I’ve been mulling this over for a few days, finally deciding to write
about it.  There was an article in the NYT on 13 January 2010
about an NEJM article:
href="http://www.nytimes.com/2010/01/14/health/research/14morphine.html">

href="http://www.nytimes.com/2010/01/14/health/research/14morphine.html">Morphine
May Help Traumatic Stress
By BENEDICT CAREY
Published: January 13, 2010

Doctors have long hoped to discover a “morning-after pill”
to blunt the often disabling emotional fallout from traumatic
experiences. Now it appears that they have had one on hand all along:
morphine…In a large study of combat casualties in Iraq, Navy
researchers reported Wednesday that prompt treatment with morphine cut
in half the chances that troops would develop symptoms of
post-traumatic stress later on…The new study, appearing in The New
England Journal of Medicine, supports the standard practice in settings
like the battlefield and emergency rooms, where morphine is often used
readily. But experts say it may have implications for the timing of
treatment and for a wider variety of traumas, like those resulting from
rape or muggings.

Overall, the NYT article is reasonably good, complete with the
necessary disclaimers about how limited the study is.  One odd
thing about it: Carey’s articles usually exhibit a noticeable bias
against the use of psychotropic medication.  I did not notice any
such bias this time.  The reason this is notable, is that often
the medications he disparages often are drugs that have undergone
multiple placebo-controlled, double-blind clinical trials.  There
have been no such studies on the use of morphine for PTSD. 

Indeed, this is an area in which it would be exceptionally difficult to
design a really good prospective study.  After all, you cannot
randomly select people to be traumatized.  And it would be
difficult to justify randomizing people to receive placebo, if they
have been traumatized in such a way that morphine ordinarily would be
given.

As it stands, the study really does not have any direct clinical
application.  (Few single studies do, of course.)  One reason
it is interesting, is that it provides some comfort that the morphine
does not makes things worse for the patient, in terms of the
psychological aftermath.  The other reason it is interesting, is
that it would be really nice to know how this works on a molecular
level.  That remains highly speculative.