In the real world, harms of lung cancer screening prove greater than expected

Mary Chris Jaklevic has freelanced for since 2016 and recently joined the staff as a full-time health care journalist. She tweets as @mcjaklevic.

Many positive lung CT scans turn out to be false alarms. Getty Images.

While hospitals and advocacy groups eagerly promote lung cancer screening to the public, some researchers are shining a spotlight on data that shows the harms of lung cancer screening occur more often than anticipated.

A whopping 56 percent of the time, current and former heavy smokers who were screened with low-dose computed tomography (CT) scans had false alarms that resulted in follow-up testing and procedures that can cause harm, according to data from a demonstration project involving 2,106 patients at eight Veterans Affairs (VA) academic medical centers.

That’s double the rate of false-positive results detected in the National Lung Screening Trial, which prompted a 2013 recommendation from the U.S. Preventive Services Task Force (USPSTF) that established criteria for screening.

Those false positives meant additional scans that exposed patients to more DNA-damaging radiation, which itself can lead to cancer, along with anxiety and sometimes invasive biopsies that risked complications such as infection.

Meanwhile, just 1.5 percent of those who were screened turned out to have lung cancer.

A new analysis of those real-world outcomes suggests some people who meet the criteria for screening might actually be better off without it.

Researchers used a prediction tool to divide those who were screened into five risk categories. They found screening was far less effective for patients with relatively lower risk, who had higher ratios of harms to benefits. For patients at the lowest risk level, nearly 7,000 patients had to be screened to prevent one lung-cancer death, and there were 2,749 false-positives for every lung-cancer death averted.

Call for a personalized approach

Tanner Caverly, MD

Anyone who meets the USPSTF criteria for screening, including those in the new analysis, has a history of heavy smoking. The task force recommends that screening be limited to those ages 55 to 80 years who have a 30 pack-year smoking history (the equivalent of a pack-a-day habit for 30 years) and currently smoke or have quit within the past 15 years.

And yet even among this generally high-risk group, “the balance of pros and cons is NOT the same for everybody,” said Tanner Caverly, MD, a general internist at the Ann Arbor VA Medical Center and co-author of the analysis, published this week in JAMA Internal Medicine, which advocates that “personalized, risk-based harm-benefit estimates” be used to help patients decide whether to undergo screening.

While some people for whom screening is recommended will benefit, Caverly said via email, “You also have a few people for whom screening is probably a bad idea, for whom screening is more likely to cause harm than good.”

He added, “And you’ll have a lot of people for whom screening is a tough call … because the benefit is smaller and the right decision is a matter of personal preference.”

Researchers also noted it’s still unclear whether the 20 percent reduction in lung cancer deaths observed in the clinical trial, along with nearly 7 percent fewer deaths overall, holds true in actual practice.

Marketing ignores downsides

Those are far more nuanced messages than the one often trumpeted to consumers, which is that screening saves lives, period.

Take the American Lung Association’s Saved by the Scan campaign, which features “screening success stories,” videos in which patients share unbridled positive experiences.

More recently, an advertorial from United Health Services, headlined “Lung cancer screening – it could save your life,” touts screening as a “noninvasive procedure using the latest technology, low-dose CT scanning.” There’s no mention of the downsides.

UHS provides a list of factors that can “increase your risk” including smoking, secondhand smoke, environmental exposure, and family history — suggesting those things might be enough to warrant screening.

As noted previously, the USPSTF criteria for screening are much more stringent and specific than this. They apply only to people over age 55 who have a history of heavy smoking.

Further, the task force said even high-risk patients should be excluded from screening if they have health problems that substantially limit their life expectancy or preclude them from having lung surgery.

Why false-positive rates can vary

Several factors could account for more frequent false alarms among VA patients compared with the clinical trial that informed the task force recommendations. According to the published trial, they include the lower mortality rates of people who volunteer for trials versus non-volunteers, newer scanners that detect more abnormalities, and varying levels of radiology and cancer expertise at medical centers that do the screening.

The VA project also included patients who were older, more often male, and more often current smokers. According to the American Cancer Society, most abnormalities that turn up on CT scans are the result of old infections or scar tissue. A U.S. News & World Report article, “Is it Lung Cancer or Something Else?”, does a good job of explaining how CT scans often pick up harmless masses.

Clearly communicate risks and benefits

Rita Redberg, MD

An editorial accompanying the analysis noted growing evidence of the need to refine how patients are selected for screening as well as “clearly communicate risks and benefits of screening.” It added, “This is especially salient in light of recent evidence suggesting that high numbers of low-risk individuals are being screened in real-world practice.”

A study based on National Health Interview Survey data showed even some people at very low risk were being screened, concluding that “unintended spillover raises concerns about overuse.” The study also highlighted a slow uptake of lung-cancer screening by heavy smokers who meet screening criteria, an aspect covered by Kaiser Health News.

In an email, editorial co-author and JAMA Internal Medicine Editor Rita Redberg, MD, said “lots of factors” encourage inappropriate screening. “The belief that ‘early detection is always better’ is a lot of it,” she said, “and everyone likes to feel like they are doing something for their health, and that is often having a test. And of course, there are profit incentives.”

“Ounce of prevention” vs. “If it ain’t broke”

As wrote last year, it’s often difficult for average consumers to grasp the drawbacks of screening, and journalists often don’t do a good job of illustrating them.

Caverly pointed to online calculators to help patients determine whether they might benefit from screening.

For those on the margin, he said, some “might choose screening if they generally like to be aggressive about taking medications and getting tests when there is a chance to improve their health. These maximalists go by the saying, ‘An ounce of prevention is worth a pound of cure.’”

“On the other hand, a more minimalist person might choose not to screen as they generally worry more about unnecessarily becoming a patient when they feel healthy, prefer to avoid taking medicines and avoid getting tests unless the benefits are large, worry more about unintended consequences of interventions, and go by the saying, ‘If it ain’t broke don’t fix it.’”

You might also like

Comments (1)

Please note, comments are no longer published through this website. All previously made comments are still archived and available for viewing through select posts.

David Littleboy

January 29, 2018 at 11:38 pm

I think you overstate the “more minimalist” person’s position. It’s much more reasonable than that. I had a bad cold last summer with a persistent cough. A shadow on an x-ray led to a CT scan that two doctors thought was non-problematic but a third thought required a follow up. What said doctor thought she saw was sarcoidosis, which would require a lung biopsy to diagnose (and would be an unlikely disease for a 65-year old white male). As a patient, getting one’s lungs carved into because of a cold is not amusing, In general, the next level of diagnostic tests are going to have serious downsides. Sure, sarcoidosis has a 3% mortality rate, but if sarcoidosis is going to kill you, it’s going to make itself known in other ways first. (My father died of sepsis following unnecessary exploratory surgery, so this is a problem I’m painfully aware of.)

I realize that doctors aren’t bad guys: if a doc sees something that might be a problem and doesn’t go after it, they’re in trouble. So that means that it’s the patient that has to put the foot down. But that’s really hard once a doc has a tiny dot on an X-ray or CT scan, an odd PSA number, or whatever. So you really don’t want to get to the point where a doc thinks there might be something there. (Yes, it’s possible that there might be screening that helps. But over the last few years, the trend towards realizing that screening isn’t working has been quite clear.)