The Huge but Essential Cost of Lung Scans
The lung-cancer screening that the U.S. Preventive Services Task Force is now recommending could add at least $3 billion annually in new costs to the already burdened U.S. health-care system. That should worry the Barack Obama administration and anyone else concerned with the system’s stability.
On the other hand, the new recommendation, still in draft form, could save 20,000 lives a year. And policy makers could recoup some of the cost by limiting liability for doctors who follow the guidelines in good faith.
The task force suggests that heavy smokers be given low-dose CT scans once a year. The test costs about $300, so if all 10 million eligible Americans get them, it would add up to that $3 billion. In reality, the bill will probably be higher, because the range of people considered to be at high risk will widen with time. Medicare and private insurers will have no choice but to pay, because the Patient Protection and Affordable Care Act mandates that they fully cover all services the task force highly recommends.
Unfortunately, the costs don’t end there. While a CT scan is a great screening tool -- it picks up more cancers than mammographies or colonoscopies, for example -- it also flags many benign spots, each of which need further investigation. And because the technology scans the whole chest, it can also pick up signs of other health conditions such as emphysema and heart disease.
The task force based its recommendation mainly on a $300 million clinical trial of current and former smokers conducted by the National Cancer Institute from 2002 to 2007. The results of this research, announced in 2010, suggest that, once the recommended annual screenings begin, patients and providers will be in for heartache as well as success: Thirty-nine percent of smokers in the trial had positive scans, but 95 percent of these turned out to be false alarms.
Across all the large trials the task force reviewed, positive screens led to confirmed lung-cancer diagnoses 4 percent to 42 percent of the time. (The higher confirmation rates were associated with CTs showing larger nodules.) Subsequent tests needed to nail down a diagnosis include more CT scans, positron emission tomography (PET) scans, biopsies via bronchoscopy and, for lesions that are especially hard to access, video-assisted thoracic surgery.
The studies tell us with scientific certainty that some patients will suffer harm in the follow-up. They may be led down a rabbit hole only to come up empty-handed or, worse, suffer complications or death from an invasive diagnostic test. In the NCI trial, there were six procedure-related deaths, suggesting that if all Americans who are at risk are given lung scans, about 1,200 will die in further efforts at diagnosis. That said, most patients who end up getting the higher-risk invasive procedures will have racked up multiple indicators beyond the initial CT scan that suggest their lesion is cancerous.
False negatives will happen, too, the studies show. About 20 percent of the time, patients will develop lung cancer after getting what appears to be a clear CT scan.
Despite all these drawbacks, the U.S. Preventive Services Task Force is actually the fifth major organization to recommend low-dose CT screening. It does make sense to scan some 320 people in order to prevent one death from lung cancer. This need-to-screen number is much lower than what’s required for mammography (more than 900 tests) and colonoscopy (more than 500). Nobody knows, however, if the number will hold up in daily practice, in which patient populations don’t have the same characteristics as the trial subjects: smokers age 55 to 80 with at least a 30-pack-year history. (Pack-years are calculated by multiplying the number of packs a day by years smoking, so someone who goes through two packs a day reaches 30 pack-years after 15 actual years of smoking.)
A doctor who sees a patient with 25 pack-years who suffers bouts of bronchitis or a smoker who is 82 but in great health (suggesting screening is more worthwhile) might be right to screen in either case, despite the task force’s recommendations, which are based on a population-level risk assessment that doesn’t factor in such variables. Consider short-term smokers who also were exposed to secondhand smoke in childhood or have worked for years in smoke-filled restaurants: The studies don’t address such scenarios at all, but it’s easy to see how these patients also might end up being screened as the tests become more commonplace.
A cascade of new diagnoses and treatments might threaten the economics of Obamacare. The law allows insurers to charge smokers a maximum of three times more for their premiums than nonsmokers pay, but that won’t be enough to cover the astronomical costs of screening. We will all have to pay the price in higher premiums.
One way to minimize this expense is to limit the risk that providers will bear by carrying out the recommendations. We already expect errors to occur; they happened, after all, even in the well-run clinical trials. Doctors who follow the task force’s guidelines in good faith should be protected against medical-liability claims and lawsuits for anything less than gross deviation from standards of care or willful harm. Lowering malpractice costs would bring down the price of CT scans.
We must also keep shrinking the ranks of the remaining 60 million Americans who smoke through such measures as tough restrictions on menthol cigarettes, which are more addictive. We should also consider an annually adjusted taxation plan that would increase the price of tobacco products on pace with the rising costs we all pay to treat tobacco-related diseases.
Such strong prescriptions to limit smoking and facilitate disease treatment would help us save 20,000 lives a year without breaking the bank in the process.
(Ford Vox, a doctor at the Shepherd Center for brain and spinal-cord injury in Atlanta, writes about health-care policy. Follow him on Twitter at @fordvox.)
To contact the writer of this article: Ford Vox at email@example.com.
To contact the editor responsible for this article: Mary Duenwald at firstname.lastname@example.org.