Hypercorrect Is a Variant of Wrong

Richard M. Plotzker, MD


March 08, 2017

Who remembers having a patient who was "euboxic"? That relic of medical terminology traces to the early days of computerized applications of medicine, so extraordinarily expanded today.

Back then, patients would have serum drawn, the autoanalyzer would run 15 or so tests, and a printer would generate a report. Each test would have a vertical line with a black dot indicating the result. A segmental line would connect each dot.

Richard M. Plotzker, MD

The normal range for each value would be depicted by a gray rectangle. If all 15 values fell in the gray boxes, that patient would be deemed euboxic. At 95% confidence intervals for each test, the likelihood of a normal person falling into that esteemed category of normality for all 15 tests would be 0.95 X 15 = 0.46, meaning less than half of all seemingly normal subjects had all 15 tests in the normal range.

As more tests were added to the panel—eventually reaching 27 before the Centers for Medicare & Medicaid Services wisely replaced them with the more rational basic metabolic panel and comprehensive metabolic panel—the prospects for being euboxic became vanishingly small. Yet, each lab abnormality needed some type of assessment if not treatment.

Hyperparathyroidism went from being a disease of bone and kidneys to a disorder of the autoanalyzer. Many other endocrine disorders—from minimal thyroid-stimulating hormone variants to incidental CT findings—similarly pose the dilemma of how much medical care to offer for these minor abnormalities, separating the trivial from the significant.

Although we can't avoid the unexpected lab or anatomic abnormality, perhaps we are making too much of an effort to seek them out by ordering tests that will take us and our patients down the road of excessive unproductive care.

Pointing Fingers

While rounding with my resident on elective last month, we saw a fellow with new transaminase elevations that had not been retested for a few days. We ordered the hepatic function panel for comparison but noted that basic metabolic panel, phosphorus, and magnesium levels were appearing on his chart every day—to the neglect of what really needed to be measured instead.

I asked the resident why this was, and he told me that the attending hospitalists insisted on it. We rounded a little more and found a few more patients who had repetitively normal or at least not dangerous magnesium and phosphorus levels.

As I encountered each hospitalist randomly over the next few days, I asked why they insisted on repetitive testing of this type on so many patients. They each denied requiring residents to order these.

Finally, the plug got pulled one day when one of our consults got riders of intravenous potassium, phosphate, and magnesium for values that were only 0.1 mg/dL below the lab norms.

Widespread Problem?

Although excessive care and its potential misadventures can be a little like Justice Stewart's definition of pornography (ie, you don't quite know what it is, but you know it when you see it), two recent contributions to the medical literature tried to analyze the downside of doing testing that would have been better not being requested. These publications attempted to view the consequences of ordering testing when the treating physician's current information was already sufficient.

The first comes from the British Medical Journal.[1] It is an American study from the Mayo Clinic looking at about 31,000 patients with stable diabetes, defined as two consecutive glycated hemoglobin (A1c) values below 7% taken 6 months apart and no other confounding illness. The accepted standard of care would be two measurements 6 months apart.

Allowing for the vagaries of people seeing multiple doctors, the study authors identified 6% of patients getting five or more tests per year and 55% getting more than two consecutive tests in less than a 3-month interval. These were stable patients, so quite a lot of them were on no medication at the beginning and end of the observation period, yet 5% of the untreated people also received excessive testing.

Although some patients just drifted along perfectly fine on whichever medication they were on regardless of how many A1c measurements took place, as a composite, more testing resulted in more changes in treatment from one class of drugs to another. Because of the means of accessing the database, the investigators could not tell what the tolerance was to the agents that were changed.

Replacing one poorly tolerated drug with another and retesting seems reasonable. However, there was a clear correlation between the number of tests and the number of drug changes, even though all A1c values were stable.

Of course, the people who did the most testing were my fellow endocrinologists.

Old Habits Die Hard

The second poke at me and my colleagues—and probably most medical school graduates—comes from the US Preventive Services Task Force, who recently updated their recommendations[2] on screening for thyroid cancer, which were first issued in 1996. Both then and now, we are advised against routine screening unless a person has high-risk factors such as a radiated gland or family history of thyroid cancer. Both palpation and ultrasound are potential screening modalities. Palpation tends to be not very sensitive and correlates with experience as an examiner. Ultrasound finds a lot of what you wish you hadn't found.

Almost 100% of patients who come to my exam room for any reason will have their thyroid area examined and a few comments made about it in the electronic health record, even though I am not looking for anything in particular.

From my first physical diagnosis course as a second-year medical student until my current status as senior physician, I have seen the more experienced docs belittling the less experienced docs for not being sufficiently thorough in examining patients. My professors would term me lazy if I did not examine the thyroid. My fellowship mentors would be a little uppity if their exam found something that my exam did not.

As residents come for elective, I have them examine thyroid glands and describe what they feel, shining a light on it from their cell phones to describe what they see. But it's all for naught, as the experts suggest that patients might be better off without our experienced fingers looking for 1- to 2-cm spheres of abnormality that might lead to surgical complications for a disorder that never would have harmed them had benign neglect prevailed.

It's just hard not to examine that inviting anterior neck when the patient is sitting on the exam table at just the right height.

So, we return to that final aphorism of the Fat Man, hero of Samuel Shem's novel, The House of God: "The delivery of good medical care is to do as much nothing as possible."


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: