I Am Afraid of Early Cancer Detection: Part III

Peter Attia, national guidelines vs. personalized care, and what I can actually tell patients.

I can’t stand celebrity doctors.

That said, I am kind of a fan of Peter Attia. This is an accomplishment for Dr. Attia because one of my criteria for calling someone a celebrity doctor is that I can’t stand them. I appreciate Dr. Attia because he’s a smart guy who seems much more interested in improving people’s lives than hawking his wares. He also spends an admirable amount of time on the un-sexy things that really matter like sleep, exercise, friends…

I’m writing about Dr. Attia because a dozen readers referenced him in comments and emails about the first two articles in this series. I really appreciated these messages because Dr. Attia’s enthusiasm for cancer screening is a good counterbalance to my own anxiety. Dr. Attia understands the issues with population data but thinks that, on an individual patient level, screening makes sense. Dr. Attia has written and said a lot about cancer screening. He posted a nice article in January on his website that summarize some of his views. His article was a response to a meta-analysis published in JAMA IM that presented pretty depressing data concerning the effect of cancer screening.[i] I’m going to outline what I think are Dr. Attia’s strongest points from this article

Dr. Attia writes “…while the risks of possible harm should be weighed against the potential benefits for the individual, the best chance of not dying from cancer comes from catching cancer early through screening.” This is the common sense rational for cancer screening.

He acknowledges that we have little to no data that our currently accepted cancer screening tests reduce all cause mortality. However, he points out that “using ACM (all-cause mortality) to determine the success of cancer screening is a nearly impossible task, at least over the period of time such studies are done.” He goes on to calculate what kind of study we would need to demonstrate a benefit in all-cause mortality with mammography.

---

“For example, if the true effect of breast cancer screening relative to no screening is a 25% reduction in breast cancer mortality, a trial would need to have 96,000 participants to show a significant reduction in breast cancer mortality, but the same trial would need more than 1.7 million participants to show a reduction in ACM, a number of subjects not achieved even by pooling multiple trials in the meta-analysis.”

---

Going further, Dr. Attia argues that our studies of screening tests do a poor job of finding the benefit of screening. He dives into the debate on whether screening studies should prioritize intention to treat (ITT) or per protocol analyses.[ii] RCTs use ITT analyses because they maintain randomization, closely mimic real life adherence, and make it harder to show a difference between the groups analyzed. Dr. Attia argues that when we are trying to find out if screening works, we should look at people who actually got screened, rather than those to whom screening was recommended – a so called per protocol analysis.

I think Dr. Attia gets all this right, but I am still afraid of cancer screening.

From a population standpoint, I think the evidence is clear that cancer screening does not make much sense. His arguments support this conclusion. The reason a mammogram trial would have to be so large is that the benefit is so small. The reason a per-protocol analysis would be needed is because so many people don’t follow through on screening recommendations.

Dr. Attia’s argument for screening at the individual patient level is strong. There is little doubt that cancer screening saves individual lives. To deny this being unreasonable.[iii] I have patients whose lives I have almost certainly been saved by convincing them to get cancer screening. I have patients in whom I have diagnosed metastatic cancer who would probably still be with us had they gotten “recommended” cancer screening. When you argue against routine cancer screening these “facts” need to be acknowledged. (I know, I know you can argue that we don’t know the counter-factual in these cases, but I am asking for a reasonable take.)

My fear persists because I also know, in an individual patient, the truth about cancer screening. Most people who are screened for a particular type of cancer will not benefit because most of the people will never get, never be harmed by, and/or never die from the cancer for which they are being screened. Individual types of cancers are, fortunately, relatively rare. I also have seen the harm I have done with cancer screening. People who have suffered complications of the tests, people who have been treated for cancers that I am not sure needed treatment. (I know, I know, the counterfactual argument holds here too.)

Our medical ethics dictate that people should make their own medical decisions once they are appraised of the balance of harms and benefits. From my experience, however, having advised probably thousands of patients on cancer screening, this does not work. People do not hear risks and benefits. During the time I discussed prostate cancer screening with all men over 50, I worked out a perfectly balanced spiel on the risks and benefits of PSA screening. Ninety nine out of 100 patients responded to my pitch with, “so Doc what should I do?”

Once a screening test comes back positive, no matter how prepared people have been to the possibility of false positives, they are devastated.

And it is difficult (very difficult) for people not to have cancer treated even if their doctors recommend observation. “Just get it out of me,” is the all too common a response. It is not only the patient these decisions effect. I have had patients who chose and tolerated “watchful waiting” or “expectant management.” When definitive treatment was finally recommended, the patients’ partners told me they were thrilled to not have to live with the stress anymore.

The complete and honest truth of what I can tell patients.

• The only thing I can guarantee about cancer screening is that there will be a monetary cost and an “opportunity cost” related to the time you spend doing the test and any recommended follow-up.

• There is a very small chance this test will save your life. (For breast cancer this might about 1/1000 if you get a mammogram every year for 10 years). Even if your test is positive and we cure a small cancer or remove a precancerous lesion we will never know whether this screening test saved your life.[iv]

• Screening could also cause harm.

  • There is a very small chance you could be harmed directly by the screening test.

  • You might also suffer from anxiety related to the test and the results. This anxiety will be especially aggravating if you get a false positive result.

  • The test could also lead to overdiagnosis, the diagnosis of a cancer that does not need to be treated. (The chance of this is about 6/1000 for 10 years of mammograms).

  • You could be diagnosed with a cancer that needs to be treated but whose early treatment provides no benefit. This means that you would have the same outcome had we waited until you developed symptoms from the cancer. In this case, the only thing that screening has done is make you live longer as a “cancer patient.”

Before I leave this topic, at least for a while, I wanted to reference two previous Sensible Medicine posts, one video, and one Sensible Medicine adjacent podcast.