Evidence, not intuition: Why lung cancer screening is not endorsed
It seems intuitive – screening all smokers with a chest CT for earlier detection of lung cancer:
Now, at the time of diagnosis, 75 percent of patients with lung cancer already have symptoms (persistent cough, blood-tinged sputum, chest pain or recurring pneumonia or bronchitis) related to advanced local or metastatic disease that cannot be cured. And by the time a lung tumor is visible on a X-ray, it usually has spread beyond the site where it arose.
Only 16 percent of lung cancers are found at a potentially curable stage; the disease is fatal in more than 90 percent of cases over all. But when tumors are found through CT scans in people without symptoms, 80 percent of the cancers are still confined to their original site and potentially curable with surgery alone.
Survival chances are directly related to the stage the cancer has reached when it is detected and treated. For Stage 1 lung cancer, the category of both of Mrs. Guettel’s cancers, five-year survival rates are 50 to 90 percent; for Stage 4, one that has spread to other parts of the body, it is a mere 2 percent.
Alas, guidelines are based on evidence, not intuition. If intuition influenced practice guidelines and recommendations, we’d still believe the earth was flat. The NY Times nicely summarizes the USPSTF’s position:
The main concern is that no study has yet proved that detecting early lung cancers with CT scans improves long-term survival. While one might guess that finding cancers early can produce a permanent cure, this has not yet been demonstrated for lung cancer found through screening. It could be that people’s cancers are found years earlier than they might have been, but they end up dying at the same time they would have if they not been screened.
Another possibility is that screening could result in overdiagnosis – finding tiny indolent tumors that would never have threatened life before the patient died of something else.
Want to reduce your risk of lung cancer? Stop smoking.
