N. Kanuya et al., DIAGNOSTIC-ACCURACY OF ESTIMATING THE NUMBER OF CL AND OVARIAN VOLUMETO PREDICT SUPEROVULATORY RESPONSE IN DAIRY-CATTLE, Reproduction in domestic animals, 32(6), 1997, pp. 291-296
The superovulatory response of 44 dairy cattle was assessed by estimat
ing the number of CL and ovarian volume per rectum before non-surgical
recovery of eggs. The study included the recovery of greater than or
equal to 3 total eggs and greater than or equal to 2 viable/transferab
le embryos which were then used retrospectively to determine the diagn
ostic accuracy of rectal palpation in predicting superovulatory respon
se. When <5 CL were palpated, the specificity (SP) of predicting the r
ecovery greater than or equal to 3 total eggs or greater than or equal
to 2 viable embryos was <0.5. The diagnostic accuracy of estimating t
he number of CL was high (both sensitivity (SE) and SP were >0.5) when
5-12 CL were palpated. The Receiver Operating Characteristic (ROC) cu
rves demonstrated that estimating the number of CL was more accurate t
han the ovarian volume in assessing superovulatory response. The best
cutoff point. where fewest mistakes were made, was when 9 CL were palp
ated. At this point of maximum accuracy, the likelihood ratio for a po
sitive test result in the recovery of greater than or equal to 2 viabl
e embryos was 6.5 and that of a negative test result was 0.3. In concl
usion, it was shown that the estimation of the number of CL per rectum
in order to predict the recovery of greater than or equal to 3 eggs o
r greater than or equal to 2 viable/transferable embryos was diagnosti
cally optimized when 5-12 CL were palpated.