%; p = 0.06). In conclusion, factors associated with ER differed among clinical settings. The consistent mare effect, the absence of a decline in ER with repeated flushing, and the potential influence of operator experience emphasize the importance of individual mare fertility and clinic-specific management when evaluating the efficiency of commercial embryo-transfer programs.