It’s tempting to begin this health tip with “Here’s the news from Lake Wobegone…” You’ll soon discover why as I tell you about two recent articles of note. Also, I’d be most interested in your interpretation from a health care consumer’s point of view.
The first article, published this week in JAMA Internal Medicine, reported the findings of an online volunteer group of 118 general internists who had agreed to participate in a unique study that tested both their diagnostic expertise and their self-confidence in that expertise. First, though, you need to know that physicians are regularly checked with test patients throughout their careers, usually as a prerequisite for recertification in their specialty. These tests can be quite complex, the physician starting out by reading the fictional patient’s history, physical findings, lab tests, and x rays. The doc may even be shown photographs of his or her “patient.” Then, using the old-fashioned multiple-choice format (but online), the physician chooses appropriate steps to manage this medical challenge, which will end either well or badly for the make-believe patient.
Until now, though, no one has examined the degree of self-confidence among physicians as they go through this diagnostic testing.
As a patient, ideally what you want is a doc who’s a good diagnostician, who listens to you, orders appropriate tests, and is reasonably confident in what she’s doing. But equally important, you want someone who, when confronted with a more complex diagnostic dilemma, is willing to say “Hmmm…I’m not sure” and order the additional tests required to reach a diagnosis and/or seek help from other sources, like a textbook, the internet, or another physician, either a colleague or by referring you to a specialist.
If that’s your kind of doctor, the article’s results are not good news.
Researchers presented physicians with four imaginary patients, two with easy diagnoses and two with quite difficult ones. Here’s a bullet-point list of results:
- The doctors’ ability to diagnose the easy cases correctly was just 55%. This means, of course, that 45% of the physicians missed one or both of the easy diagnoses.
- Physicians attempting the difficult cases fared very poorly, with only 5% diagnosing the challenging cases correctly.
- But the really bad news was the degree of personal self-confidence these physicians had in their diagnostic prowess–their reluctance to say, “I’m not sure. I need you to have more tests. I think you need to see a specialist.” In fact, and this is decidedly creepy, they were almost as confident misdiagnosing the difficult cases as they were misdiagnosing the easy ones.
There were all sorts of explanations for why this disquieting “wrong diagnosis made with confidence and reluctance to seek additional information” occurred. I’m sure everything has been suggested, ranging from attitudes learned in medical training or being overworked with too many patients to an ego-led belief in their own personal “brilliance.”
Doctors are docked financially for getting diagnostic help
To me the most disturbing possibility for this situation is a consequence of physicians being financially penalized by insurance companies for ordering “too many tests” or making “too many specialist referrals.” How many doctors have to force themselves into an artificial comfort zone, convincing themselves they’re satisfied with decisions that are clearly beyond the information they have available or clearly beyond their level of expertise?
As a quick example, if an insurance company penalizes a doctor for ordering too many MRIs, he’ll cut down on ordering them, to the delighted approval of the insurance company. After a year, if nothing goes wrong, he’s probably grown over-confident in his “I can make a diagnosis without an MRI” position. But when someone from the sheriff’s office delivers his missed-diagnosis malpractice suit, during the subsequent trial his excuse “but the insurance company told me to cut down on MRIs…” will get him exactly nowhere.
On that cheery note, here’s the second piece of bad news this week, very much related to the first.
The obscenely profitable UnitedHealthcare recently announced it was cutting thousands of physicians from its networks. Physicians in 11 states who were participating in one of United’s programs called Medicare Advantage received—literally–an unmarked plain envelope (many of the notifications were inadvertently tossed out as junk mail) informing them they were no longer “in network.” This meant that tens of thousands of patients had to scramble care elsewhere, records had to be transferred, and prescriptions required rewriting by doctors who didn’t even know these patients. I’m sure many of you have been through this.
The main criterion for a physician being dropped from a network is always (always always) a financial one. The physician or her group is red-flagged by the insurance company as an “over-utilizer.” This means too many diagnostic tests, too many referrals to specialists.
In other words, just when we learn that in order for physicians to be more effective they need to be willing to order more tests and refer patients to colleagues or specialists, BANG! They’re offed by the Godzillas of the health care system. It’s already started with UnitedHealthcare, and the other giants are watching the mounting legal battles (restraining orders have been filed by medical groups against United), but believe me, this is the future.
And that’s the news from Lake Wobegone. What do you think? Your comments are invited.
David Edelberg, MD
PS: I was mightily pleased to learn that one of the solutions suggested to enhance physicians’ diagnostic expertise was something I’ve been doing for years (you who are my patients can affirm this). It’s called Point of Care Learning and what it translates to is having immediate help from both colleagues and the internet. I probably say to my patients “Give me a second while I look this up…” at least a dozen times daily.
As a quick and personal example, although I wasn’t part of the survey, I did tackle the cases that were presented to the doctors in the article. I correctly diagnosed the two easy cases, but admittedly joined the 95% who were stymied by the two difficult ones. However–and to me this is a big however–with the help of good old Google in about two minutes I got enough information to make (with pretty good confidence) the likely correct diagnosis of both.