Furthermore, identifying appropriate research sites
might be more difficult with embedded trials, if otherwise
ideal sites anticipated difficulties in recruiting
pregnant women (owing to limited expertise with this
participant population among the investigators at the
site, limited access to this participant population and
cultural norms precluding research involving pregnant
women). This difficulty could be addressed by not
requiring all sites to enroll pregnant women. However,
this practice could give rise to other difficulties if the
number of sites enrolling pregnant women was sufficiently
low as to limit the generalizability of the data.
Finally, including pregnant women in late Phase II or
Phase III trials might delay availability of data from the
overall population if pregnant women were enrolled at
a slower rate than nonpregnant adults.