Monday, April 11, 2011

JAMA hides its review of Overdiagnosed

I wonder if the people at JAMA are ever going to understand that the journal would have even more influence if they made articles of a general nature free and available for all to read. Especially when the article makes a recommendation to the general public.

Here is one example, a thoughtful book review in the April 6 edition written by Dr. Leonard Berlin. The topic is Overdiagnosed: Making People Sick in Pursuit of Health, by H. Gilbert Welch, Lisa M. Schwartz, and Steven Woloshin.

Here is the part of the review that JAMA lets you see:

A little-noticed but nonetheless profound evolution in medical practice took place in the latter half of the 20th century. Until that time, individuals visited their physicians only when they experienced symptoms and worried they were ill. Now they began to visit their physicians to undergo testing to detect occult disease, even if they felt healthy and were asymptomatic. A new mantra worked its way into medicine in the United States and was quickly adopted by the public: get tested, diagnose disease early, and be treated while the problem is “small” before it becomes “big” or, even better, before a potential disease becomes a reality. This was, and still is, the right thing to do—correct? Well, maybe not, claim [the authors]. Such testing of healthy individuals all too often results in overdiagnosis and overtreatment, because in reality there is “nothing to fix.”

The next sentences are: As a result, contend the authors, although a few will be helped, many will be overdiagnosed, and some will be harmed. This theme permeates the book.

This is very important stuff. A scientist friend reminds me that, as a general matter, the predictive value of a screening test is mathematically related to the prevalence of the disease in the screened population; so the less prevalent, the tougher it is to have a high predictive value.

You can get a good sense of the book itself from these excerpts published in the New York Times. (Hey JAMA, please note: They include way more than 150 words!) But, let me give you some more excerpts from the book review:

That the prevalence of overdiagnosis has reached epidemic proportions, according to the authors, is attributable to the changing guidelines regarding numerical values of what is considered normal and abnormal as well as to technological advances in radiologic imaging.

According to data documented in the book incidentalomas are found in the lungs of 50% of otherwise “healthy” persons undergoing computed tomography of the chest; in the kidneys and livers of 23% and 15%, respectively, of those undergoing computed tomography of the abdomen; and in the thyroid glands of 67% of those undergoing ultrasound examinations of the neck. The chance that the incidentaloma could represent a lethal cancer is considerably less than 1%.

In all of these areas, the authors argue that the quest to find potential disease which may not exist or, if it does, will never harm the patient often leads to unnecessary surgery and medication that may cause serious adverse effects as well as patient anxiety and ever-increasing costs.

And the important conclusion:

Overdiagnosed is a provocative, intellectually stimulating work. As such, all who are involved in health care, including physicians, allied health professionals, and all current or future patients, will be well served by reading and giving serious thought to the material presented.


Anonymous said...

How right you are

Anonymous said...

This is just one more symptom, Paul, illustrating the point that physicians do not yet see. It is aptly explained by a translated retweet from e-patient Dave:

MT @bettinepluut: "Patiënt ctrd care is not the route to a point - it IS the point."

Indeed. Now, I am not going to get into the mess about paid subscriptions to journals but, at the very least, the JAMA should make its content which is directed at the public - which is patients after all, correct? - freely available to the public. Even the NEJM has taken some steps in this direction.

It is time to truly rethink EVERY aspect of our care in the context of 'patient-centered'. Thanks for bringing this to our attention.

nonlocal MD

Barry Carol said...

Since my employer has a subscription to JAMA, I was able to read the whole article, which is only one page long in any case. What I wonder about is who was responsible for lowering the threshold for hypertension from 160 / 100 to 140 /90 mm Hg, the blood glucose threshold for diabetes from 140 to 126 mg/dl, the cholesterol threshold for hyperlipidemia from 240 to 200 mg / dl and the T-score threshold for osteoporosis from
-2.5 to -2.0? All of these changes transformed millions of patients previously considered healthy into patients needing medication or other treatment. Do other countries use the same standards? If not, why not? Also, this seems like an area where tort reform would be helpful in protecting doctors from lawsuits based on the failure to diagnose a disease or condition. At the same time, I can understand that patients might want tests that are neither painful nor invasive if they might either help them or provide reassurance even if the tests are not cost-effective at the population level. After all, someone else is paying the bill, at least for the most part. Perhaps more patient “skin in the game” would be helpful as well.

Anonymous said...

What the medical industry gets, that public health advocates often do not, is that individual interests in preserving the somatic representation of themselves is a far greater incentive than fair or wise allocation of public and medical resources. If I get a .5% edge on cancer, it is in my interest to get screened. A little pain and discomfort to avoid cancer? Anxiety? How about the anxiety of telling your family you are dying? Doctors may experience patients with cancer daily, but people who experience the pain of watching someone they love die with the disease have a valid experience to honor in our calculations as well.

Risk estimation and human psychology mismatch is well documented. If we want to sell overdiagnosis as a problem, we'll need to make a stronger individual harm argument. It could start with honest informed consent that includes patient education. For example, radiation exposure is often reported without a discussion of cumulative effects. Physicians also could report how many surgical infections their department had last year, and what were the patient outcomes. But how is it in their interests to do so? Industry knows that they do not need to make an intellectual argument, just one that hits the heart. Otherwise, no orphan drug would ever be developed.

If the science of medicine wants better impact (closed circulation journals? 17 years bench to bedside, anyone?) they might aim to educate the very frightened aging population who continue to go to the polls. That will have a bigger influence on health resource distribution than science. This is something industry already knows.

Marya Zilberberg said...

Archives of Internal Medicine is doing a "Less is more' series to address this point exactly. My letter this week shows exactly how much "skin in the game" a patient has when getting unnecessary testing. The letter is found here: