This piece, “Sensitive or Specific?", appeared on page 21 of the Fall 2024 issue of Chicago Life Magazine.
Too much or too little medical testing is bad for you. Too much leads to excess intervention, cost and worry. Too little means missing important diagnoses. Sensitivity and specificity are powerful statistical tools that help make such decisions.
It can be viewed using this link: Chicago Life Magazine - Sensitive or Specific?
--
“Just do it every time you pay your phone bill.” That was my standard advice to female patients on how to remember to do their monthly self-breast exam (SBE), which I nagged about over the course of a couple of decades of medical practice. In 1950 the American Cancer Society recommended that all women, starting in high school, do a methodical self-exam every month. This routine was expected to catch lumps early in the course of their possible progression to breast cancer, resulting in better prognosis, including a higher cure rate. It makes perfect sense. Find the pathology early, before it's spread, and treat it by appropriate means--surgery, chemo and/or radiation.
By 2003 the Cancer Society guidelines had made the self-exam optional. They dropped it altogether in 2015. The argument against doing a monthly SBE is straightforward. In medicine good evidence should always trump even the most reasonable-sounding conjecture. Research papers started appearing in the 1980s and ‘90s that compared the actual course of breast cancer in women who performed a monthly SBE to those who didn't do it. The research showed that women who'd discovered lumps by doing an SBE fared no better, in terms of stage at diagnosis and all other outcomes, including longevity, than those who had encountered breast lumps in the normal course of life or whose doctors had found them on formal exam. What the SBE women did experience though was more procedures, with their attendant pain, inconvenience, complications, cost and worry.
Please keep in mind that you must consult a clinician right away if you ever do find a breast lump and it doesn't disappear with your period.
Data is just the first step in a discussion about screening recommendations. How far to go to catch disease is not a simple question. There are tradeoffs.
A crucial category of tradeoff is between sensitivity and specificity. Diagnosing diabetes based on blood sugar level provides a clear example. If you choose a high cutoff point for normal blood sugar it means that when you say a person has diabetes, based on a glucose reading that is over the mark, it is likely they really do have the disease. So it's a pretty specific test. But how about all the folks whose glucose intolerance you'll miss because they don't have such high numbers? By setting the cutoff point for normal too high the test may not be sensitive enough to identify a lot of people who ought to have attention paid to their blood sugar.
So you specify a lower blood sugar level for diagnosing diabetes. In 2021 the National Institutes of Health actually chose a number so low that it labeled 38% of US adults as at least pre-diabetic, a clear example of turning the sensitivity dial up way too far. Not surprisingly, the American Diabetes Association changes its recommended cutoff points for blood sugar screening sometimes as often every few years.
Every medical test is subject to analysis of its sensitivity and specificity, which are calculated employing standard statistical methods. Where to set a balance point between sensitivity and specificity when designing, evaluating and implementing a test is ultimately a matter of judgment. What is the value of detection and how much will it cost in monetary and human terms? Though mathematics can help to quantify consequences, the actual decision-making depends on value, a consummately human judgment.
The self-breast exam ultimately presented little dilemma. It yielded way too many benign diagnoses at the cost of expensive, painful and worrisome procedures, without actually improving outcomes for malignancies. That is, the SBE was far too sensitive and, it follows logically, far too non-specific to continue recommending it.
American medicine is largely biased toward sensitivity and away from specificity. The prevailing mentality makes it a priority to find pathology and treat it, with lesser regard for cost in dollars, life disruption and worry. Money is, unfortunately, one of the main motivators in the direction of sensitivity. The more positive screening results, the more diagnostic workups, procedures and follow-up, all of which, in our mostly fee-for-service medical economy, translate to more revenue. The profit motive is pervasive and unseen, sort of like water is to the fish that swim in it. People in healthcare rarely will admit, including to themselves, to what extent their medical decisions are influenced by money.
There are tons of screening programs whose bald-faced purpose is to generate business, with little real regard for value. The total body scans they advertise on TV are a great example of revenue-driven services posing as real medical care. Repeated studies have shown that this sort of imaging on healthy people turns up life-changing serious diagnoses exceedingly rarely. But all too frequently it does start patients down twisty, unnecessary paths of more testing and anxiety and unfailingly lightens their wallet by, at a minimum, the cost of the scan.
Even without the profit motive American medical culture embodies a pervasive prejudice toward doing more, leaning ever more heavily on technology and intervention and less on the sorts of human relationships that reveal values. In healthcare, as in every other human endeavor, it's ultimately values that ought to drive things.
I seem to have wound up where I always do when I talk about medical care, describing how the tapestry of good care must be woven with a weft of modern science on a warp of understanding individuals, their communities, cultures and values. Tools such as sensitivity and specificity yield important insights that help medical decision-making. Applying them to accomplish what is best for patients is where the consummately human art of medicine comes in.
With the self-breast exam a thing of the past, I still hate to waste the monthly phone bill prompt. How's this instead? As you make your payment, remind yourself of something you are grateful for, like all your body parts that still work well. And men should do it too.
