When Doctors Cry “TMI!”

Robert Hanlon

Recent stories have highlighted genetic testing that may predict an 87% likelihood of getting a certain rare type of cancer.  These reports provide a tantalizing, if scary, hope of informed decision-making about medical treatments.  But don’t be fooled.  To date, only a very small percentage of the population and a very few types of cancer can benefit from such sharp analytic tools.

The vast majority of discussions in the health care community run exactly in the opposite direction.  Most of the news reflects a dramatic trend toward cutting back on certain preliminary diagnostic tests.  In particular, mammograms to screen for the more common forms of breast cancer and PSA tests for prostate cancer are the principal screening devices available to most individuals.  These have come under mounting criticism as being poor indicators of disease and not very effective in reducing death rates.

Many scientific groups and guardians of health care costs have responded to these observations by calling for the expansion of the time between the administration of these tests, or by choosing a later start or stop age for their use, or even eliminating them altogether.  While there are some real costs to these screening devices, economic and personal, one of the main rationales for the proposed reduction in use is that screening leads to the widespread administration of secondary testing.  Those follow-up tests are more costly, more intrusive and carry their own health risks, such as possible infections, exposure to radiation and other complications. 

Even more importantly, the critics argue that widespread use of screening devices leads to the overuse of full blown treatment such as surgery, radiation and chemotherapy, in many cases, on patients who either do not actually have the cancer sought, whose cancer is not likely to be life threatening or debilitating, or who may not be helped by the treatment provided.  The solution to this problem, they say, is to drastically reduce the use of the initial screening techniques and thus reduce unnecessary further testing and treatment.  In effect, the screening gives us data we don’t want or need – too much information.


Screening tests do not inevitably compel more testing or mandatory treatment.  Decisions by physicians and patients drive the next steps.  Those decisions can be sound, based on medical evidence in hand, family history patterns and statistical probabilities.  Or they can be rash and ill-informed, based on fear, loathing and a please-don’t-sue-me mindset.  It isn’t the weak data that is leading to overtesting and overtreatment, it’s weak decision-making.

The screening tools for many cancers, such as mammograms, PSAs and similar analyses, are crude tools at best.  They have numerous false negatives and false positives.  But they either have some diagnostic value or they have none.  If they have some utility, until better screening techniques are available, we need to use them, but we need to use them smarter.  If they have absolutely no utility, we need to admit it and move on.

Let’s look at an example.  In a recent New York Times article questioning the routine use of pelvic exams for the screening of uterine cancers and other gynecological issues, a physician described the outcome of his exam.  “In my experience as a practicing gynecologist, I frequently have had to take patients into the operating room because I found an enlargement during a bimanual pelvic exam.”  The doctor then comments, “I then follow up with a sonogram which shows a mass, but I can’t tell what the mass is without surgical exploration. Yet nearly always it’s benign.”  Have had to take patients into the operating room? Sez who?

What’s missing in this report is the concept “then the patient and I have to decide what to do next.”  An enlargement in the patient’s abdomen is either there or it is not.  If it is there, what are the odds that it is dangerous enough to demand further action?  That may depend on several factors – the patient’s own age, health and personal history, the family risk factors, the overall risks in the general population and perhaps several other factors (geography, occupation, etc., etc.).  These factors should inform a decision as to whether additional testing or surgery is advisable. 

Let’s assume that, before the exam, if a mass were detected, those factors make it 75% likely that the patient has a serious disease well treated by surgery.  Most doctors and patients would decide to proceed aggressively.  Perhaps those factors combine to make the risk 1%.  Most would choose to avoid dramatic intervention.  The cases in between (admittedly the majority of them) are tougher calls, but the decision process is the same.  How much risk will you tolerate before you decide to go the heavy intervention route, or even one step further down that path?  Those are decisions that patients and doctors make every day and should be making.

But what if we don’t do the exam?  The mass is still either there or it’s not.  The benefits of intervention, or lack thereof, based upon the patient’s risk factors, are the same, whether or not the exam is done.  All that’s lacking is the data point of whether or not there is a mass.  How can that lack of information improve the patient’s life?

Many of the arguments defending the avoidance of the information relate to patient anxiety and bad decision making.  “There’s a one in a million chance I have cancer? TAKE IT OUT! TAKE IT OUT!”  Also justifying questionable intervention is the occasionally voiced fear of malpractice claims.  “If I find anything, I’m going to have to order every test and treatment in existence so she can’t say I didn’t do EVERYTHING”.  This is simply abdication of responsibility on the parts of both the patient and the clinician.  Sensible doctors (and patients) reject the “give-me-antibiotics-even-though-you-are-sure-it’s-a-virus” approach.

One of the “solutions” proposed to deal with less-than-definitive screening is to delay the onset and reduce the frequency of testing.  This would inevitably save some money and lessen the exposure to uncomfortable tests, some of which many carry a small risk of their own.  However, this does nothing to improve the diagnostic value of those measures.  To the contrary, the delay in starting postpones establishing the baseline measure that is often the most useful tool in making determinations as to whether there is a change in the underlying structure or level being measured.  For example, many scientists believe that a change, and especially the rate of change, in PSA are better indicators of the existence and the possible aggressiveness of prostate cancer than a single supposedly “high” reading.  Fewer or later data points don’t increase good decision making, and may actually impede a rational decision process.  This doesn’t mean that, once baselines or patterns are established, the frequency of repeat testing might be extended, especially for conditions that are generally slow to develop.  But first, we need to try to distinguish those patients who are at high risk for a rapidly developing condition from those who have time to watch and wait.

Recent breakthroughs in the identification of genetic markers that may signal the likely incidence and aggressiveness of cancers are welcome additions to the diagnostic toolbox.  These will be especially useful in helping doctors and patients engage in an informed decision making process about further testing and treatment.  Some of these may be good enough to modify, or even eliminate, our reliance on some of the current screening methods.  These are still few and far between.

But let’s be logical.  More data is just that – more data, not “too much information”.  It’s what we do with the data that matters.  It’s time to develop better and braver methods of talking about the real risks, rewards and benefits of each test and treatment before embarking on an eradication juggernaut.  It’s also time to reject an ostrich-like approach and cries of “TMI”.


Robert Hanlon is a seven-year “survivor” of Stage III prostate cancer, numerous PSA tests and digital exams, a biopsy, a radical prostatectomy, radiation, chemo, and hormone therapies. He believes he owes his life to good decision-making on the part of his wife, his medical team and, to a lesser degree, himself.