IN THE 1970s, H. Gilbert Welch drove an ambulance as a college job in Boulder, Colorado, often blaring out Elton John鈥檚 Someone Saved My Life Tonight. Wanting to save lives led him to study medicine, but he came to realise that saving lives wasn鈥檛 as clear cut as he thought. Sometimes, he found, it can be better to do nothing.
Welch became a physician and academic researcher, and he has spent the last 25 years warning of the dangers of overzealous medicine. He worries that doctors are detecting problems too early, convincing healthy people they are sick, and treating them too aggressively.
Advertisement
His latest research, published in December in the Journal of the American Medical Association, is a case in point. He has found that in US hospital regions with high rates of CT scans 鈥� which are typically ordered to check the lungs and abdomen 鈥� . So what is going on? When doctors look at the images, they can see the kidneys too, and often stumble on innocuous cancers, says Welch. 鈥淚t鈥檚 leading some people to be treated for disease that was never going to bother them.鈥� And at significant risk: 1 in 50 of those who underwent the surgery died within a month.
A professor at the Dartmouth Geisel School of Medicine who only stopped practising medicine five years ago, Welch has written , as well as dozens of journal articles and . He travels the globe to speak to fellow doctors and researchers. With biomedical companies designing ever more tests, such as breath-tests for cancer, the problem seems poised to worsen. 鈥淚t鈥檚 a very frothy industry right now,鈥� says Welch.
TheJAMA study was inspired by a patient we will call Robert, who came to Welch at a Veterans Affairs medical centre in Vermont, complaining of lingering hoarseness. Welch referred him to a specialist, who found a small tumour on his vocal chord. The tumour was removed and his hoarseness went away. Then Welch had to call Robert back. Somewhere along the line, a CT scan had been taken of Robert鈥檚 lungs, which showed his chest was fine but revealed a cancer in his kidney. This was, in medical terms, an incidentaloma. 鈥淗e was just so funny about it,鈥� Welch recalls. The urologist wanted the kidney out, and Robert said to Welch, 鈥淐鈥檓on, you鈥檙e kidding me, doc. You just did surgery in my throat and now you鈥檙e going after my kidney? Let鈥檚 you and I talk about this.鈥� So Welch challenged the urologist. He followed the cancer for 10 years with imaging; it stayed the same size. Robert eventually died of pneumonia.
Vanishing cancers
鈥淚 was taught in medical school that once a cancer was formed, it was going to relentlessly progress to metastatic cancer,鈥� says Welch. 鈥淲e now know it鈥檚 a whole lot more complex than that.鈥� Cancers can grow quickly and slowly; some even vanish on their own. There are the bird cancers, which have already spread before tests notice them; the rabbit cancers, which can be treated before they spread if caught early; and the turtle cancers, which never spread. The problem, says Welch, is 鈥渢here鈥檚 a whole lot of turtles out there鈥�, but doctors and patients alike want to treat all cancers.
A new test that worries Welch is liquid biopsy, which identifies pieces of 鈥渃ell-free DNA鈥� in the blood to determine whether someone has cancer, and how bad that cancer is. 鈥淵ou think, 鈥楬ow could you possibly argue with that?鈥� until you look under the hood,鈥� says Welch. We all have cell-free DNA in our blood, and liquid biopsy analyses about 2000 different mutations in this DNA. An algorithm then determines what thresholds and combinations of mutations equal cancer. Welch worries about a future in which people are told: 鈥淵ou have a positive liquid biopsy, but we don鈥檛 know where the tumour is, so we鈥檙e gonna have to start looking.鈥�
Richard Baker, a radiologist in Madison, Wisconsin, worked with Welch at the Veterans Affairs centre. As a result of Welch鈥檚 influence, and against his own financial interest, Baker often dissuades his patients from getting a biopsy on their thyroids after imaging has found a nodule, even though that is why they are seeing him. 鈥淭hyroid biopsies are skyrocketing in this country,鈥� says Baker, yet , and treatment carries risks of its own. 鈥淭hese are difficult ideas for both patients and physicians to accept,鈥� he says.
鈥淵ou just did surgery in my throat and now you鈥檙e going after my kidney?鈥�
In 2016, that screening in the US had found many more non-progressing breast cancers in the 20 years up until then, but helped very little in catching fast-progressing cancers early on. In looking at women who were screened every year for a decade from the age of 50, he found that for every 1000 of those women, roughly one will avoid death through breast cancer, more than 500 will have at least one false alarm and 10 will be treated needlessly.
Welch . Taking this sort of position doesn鈥檛 win popularity contests, and Welch decided early on to direct all profits from his books to charity to avoid the criticism that he is making provocative arguments to cash in.
鈥淸Welch] has had an enormous negative impact on the practice of medicine,鈥� says Daniel Kopans, a professor of radiology at Harvard University. He disagrees with Welch鈥檚 research on a number of methodological points, and on his larger conclusions as well. 鈥淎ddressing overtreatment by stopping screening is like removing the engines from our cars to stop automobile accidents.鈥�
Kopans believes in the life-saving good of mammography, and he isn鈥檛 alone. Likewise, many healthcare providers stand in Welch鈥檚 camp. One side emphasises the lives saved by mammography. The other side puts more weight on the very common post-mammography anxiety women experience as they wait for a biopsy of a suspect mass, and on the risk of undergoing chemotherapy for a cancer that would have gone forever unnoticed. Welch thinks women should have mammography鈥檚 risks and benefits explained, then be encouraged to choose for themselves.
When Welch began practising 30 years ago, the suggestion that screening was responsible for overdiagnosis was a radical one. Now, thanks to the work of Welch and his ilk, the debate isn鈥檛 whether overdiagnosis occurs, but how big a problem it is.
Welch suggests it is time we reassessed what medicine is for. 鈥淒o people want medical care as a way to deal with acute problems; things that are bothering them? Or do they want to take the power of medicine to look hard to try to find things wrong with them?鈥� he says. Because in this age of super-sensitive diagnostics, seek and ye shall find.
This article appeared in print under the headline 鈥淟ook too close and we鈥檙e all sick鈥�
