Dr. H. Gilbert Welch: “The More Tests You Do, The More Likely One Of Them Will Be Falsely Abnormal”

It’s that time of year, time to revist this sage advice:

H. Gilbert Welch (MD, MPH, Professor of Medicine at Dartmouth University, author of Less Medicine, More Health, and Overdiagnosed: Making People Sick In The Pursuit Of Health) says we’re testing too much … blood tests, colonoscopies, endoscopies, mammograms, ultrasounds, MRIs, CT scans, biopsies, gene tests. He says all this testing isn’t making us any healthier. To the contrary, by digging incessantly for some tiny abnormality – abnormalities that may never harm us, abnormalities that BigPharma and BigHospital and BigMedicalEstablishment can treat – it leaves consumers with unnecessary stress and undesirable side effects, not least of which is the pain in their wallet.

Mark Cuban, owner of the Dallas Mavericks, started a firestorm with this tweet:

Comprehensive blood tests every 3 months. In a healthy person. Wow.

Charles Ornstein (CO), a senior editor for ProPublica, interviewed Dr. Welch about that tweet:

Mark Cuban’s Advice A ‘Recipe For Making All Of Us Sick,’ Expert Says, ProPublica, 6 April 2015

Some excerpts:

(CO) Is quarterly blood work for healthy people a good thing?

No. It’s not. This is potentially a recipe for making all of us sick.

(CO) Explain.

I guess the first thing to say is that we all harbor abnormalities, and increasingly our technologies are able to detect them–be they biochemical, be they structural. We can see things down to millimeters in size; we can measure things down to parts per billion; and we can sequence the whole genome. That’s 3 billion data points.

So there’s no shortage of biometric data that people could be collecting on themselves regularly, and by the way, there’s a huge financial interest in having people do that. The market of the well is a huge, huge market.

The problem is you’ll always be catching things out of what we would say is normal. This is anticipatory medicine at its worst, where you’re really focused on what could be going wrong in the future and you’re trying to pick up [a] signal.

The problem is there’s so much noise, because the human body is a living organism. Variation is the very essence of life. People will start reacting to this data. I also think it’s really important to label it what it is: data. To me it only becomes information to the extent that it accurately predicts something will happen in the future, and it only becomes useful knowledge – a higher level piece of information – if we can do something about it.

(CO) Cuban argues that any misdiagnoses and unnecessary treatment rests with the doctors, and not with patients owning their data. Do you agree?

I think there’s a misunderstanding that diagnosis is some super clear black-white kind of distinction, when in fact there are 1,000 shades of gray in between. The time you get into that gray is when you’re dealing with people who feel fine and have some detectable abnormality. That’s how we get into it in cancer screening. We’re looking for very early signs of disease. There’s going to be great pressure to react to those abnormalities.

It’s not fair to say the pathologists have misdiagnosed. Undoubtedly there is some misdiagnosis, but their standard for what constitutes cancer is the appearance of individual cells, and how they relate to one another – the architecture of the cells. That was a perfectly good standard when you were sending them cancers that you could feel, things the size of golf balls. But when you start sending them microscopic collections of cells, expecting them to make some prediction about the dynamics of that process and how that will interact with the host – that’s you – it’s understandably going to be fraught with uncertainly. It’s all going to be probabilistic.

(CO) Cuban makes a distinction between making a diagnosis, and collecting a series of data points to benchmark yourself.

The more tests you do, and this is only the statistical process, the more likely one of them will be falsely abnormal. And the more times you do it, the more chance that something will be falsely abnormal.

There will be great pressure to take actions and that’s how people will get hurt. It’s going to distract them from the more positive things that they can do now.

It gets down to what health is. What I’m worried about is allowing health to be defined as some set of biometric measurements. … Health is about more than a bunch of physical measurements. It’s about a state of mind and we have to be careful not to undermine that state of mind. Ironically, part of health is not being too focused on it. … Much better for people to develop good relationships, have good friends, be outside, eat well, find things that produce meaning in their lives.

I agree with Dr. Welch:

  1. The more tests you do, the more likely one of them will be falsely abnormal.
  2. Treating people who are well has become a huge market.
  3. Health is about more than biometric measurements.

From what I can tell, not a lot of people agree with Welch. People like to test themselves. As Paul Roberts writes in Impulse Society, we have become so inside ourselves, so self-focused and narcissistic that meticulous analysis of our every cell and secretion, every bit of bile, phlegm, feces, and blood, is all that matters.

But test-chasers are playing into the hands of businesses that are more interested in making money than in making sure people live to a ripe old age with all their faculties. How do they make money? The tests may be free (covered by insurance), but the treatments aren’t. Employers have been a key player in getting people onto the testing treadmill by requiring workers to have annual screenings, penalizing them if they don’t. It’s not because they care about their employees’ health, or at least not only. It’s because insurance companies offer employers plan discounts for getting people into the system. I mean, the more people you can test for cholesterol, the more statins you can sell.

5 thoughts on “Dr. H. Gilbert Welch: “The More Tests You Do, The More Likely One Of Them Will Be Falsely Abnormal”

  1. mboydp

    Overall I agree with this and want to read it more closely. For instance, this is my objection to 3-D mammograms: A) they detect minor defects that would never develop into anything serious but hence require further diagnoses, etc. B) they deliver far more radiation to the breast than 2-D digital; C) the radiologists don’t necessarily know what they’re looking at, so they biopsy it unnecessarily, and D) they provide more false positives than 2-D digital.

    Reply
  2. Bix Post author

    I don’t think people who are healthy with no symptoms should be getting colonoscopies. If you are concerned, test for blood in the stool or do a sigmoidoscopy first. They are less invasive, less risky, and a LOT less costly.

    Same with mammograms. This: “Breast cancer screening does not reduce mortality“:
    “The important result is that we do not find a beneficial effect of breast cancer screening [mammography] any longer.”

    I even think blood tests are getting out of hand. Dozens of markers … why? When the therapy ofttimes is lose weight, eat fruits and vegetables, take a walk. Why not do them first?

    Once you get started on the testing treadmill, it’s hard to get off.

    Reply
  3. mboydp

    On the other hand, a friend just went for her yearly gyne exam, which always includes a pap smear. She’s been fine for 6 years, but this 7th year they discovered aggressive uterine cancer, which is going to be treated w/ chemo, radiation, etc. Plus a complete hysterectomy, of course.

    Reply
  4. Magnolia

    It’s hard enough for me to get in for my yearly exam. I don’t want to spend any more time in a doctors office or a hospital so this sounds good to me. On the hand if you have genetic diseases in ones family maybe that needs to be checked out more often. Also you have to be suspicious if you are being conned.

    Reply

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s