April 2017

On Overdiagnosis

I highly recommend Dr. Gilbert Welch's book "Over-diagnosed: making people sick in the pursuit of health." In it he talks about patients getting caught up in a cascade of overdiagnosis. I once had a patient who went to the ER for an upper respiratory infection. A chest X -ray was done and it was negative except for a possible lung mass; the radiologist recommended a follow-up CT scan of the chest which I ordered. The CT scan showed a possible mass, so this time the radiologist recommended biopsy. At this point I referred the patient to- Pulmonology. The pulmonologist asked the interventional radiologist to do a fine needle biopsy, hoping to avoid an open biopsy; the result came back inconclusive. The pulmonologist now ordered a PET/CT scan to measure metabolic activity in the lung, and it was read as being consistent with a right lung cancer. Pulmonology now wanted to refer patient to a thoracic surgeon, and I assumed (wrongly) that this was for an open biopsy. I recommended Hershey over Johns Hopkins because the nursing was better; but the patient had done his research and he wanted to go to "the best"; at his insistence I referred him to Thoracic Surgery at Johns Hopkins.

On a busy day I got a call from Johns Hopkins Thoracic Surgery; it was from a resident who told me that the chief of Thoracic Surgery had taken out my patient's right lung and sent it to pathology; the pathologist found no cancer in it, only evidence of an old fungal infection. I was stunned and angry; angry with Hopkins, but also with myself for not calling a halt to the cascade. Apparently the pulmonologist and the head of Thoracic Surgery did not know that the PET /CT scan should not be used for cancer diagnosis, but I knew it. I had learned about it not from the medical literature which is dominated by the drug and medical equipment companies, but from an investigative (follow the money) article in the Washington Post. The PET /CT scanner was developed in 1976, but Medicare (CMS) refused to pay for it as it did not work for diagnosis. It took the company almost thirty years to buy enough influence in Congress to force CMS to approve its use. CMS approved its use not for diagnosis, but only for staging of cancer, i.e. following the course of the cancer through treatment. CMS had dragged its feet for decades because it knew that doctors, likely upon the advice of the sales reps, would soon be using it "off

label" to diagnose cancer. Based on the PET fCT, Hopkins believed that my patient had extensive right lung cancer, and so did not see the need for doing an open lung biopsy prior to resecting his lung.

Lessons? To paraphrase a recent story in Medical Economics, doctors know that hospital ratings "are not worth the paper they are printed on." The best doctors according to the ratings are such because they bring in large research grants, and that has nothing to do with being a good clinician. In fact, the more beholden to industry and its grants, the less clearly a doctor sees medicine. Overdiagnosis is just as dangerous as underdiagnosis, and doctors must learn to resist ordering unnecessary tests and ineffective screening that lead to a cascade of bad care.