Imagine you are sitting in a 1000 seat theater filled with former or current heavy smokers, and you are told that Medicare will now pay for a test that, if done every year for 5 years on everyone in the theater, will save 3 people in the theater from dying of lung cancer. Would it be worth it? What if you were then told that two-thirds of people informed that they may have lung cancer with the test actually don’t have lung cancer, and it will require further testing, some of it harmful and life threatening, to prove that they don’t actually have lung cancer? In fact, while 3 lives are saved from lung cancer in the theater, 6 people without lung cancer will suffer significant harm from the test, and a large number of people without lung cancer will be told that they may actually have cancer. On top of this, about a fifth of cancers detected and removed after testing would not have caused harm if left alone, but the surgery to remove them will lead to some unnecessary deaths. Read our op-ed in the Baltimore Sun to see what you think about the new $9 billion screening test approved by Medicare.
Recently, the Centers for Medicare and Medicaid Services (CMS) declared that Medicare would now be paying for lung cancer prevention with spiral CT scans. The scans would be offered annually to people between the ages of 55 and 77 with at least a 30-pack year smoking history (calculated by multiplying packs per day by the number of years smoking) and who either still smoke or who quit less than 15 years ago. The controversial decision came four years after the National Lung Screening Trial (NLST) of 53,000 current or former heavy smokers demonstrated what was characterized as a 20 percent reduction in lung cancer death among those who had annual spiral CT scans of their lungs.
Before meaningful shared decision making between doctor and patient can ensue, it is important for both to understand the actual risks and benefits of spiral CT screening in absolute terms. The NLST data, for example, is presented in relative risk terms: 20 percent reduction in lung cancer death. But, it is generally acknowledged that relative values are misleading and that they distort and misrepresent relevant health benefits and risks. Therefore, it is imperative that data are presented in absolute terms for doctors and patients to have constructive conversations about risks and benefits.
What are the absolute numbers in lung cancer screening? Over five years, the people screened with spiral CT scans had approximately three fewer lung cancer deaths per 1,000 screened than those who were not screened. Put another way, 997 out of 1,000 people screened did not benefit from the test compared with those who were not tested. But given that the benefit was the avoidance of lung cancer death — a significant benefit — we must ask whether there were any downsides to screening. That question must be part of any conversation about risk/benefit as well, and, as is the case with the benefits, those numbers must be presented in absolute terms.
One of the perils of such screening has been a very high false positive rate, meaning that in many people with abnormal CT scans there is no cancer present, and only with further testing can that fact be determined. Thus, spiral CT scans will lead to a significant amount of unnecessary testing and, subsequently, to potential harm. Because of this, CMS has stipulated that it will pay for a single doctor's visit for eligible patients to discuss the risks and benefits of lung cancer screening.
More specifically, using NLST data, 675 out of 1,000 people with abnormal CT scans had false positives. Therefore, 675 out of 1,000 people who were told that their scan was problematic required additional, unnecessary tests to determine that they did not have cancer. For some patients, that level of stress and uncertainty may dissuade them from getting the test. Further, 230 out of 1,000 people with abnormal CT scans required high radiation PET scans and even lung biopsies to determine that they were in fact cancer free. Overall, it was determined that 6 out of 1,000 screened people without cancer suffered major complications from some form of intervention due to false positive testing.
It is estimated that 180 out of 1,000 cancers detected by CT scans would, if left alone, cause no harm, but because of their detection they will be removed. Lung cancer surgery can result in high mortality, with an estimated 30 day death rate of 50 out of 1,000 people, and thus people without lethal cancer may be harmed and even killed unnecessarily by over-detecting nonlethal tumors through sensitive CT scans. It is estimated that 7 out of 1,000 screened people have unnecessary surgery and are thus subjected to potentially fatal harm without benefit due to screening.
This current recommendation for screening will be conducted on roughly 4 million Americans at a cost of $9 billion, according to current estimates. It is possible that the rigorous standards mandated by NLST rules will not be met in the community and can lead to even more false positive results and more unnecessary harm. But for the doctor and patient sitting in an exam room discussing the risks and benefits of the test, the use of absolute risk/benefit numbers will allow each patient to make a reasonable personal assessment of whether spiral lung cancer screening is appropriate for them.