| Abstract: | This report compares estimated gestational ages from published cubic spline curves to gestational ages estimated retrospectively from delivery dates in 28 pregnancies from ten common marmosets (Callithrix jacchus). Both CRL- and BPD-based estimates of gestational age were closely correlated with delivery-based gestational age estimates. Of the three ultrasound machines used, the one with 16 shades of gray and a sequential linear array overestimated gestational age during early pregnancy, based on CRL measures. Measures from the other two machines (64 or 264 shades of gray; linear sector and annular array or electronic phase array) were similar and resulted in a correlation of the two estimates of gestational age of 0.94 and a mean difference between the two estimates of 0.16 days with 80% of CRL-based gestational age estimates being within ± 5 days of the delivery-based estimate. The reliability of BPD-based estimates of gestational age was strongly related to pregnancy outcome. BPD-based estimates underestimated gestational age in poor outcome pregnancies (i.e., those in which infants died within 7 days of birth) but not in good outcome pregnancies. The combined CRL- and BPD-based estimates on poor outcome pregnancies suggest that there was less growth in BPD in late gestation for those pregnancies that resulted in nonviable offspring. For good outcome pregnancies, the correlation between BPD-based and delivery-based estimates of gestational age was 0.871 and the mean difference between the two estimates was ?0.06 days with 83.3% of BPD-based estimates falling within ± 5 days of delivery-based estimates. |
