First, Do No Harm

I find that it takes more effort for me to talk patients out of doing tests than it takes to get them to submit.  It doesn't seem like it should be that way.

I need you to undergo a test where you fast for 12 hours, show up to the hospital 3 hours before the sun comes up, strip down and slide into a breezy hospital gown, have someone stab at your arm for blood specimens and an IV, sit around anxiously waiting for your turn to come up, get dragged onto a cold procedure table, have total strangers shave and scrub your groin, and then have a doctor come at you with every manner of sharp instruments aimed at parts of your body where the sun rarely shines . . .

That seems like the kind of deal I'd take only if my life depended on it.  Yet I'm frequently in the position of trying to convince a patient not to pursue further testing, even if it is invasive, uncomfortable, expensive, and, in my opinion, not needed.

A typical example is the patient I'm asked to see who has a slightly abnormal stress test but no symptoms consistent with coronary artery disease.   The indication for the stress test are typically somewhat nebulous—maybe "screening" or "family history" or simply "it's been a while since I had one"—and the abnormality on the stress scan is usually a low risk borderline defect.  But since the test was not normal (strictly speaking) the patient comes to me to learn what the next step in the process is.

"Exercise, weight loss, tighter cholesterol and blood pressure control," is my answer.  "That's the next step in the process."  Most patients stare at me in disbelief.  That's when I get myself comfortable in my chair and launch into a discussion about the issue of false-positivity in medical testing and the value of pre-test probability.  Even if the test is accurate, I go on, we have to keep in mind that coronary angioplasty in low-risk, asymptomatic patients is of questionable benefit and, in the end, the only intervention that makes an impact on reducing the likelihood of heart attack is risk factor modification.  In other words, exercise, weight loss, tighter cholesterol and blood pressure control.

It turns out that it takes me a lot longer to deliver that lecture than it does to say "let's set up your cath for next Tuesday."

I occasionally run across people who breathe a sigh of relief when I tell them that I think their hearts are fine and don't need further testing, but they are in the minority.  These are patients who are grateful to hear that they can take matters into their own hands—rather than turn themselves over to the hands of the angiographer—and engage in a real effort to curtail their risk of heart attack, stroke and death.  I remind them that we can always arrange an angiogram in the future if ever they begin having symptoms of angina (chest pain, shortness of breath with exertion).

On the surface this looks like yet another example of how we rich Americans demand unlimited access to the fanciest and most expensive technology modern medicine has to offer—the my-foot-hurts-so-I-demand-an-MRI mentality that has pushed our national health care bill to a critical level.

Personally, I think this pattern has arisen mainly as a result of people with good intentions trying to do what they think is the best for themselves (in the case of the patient) and for their customers (in the case of the doctor).  Patients have developed an overly optimistic trust in the system that leads them to believe that all testing is accurate, all screening is necessary, and all disease can be prevented or cured if the right steps are taken.  Doctors want desperately to cure their patients, or at least offer some respite from their symptoms, and know that they have at their disposal an arsenal of tests and procedures that at least sometimes produce this result.  Moreover, the last thing a doctor wants is to have a patient come to her for further evaluation, and then walk out and have a heart attack (or stroke, or cancer, or whatever) after the doctor tells her patient that nothing is wrong--not out of concern about a malpractice case, but rather because the doctor legitimately wants the best for her patients and feels the weight of failure anytime she makes a mistake.  In most cases it's just easier on both the doctor and the patient to order that next test or perform that next procedure.

History plays a role, too.  In the world of cardiology in particular we have a long legacy of coronary interventions that are followed by the legitimate need for frequent repeat stress tests and cardiac catheterizations.  As I detailed in a previous post, in the early days of angioplasty—before the advent of stents—it was not uncommon for patients to come back repeatedly to have their coronary arteries imaged and ballooned.  Most seasoned cardiologists (that's my polite way of saying old) came through their training at a time when we routinely scheduled stress tests every 6 months or yearly to follow the potential failure of the patient's last angioplasty.

It's interesting to note that this pattern persists despite newer stents that reliably ward off the problem of angioplasty failure.  Last week's issue of the Journal of the American College of Cardiology published a report detailing the use of routine stress testing in patients who had recently undergone successful coronary stenting or bypass surgery.  Using a national health insurance claims database the researchers determined that 59% of recently-revascularized patients underwent some type of stress test in the ensuing 24 months.  As you might guess, the yield of such testing was remarkably low—only 5% of them ended up actually needed additional coronary intervention in response to their testing.  In other words, 95% of the tests done on this population led to no change in the care they received and no improvement in their outcome.

It's not hard to see the legacy of failed angioplasty in the findings of this study.  Cardiologists order these tests because their experience has shown them to be beneficial—only, in the current age of highly-successful intervention, this learned pattern is no longer needed.  It's for this reason that 5 years ago the American College of Cardiology Foundation proposed "appropriate use criteria," based on objective data rather than subjective anecdotal experience, to aid doctors in selecting what types of patients should undergo stress testing.  Very low-risk patients, like the one I highlighted in the vignette above, should not be sent for a stress test in the first place (asymptomatic with very few risk factors for coronary disease).

It goes without saying that the downside to such extensive testing is the very real possibility of complications that arise once an abnormal stress test is identified.  Angiography is an invasive procedure that carries with it a risk of bleeding, heart attack, stroke and death.  Most patients sail through this procedure without untoward effect, but there are a rare few who suffer despite our best efforts.

Most doctors sincerely want the best for our patients.  We are driven by a legitimate desire to do whatever it takes to find and fix whatever problem a patient brings to us, regardless of cost or complexity.  But sometimes the best is to do less rather than more—even if it takes some explaining.