Traumatic injury is an enormous
public health problem and is a leading cause of death
for children and young adults. The heaviest toll comes
from head injuries, primarily due to traffic collisions.
Intracranial pressure is the great killer, so trauma specialists
have naturally assumed that reducing inflammation would
minimize the effects of brain injury. This is the fundamental
rationale behind the widespread use of corticosteroids
to treat head trauma — a rationale that now appears
to be completely wrong.
The trauma centre is a notoriously
unsuitable environment to be messing around with placebos,
so the benefit of this treatment was mostly taken on
faith. However, in 1998 the corticosteroid randomization
after significant head injury (CRASH) trial was initiated
to settle the question once and for all.
Some 239 hospitals in 49 countries
were willing to overlook the ghoulish pun in the acronym
and participate in the study. Eventually, 10,008 patients,
who all showed signs of neurological injury, were treated
under the trial protocol. With a Glasgow coma score
of 14 indicating a healthy alert state, many patients
were in very bad shape with 40% having Glasgow scores
of 8 or less, and 31% having subarachnoid bleeding.
The trial was supposed to stop
once 20,000 patients had been treated, but the data
monitoring committee found such alarming results that
the trial was halted halfway through, not because the
placebo patients were dying, but because the corticosteroid
patients were.
Of 4,985 patients allocated corticosteroids,
21% died within two weeks of randomization, compared
with 18% of the 4,979 patients given a placebo. The
findings were the same for all degrees of injury severity.
The researchers are currently analyzing six-month followup
data to look for effects on subsequent disability.
The CRASH findings could also be
relevant in spinal cord injury, where corticosteroid
use is still very much in favour. All the research done
on this question to date has only covered about 500
patients. "Use of corticosteroids in spinal cord injury
should remain an area for debate," argue the authors.
Commenting on the findings in the
The Lancet, German trauma specialists Drs Stefan
Sauerland and Marc Maegele wrote: "Most clinicians expected
the trial to confirm the benefits of steroids, while
others suspected that the effectiveness of steroids
would turn out to be small or non-existent. The results
of CRASH are therefore a complete and alarming surprise
for all."
"When extrapolating the results
of the CRASH trial to the annual incidence of severe
head injuries worldwide, it is frightening to calculate
how many patients might have been harmed by corticosteroids,"
they say.
Most trauma centres have already
abandoned corticosteroids for head injury, but evidence
suggests that is truer in Europe than in North America.
A 1996 British survey found that just 14% were still
using corticosteroids, but a 1995 survey of US centres
found 64% of trauma centres still use this treatment.
No data is available on Canadian use.
Nobody has any idea what mechanism
led to the deaths in the corticosteroid group. There
was no evidence of extra infections or gastrointestinal
bleeding among these patients — something else
is going on. "CRASH partly shakes our pathophysiological
understanding of what is of primary importance after
traumatic brain injury," say Drs Sauerland and Maegele.
"The key message of CRASH, however, is that applying
treatments with unproven effectiveness is like flying
blindly."
|
