Pulmonary embolism (PE) affects about five in 10,000 Canadians
and is deadly if left untreated. On the other hand, only
one in 10 patients who presents to the emerg with a suspected
case actually has one. Trouble is, there's no easy way
to determine who's who.
Doctors are faced with a conundrum.
The lung blockage kills many patients because of missed
diagnoses due to vague and non-specific symptoms, but
the treatment for it also carries risks and shouldn't
be given lightly. The best diagnostic test, pulmonary
angiography, is also considered unnecessarily dangerous,
not to mention expensive, given the relatively small
number of diagnoses.
For years, researchers have sought
a way to eliminate the low-risk patients from the testing
process early, based on predictable characteristics.
Three studies in the January 23 issue of the Archives
of Internal Medicine set out to do just that: eliminate
invasive testing in low-risk patients and so save time
and money and reduce complications.
TURNING
ON A DIMER
The most popular predictive tool for PE is the Wells
algorithm. It considers seven clinical features that
can be gleaned from a quick exam of the patient and
their medical history. But the Wells scoring system
doesn't spare patients from more invasive techniques
like pulmonary angioplasty. Rather, the Wells scale
is used to establish a 'pretest probability,' which
then guides the physician's interpretation of other
test results.
A group of researchers from France
set out to determine if a simple blood test, known as
the D-dimer assay, combined with a low Wells score could
exclude PE in patients with a history of blood clots.
The test flags clotting problems by measuring levels
of D-dimers, by-products of clot formation.
The researchers looked at the results
of two previous prospective studies that included 1,721
consecutive patients, 308 of whom had a history of venous
thromboembolism (VTE). All patients whose test revealed
low levels of D-dimers had no recurrent events in the
succeeding three months.
In an accompanying editorial, Dr
Lisa Moores of the Walter Reed Army Medical Center was
cautiously optimistic about the results. "The combination
of a low pretest probability and a negative D-dimer
test result should obviate the need for further testing,"
she wrote. Unfortunately, both she and the authors themselves
point out that the chances of a negative D-dimer test
are low among patients with a history of VTE. So while
this approach can reliably single out patients who can
safely go untreated, it doesn't single out very many.
WHAT
ABOUT DEAD SPACE?
A team from the Ottawa Hospital took things a step further.
They assessed whether a combination of three bedside
tests, the Wells scale, the D-dimer assay and alveolar
dead space measurements (which requires patients to
undergo blood gas sampling) could exclude PE. Among
patients scoring four points or less on the Wells scale,
those who got negative results on either of the two
diagnostic tests were left untreated.
Their outcomes were then compared
to those of patients in whom PE was ruled out by a ventilation-perfusion
(V/Q) scan, a test which shows if areas of the lung
are not receiving blood because of a clot. Rates of
pulmonary embolism were similar, at 3% for the V/Q scanned
patients and 2.4% for the "bedside investigation" patients.
This technique enabled the researchers to reduce the
use of diagnostic imaging by 34%.
The problem, wrote Dr Moores, is
that not that many centres have the facilities needed
for reliable alveolar dead-space measurements. Moreover,
blood gas sampling isn't currently a routine test in
suspected pulmonary embolism.
RISK
TOP TEN
The final study is probably the simplest of all. It
distilled the high-risk factors down to 10 key criteria:
1. Age 70 or older
2. Hx of cancer
3. Hx of heart failure
4. Hx of chronic lung disease
5. Hx of chronic renal disease
6. Hx of cerebrovascular disease
7. Altered mental status
8. Pulse rate >/- 110 beats/min
9. Pulse rate >/- 100 mm Hg
10. Arterial oxygen saturation < 90%
The team, led by Dr Drahomir Aujesky
of the University of Pittsburgh, retrospectively followed
over 10,000 patients for one month after diagnosis with
PE. Patients with none of the 10 risk factors, who accounted
for 21.6% of the sample, were deemed low-risk and eligible
for discharge. Thirty-day mortality was just 1.5% in
these patients, compared to 9.6% for the overall sample.
The great thing about this approach
is that it relies on data that's easily obtained with
a thorough history and physical exam. But only a minority
of patients will qualify as low-risk using these criteria.
"Many patients presenting with suspected PE will have
one of these comorbid conditions, making me question
the true clinical utility of this model," wrote Dr Moores
in an accompanying editorial.
Though all three approaches proved
reliable, ultimately too few patients are being singled
out to make any of them the gold standard.
|
