Unfortunately I am not an ID expert, but, as a neonatologist who has been
in this clinical situation I am inclined to agree with Jack's premises.
There are a few interesting points about CMV infection in developed countries:

1.  It is common:  1-2% of pregnant women may acquire CMV in the year of
their pregnancy (or indeed any other year).  As a result antibodies to CMV
(IgG class are common in mothers and infants).
2.  It rarely causes serious complications:  most of these pregnant women
will never know that they had CMV and their children will be fine.
3.  When it causes complications in a small fraction of pregnancies it is
serious:  there is a risk of multiorgan involvement and long-term
neurological sequelae.
4.  Infants who are infected may shed large quantities of virus in their
body fluids for many months.  We often make the diagnosis from culturing
the urine.

Given #4 I have difficulty with the recommendation to discontinue
lactation: especially as many more women around the world, unknown to them,
had CMV in pregnancy and breastfed without ill-effect.  It sounds like the
opinion of an ID expert should be sought before carrying out the advice to
discontinue breastmilk:  this baby needs every advantage it can get.

Khalid