There’s been an awful lot of Internet hullabaloo about “unnecessary testing” lately. The Choosing Wisely program keeps trying to assert that we should not perform any “unnecessary” tests. Recently, a paper was published in the Journal Academic Emergency Medicine alleging that “overordering of advanced imaging may be a systemic problem” since many emergency physicians believe that such testing is “medically unnecessary.” The paper was based on surveys that were presented to emergency physicians and the work was at least partially funded by the Veterans Administration and the National Institutes of Health — which should be considered a conflict of interest. After all, don’t the entities that pay for the testing stand to gain to gain the most from limiting “unnecessary” testing?
But now that the paper has been published, the media has been whipped up into a frenzy, stating that doctors admit they perform “unnecessary” testing and are single-handedly bankrupting our medical system.
Consultant Magazine has an article stating that “ordering unnecessary imaging tests ‘may be a systemic problem.’”
NewsMax Health states that “97% of ER docs order unneeded tests out of malpractice fears”
HealthLeaders Media trumpets that “97% of ED physicians order unnecessary imaging tests”
Even Time Magazine has a headline stating that “your doctor likely orders more tests than you actually need.”
However, I’d bet my white coat that if you asked any of the study authors or any of the authors of the articles in these prestigious magazines what the definition of an “unnecessary test” was and how to prospectively determine whether or not a test is “unnecessary,” they would all look at you with blank stares and shrug their shoulders.
Unfortunately, defining the term “unnecessary” is more difficult than it seems. The American Heritage Dictionary defines “unnecessary” as being “not necessary.” In turn, “necessary” is defined as being “needed or required.” “Needed” is defined as “A condition or situation in which something must be supplied in order for a certain condition to be maintained or a desired state to be achieved.” This definition gives us a little bit of help, but is still vague as it applies to medical care. When discussing advanced medical imaging, a necessary test would be that which must be performed so that a “certain condition can be maintained or a desired state can be achieved.” Ruling in or ruling out a disease process would seem to fit that definition.
Based on the paper’s abstract, it appears that almost 500 emergency physicians were given a survey and 97% of them stated that at least some of the advanced imaging studies (CT scans and MRIs) they personally ordered were “medically unnecessary.” In other words, the researchers took a politically charged statement and, using a fallacy of definition, created a statistic which is in itself both untrustworthy and sensationalistic.
The questions used in the survey aren’t available. If you want to look at them, you can “rent” the article for $6 or you can splurge and purchase the article for $38 – neither of which I’m planning to do. However, consider the questions that would be raised if the researchers asked 500 people whether they were “good parents” and then published a study saying that, based upon survey data, bad parenting did not exist in the United States. Or consider a study asking doctors if they acted in “professional manner” and then declaring that unprofessionalism does not exist in medical care because 97% of doctors answered “yes.” Those studies would get laughed out of Medline because it is easier to see the bias in asking people ambiguous questions when the terms of the questions haven’t been defined. That’s why I’m very surprised that some of these authors would put their names behind a paper with such dubious conclusions.
Shortly after the JAEM paper was published, Lenny Bernstein then published an article in the Washington Post noting that patients with low back pain who were first sent for MRIs instead of first going to physical therapy paid an average of almost $5000 more for their medical care. The reason was apparently that MRIs tended to show all kinds of “benign changes” in the patient’s backs. The article is based on a study in the Journal “Hospital Services Research.” Again, it will cost you $6 to rent or $38 to own this piece of research. It would be nice to know whether or not the authors, two of whom were physical therapists and one of whom was a researcher in “Clinical Quality and Outcomes Research”, discussed whether there were any adverse outcomes associated with proceeding directly to physical therapy as opposed to obtaining advanced imaging. It is difficult to draw any conclusions from the abstract other than physical therapists are advocating more physical therapy for low back pain.
Finally, also published right around the same time as the above two articles, there was an article in the Los Angeles times noting how Kaiser Permanente was ordered to pay a young woman more than $28 million after delaying an MRI that could have detected an aggressive cancerous tumor. According to the article, the patient was 17 years old and experiencing severe lower back pain. She and her mother repeatedly requested an MRI of her lower back, but the patient was instead told that, at 125 pounds, she had too much “belly fat” and needed to lose weight. After three months, Kaiser ordered the MRI and found a large fast-growing cancer in her pelvis. Ultimately, the patient required amputation of her right leg, removal of half her pelvis and part of her lower spine in order to remove the tumor. The patient’s lawyer argued that her leg and pelvis could have been saved had the MRI been performed and the cancer diagnosed earlier. Kaiser argued that the cancer was already so large that the patient would have lost her leg even if they had ordered the MRI sooner.
Admittedly, it is difficult to determine whether care was appropriate based upon an incomplete clinical picture, but I think we can safely assume that the patient falls into the “don’t image” guidelines that ACEP has created for the “Choosing Wisely” initiative (see guideline #8).
Note that the Choosing Wisely guidelines for acute low back pain are different than the low back pain “red flags” that are traditionally considered an indication for diagnostic imaging. This patient had at least two “red flag” symptoms.
The Kaiser case also raises an issue as to whether or not the jury award was unreasonable and based upon emotion rather than based upon the patient’s injuries. However, consider that the award must provide for lifelong care of a young adult who had to have her leg amputated, half of her pelvis removed, and part of her spine removed. She’s going to require quite a bit of care and is going to go through a lot of suffering for the rest of her life. This case took place in California where there are caps on the “suffering” or “non-economic” portion of the damages. We don’t know how much of the award was based on future medical care and punitive damages (neither of which are subject to the caps) versus noneconomic damages which would be capped at $250,000. If the patient was able to introduce evidence that Kaiser had a pattern of engaging in denial of care to increase profits as the article suggests, punitive damages would be more likely to be imposed.
The thing I found most interesting about the Kaiser case was that Kaiser did not argue that the MRI of the young patient’s lumbar spine was an “unnecessary” test. Nobody in the comment section of the article argued that the MRI was an “unnecessary” test, either. Why? Because the results of the test were positive and anyone who argues that a test is “unnecessary” when it shows gross abnormalities needing immediate therapy would be viewed as an idiot. Had the exact same MRI in the Kaiser case been normal, everyone would have rolled their eyes, shook their heads, and proclaimed what a waste of money it was to perform the test. However, because the MRI was grossly abnormal, the consensus is instead that the test was not only “necessary”, but that it should have been performed much sooner.
This concept underscores why the JAEM article is so misleading. The authors don’t adequately define the terms on the surveys that they provided and, as a result, the conclusion they base on those ill-defined terms do not pass scrutiny. When we define the utility of a test by that test’s results, we engage in medical mumbo-jumbo which neither improves the health of patients nor improves the practice of medicine.
It sure makes for some great “unnecessary” headlines, though, doesn’t it?