Is the opportunity for Medicare recipients who are heavy smokers to receive yearly low-dose CT scans a boon or a problem? This New York Times analysis not only provides a little history about the recent Medicare decision to fund the procedure, which seeks to find lung cancer in its earlier stages, but also shares plenty of detail and context as to the likely benefits and harms of getting screened. Cost was the only one of our 10 criteria that wasn’t satisfactorily addressed. Excellent reporting on a complicated issue.
Among the illnesses caused by smoking, lung cancer is a big one. It is the number one cause of cancer death in the U.S. and most lung cancers occur in smokers. So short of stopping smoking, what can be done to decrease the risk of dying from lung cancer? Before the National Lung Screening Trial (NLST), the answer was nothing. This trial demonstrated that compared with a chest x-ray, a yearly low dose computed tomography (CT) scan for three years reduced the likelihood of dying from lung cancer by about 20%, or by 5 in 1000 people (0.5% absolute reduction). Depending on one’s perspective, a 20% relative reduction or 0.5% absolute reduction, may or may not be important. And that is the focus of this article. How do you educate individuals who are eligible for this test so they can make a personal choice that’s right for them? Traditionally, doctors have simply told patients what to do. “I think that screening for colon cancer with a colonoscopy is a good idea for you.” However, this article points out that in addition to a small potential for benefit, there is also a real potential for harm or at least inconvenience. Almost 1 in 3 will have an abnormality seen on the CT scan that will require special follow-up. Of those abnormalities, only 6 in 100 will actually turn out to be something serious enough to require a biopsy or treatment. So patients and doctors are going to need to find ways to communicate these potential risks and benefits and help individuals assess preferences and other health or life issues that may influence the choice to screen. How well we do this will determine whether translating the results of a research study done under strict controls by experienced investigators translates into the same benefits in the broader population.
The story describes the case of one patient who paid $95 out of pocket for a lung cancer screen in 2013. But while the story does attempt to at least provide some cost perspective with the coverage, we’d note that under the Affordable Care Act, approved preventive tests do not carry copayments, so the cost of the initial scan is not really an issue for eligible patients. And in any case, the real costs to individuals and society are not really addressed by the cost of one scan. The story notes high up that the number of Americans who would be eligible for Medicare-funded scans would be in the millions. So the cost to Medicare and private insurers to fund this program will be, well, gargantuan at that scale. There is the cost of the initial test, annual testing if negative, and potentially more testing if an abnormality is found. This is not going to be a cheap program.
The story does an admirable job of explaining the modest benefit of these scans. The story’s effort to explain the statistics in several ways, including the use of absolute numbers, could serve as a model for other journalists. Example: “In the national trial, those screened with chest X-rays had about a 1.7 percent chance of dying from lung cancer during the study period; in the CT scan group, it was about 1.4 percent. For every thousand people screened with a low-dose CT, three fewer died of lung cancer.”
Another way to express this benefit would be the number need to treat (NNT) — the number of patients screened to prevent one lung cancer death. In this case, the NNT appears to be around 1 in 400. This is considerably better than other screening programs such as breast or prostate cancer.
This story devotes considerable space to explaining the potential problems invoked by false positives, again with facile use of numbers. The one narrative offered in the story elaborates on this element by focusing on an elderly retiree whose scan unearths a “hot spot” on her lung that, after considerable angst and pain caused by diagnostic procedures, was found to contain no cancer.
Another harm that could have been added is that the screening x-ray involves radiation, which may cause cancer itself. Though the scan is “low dose”, it isn’t “no dose,” and we know that screening not only will prevent death from cancer, but it will in fact be the cause of some of the cancers.
The story references a large screening trial involving more than 53,000 smokers and former smokers. It could have provided the reader with more information about how the study was carried out — for instance, emphasizing that to achieve the benefit seen in the trial, you need a yearly CT scan until either you age out or, if you’ve stopped smoking, you reach 15 years since quitting. But the story contains plenty of context about the study, especially with respect to the types of patients enrolled and the harms observed in the study. It also noted that treatment for those found to have cancer “involved major medical centers with specialized radiologists and surgeons — the ideal setting” — suggesting that the same results may be difficult to achieve at less specialized centers. We’ll rule this Satisfactory.
In some ways, this story is an example of a way to avoid disease-mongering. Because any Medicare-approved procedure will, by definition, be carried out on elderly individuals, the story’s reflections on “competing mortality”—the tendency for older folk to be saddled with multiple chronic diseases—emphasizes the importance of situating lung cancer among the elderly in a larger matrix of health issues. Lung cancer is clearly an important disease, but, as one physician/researcher noted in the story, “If I find a teensy lung cancer in a 77-year-old with heart disease, I may not have done him any favor.”
The sources in this story are clearly identified, but no effort is made to articulate the presence—or absence—of conflicts of interest. However, a quick search of these individuals does not reveal any major conflicts, so we’ll rule this Satisfactory.
The single study discussed in this story—the National Lung Screening Trial—compared low-dose CT scans with more traditional X-ray diagnoses. But in reality, the comparison is between low- dose CT and not doing one at all. No one is recommending a regular x-ray of the chest.
The issue of interest here is whether low-dose CT scans are available for older, heavy smokers to screen for lung cancer. While this is satisfactorily addressed (it’s the main topic of the story), it isn’t entirely clear from the story if screening has actually begun or when it will. Programs to do the shared decision-making for Medicare patients haven’t been established and Medicare hasn’t yet released codes for billing for the initial visit where the issue is reviewed. It’s also unclear if commercial insurers are covering and for whom.
The story makes it clear that low-dose CT scans are now in use. What is novel here is a recent decision by Medicare to cover the cost of such scans for current or former heavy smokers who meet specific criteria.
We see no evidence that this analytical narrative relied on a news release.