Share
All the best evidence protocols are worthless in the face of
the wrong diagnosis. With the efforts to standardize medical treatment, reduce
costs and enhance patient safety, planners pay little attention to missed or
wrong diagnosis. Disease in my formal medical training was defined as a maladaptation
to the environment, and at some level, indeed it is. Medical science and the
physician’s art define the causes of that maladaptation as a diagnosis or
syndrome. Unfortunately many diseases have common signs and symptoms. The body
has a limited number of ways to respond to disease and there are so many
diseases and syndromes – more than we can remember. The identification of a diagnosis, therefore,
may be uncertain at best.[1]
|
Two thousand four hundred and fifty years ago Hippocrates
confronted the challenge of diagnosis with terms meticulously describing the manifestations
of disease. His precise descriptions and inductive reasoning lead to the abandonment
from mythology in diagnosis and from the concept of disease as a punishment by
the gods. His diagnosis was correct in that it was a precise description of its
effect.
Today, however, with far greater knowledge of
pathophysiology we struggle with identifying the cause of our patients’
problems. We try to identify the diagnosis in order to relate a treatment plan that
is safe and yields a prognosis favorable to the patient. What is the problem? With what little research
there is on wrong diagnosis and missed diagnosis, we find up to a 20% error
rate. Autopsy findings prove error rates much higher. In some cases, such as
pulmonary embolism or aortic aneurism, the error rate runs as high as 60%[2]
Outpatient ENT errors run high as well and most hospital admissions for pneumonia
are signed out as community acquired pneumonia unidentified.
Many causes lead us to submit an erroneous diagnosis.
Perhaps the greatest problem is the number of layers of cause and effect that
we deal with. Is it a gross diagnosis or a tissue diagnosis or tomorrow a bio-molecular
diagnosis – even genomic? Is it left
lower quadrant pain, a left ovarian mass, a corpus lutein cyst, a seromucinous
cyst adenocarcinoma and if the later, what proteinomic predisposition?
Autopsy has gone out of style. Two hundred years ago Marie Francois Bichat wrote
over his autopsy room in Paris, “Death comes to the Aid of the Living.” The abandonment
of autopsy must be economic. Maybe insurance companies refused to pay for it.
Patient’s families have always objected. The hospital does not want the
responsibility. The pathologist finds it a low marginal return activity. For whatever
reason, autopsy -- except for forensics -- seems a lost art. We find clinical
pathological conference and morbidity and mortality conference replaced by the x-ray
conference. With the abandonment of autopsy we have lost the gold standard for definitive
diagnosis.
The CAT scan has replaced the autopsy in an intellectual sense, and that
is all wrong. The CAT scan is expensive, far more so than an autopsy and is by
no means definitive at the tissue, cellular and molecular level. Although X-ray
conferences are contemporary, and bring together multiple specialties and the
diagnostic suggestions of the group, they still may still fail to yield the
right answer. CAT scans done extensively in place of careful differential
diagnosis and basic cognitive disciplines cost the patient and the health care
system enormously. With an initial wrong presumption the addition of added
unnecessary tests and procedures lead to decreased patient satisfaction, safety
and poorer outcomes. --- “First, do no harm!”
Patients are different. There molecular biology may be as varied as their
noses. Statistically derived presumptions, may fail to account for geography,
ethnicity or individual circumstances. Additionally, the physician may decline
to write a diagnosis that may cause the patient to be ineligible for insurance,
induction into the military or some other sensitive job such as hospital
privileges for a doctor or class I medical for an airline pilot. These are
usually minor omissions, but together they deny any meaningful analysis of a
database based on the diagnosis.
Our very regulations seeking standardized care may cause a more insidious
error rate based on the diagnostic requirement for laboratory or other
diagnostic studies. The same may apply to treatment plans – a treatment looking
for a diagnosis. A similar diagnostic distortion occurs, when the diagnosis is
altered or selected for reimbursement purposes. Hopefully these alterations are
harmless to the patient, but in some context can lead to a wrong presumption in
a future event that misses a lifesaving solution.
Overconfidence, assumption, snap diagnosis and omission of essential
parts of the history and physical contribute to errors as well.[3] We do not keep statistics
on wrong diagnosis, but the trial lawyers do, and they can tell us something. Graber et al chronicle the size of the missed
diagnosis problem which results in between 40,000 and 80,000 deaths a year in
the US. The problem may actually be greater.[4]
Cognitive scenarios as presented by primary care specialty boards, the
review of systems together with listening to the patient, a complete history
and physical at least for the organ system involved will go far. But the
association of findings with the vast number of possibilities suggested by
these findings presents a challenge to the clinician. It is a challenge that
medical schools can meet with appropriate lists, current terminology and
criteria for diagnosis -- most of all a vigorous
CME program. Johns Hopkins will hold a conference on wrong diagnosis this
month.[5]
[1]
Delaney;
Munro: Diagnostic difficulty and error in primary care; Family Practice 2008;
25 (6) 400-413
[2]
Berner,
Graber; Overconfidence as a cause of diagnostic error in medicine; Am J Med
2008;121: S2-S23
[3]
Bringing Diagnosis Into the Quality and Safety Equations; Graber ML;
Wachter RM, Cassel CK. JAMA. 2012; 308(12): 1211-1212.
I see the greatest contents on your blog and I absolutely love reading them.cme conferences
ReplyDelete