About a third of patients presenting
with dyspnea in the emergency room owe their breathing
trouble to acutely destabilized heart failure. Catching
all of these heart failure cases has always been a challenge.
In a new Archives of Internal
Medicine study, in the April 14 issue, doctors and
researchers from Harvard Medical School and Massachusetts
General Hospital argue for more use of natriuretic peptide
testing to support diagnosis. Their case is an eminently
sound and sensible one, but the methodology and reasoning
used to get there is a little strange, to say the least.
Two arguments are made by the researchers.
One is that natriuretic peptide testing, specifically
amino-terminal pro-B type natriuretic peptide (NT-proBNP)
testing, can aid clinicians to make more accurate diagnoses
of heart failure (HF) in emergency department dyspneic
patients. This has already been shown in other settings,
and is again shown here.
Their other argument is that "clinical
uncertainty" in the diagnosis leads to worse outcomes
and longer hospital stays, and that a higher proportion
of those uncertainly diagnosed actually turn out to
have heart failure.
The study uses data from previous
research in which clinicians were asked to rate, as
a percentage figure, the likelihood that each patient's
dyspnea was due to HF. Any rating of under 20% or over
80% was declared to be "clinical certainty." Anything
in between was declared to be "clinical uncertainty."
SEEKING
CERTAINTY
It then goes on to show that, of 592 patients in the
study, the 31% who had an "uncertain" diagnosis were
more than twice as likely to be declared heart failure
patients in a later adjudication. They also had higher
mortality, were more likely to be admitted to hospital,
and stayed there longer.
But hang on. What about the patients
with certain diagnoses? It turns out that more than
four-fifths of these got "certain" diagnoses because
the physician judged their chances of having heart failure
to be less than 20%, while only a small minority got
"certain" diagnoses because their chances were rated
over 80%. In fact the largest single category by far
was those whose chances of having HF were rated at 0%.
Other studies, and this one too,
have shown that clinicians are very, very good at accurately
excluding diagnoses of HF in dyspnea. So when the study
claims that patients with uncertain diagnoses fared
worse, what they're really saying is that a group of
patients in whom physicians refused to rule out HF after
examination fared worse than a group in whom the vast
majority had seen HF ruled out upon examination. Now
there's a surprise.
|
