Lead time bias is a common phenomenon to watch out for in screening studies, though it can be relevant in cancer interventional studies as well. People usually don’t go to the doctor for a disease until they experience symptoms. The goal of screening is to identify a disease before symptoms occur, ideally so that earlier treatment will lead to a better outcome. But identifying a disease before symptoms occur can lead to lead time bias, which artificially inflates the survival time of someone with a disease because their diagnosis was identified earlier than it would have been had they waited until they experienced symptoms. As described by Abi K. In this YouTube video, “The earlier you diagnose a disease, the longer a patient will appear to survive—even though we just started counting earlier.”
Deeper dive
For chronic diseases like cancer, a person is usually asymptomatic for a while before they become aware of the symptoms that cause them to go to the doctor. But if they are screened for a specific cancer, such as with a mammogram (breast cancer) or lung scan (lung cancer), a physician may discover they have cancer before they begin experiencing symptoms. Lead time is how much time passes between disease detection due to screening and the onset of symptoms (that would have led to diagnosis).
Lead time bias occurs when a person’s survival time is counted from the moment of the diagnosis following screening and then compared to the survival of someone whose diagnosis occurred after the onset of symptoms. Those two survival times cannot be compared because a) the person with symptoms almost certainly had a more advanced form of the disease at the time of diagnosis, and b) the person who was screened got “extra” survival time that occurred between the time they were screened and the time they would have experienced the symptoms that would send them to the doctor. That means the person who was diagnosed earlier and begins treatment earlier will appear to have a longer survival time than the one diagnosed later — even if their treatment does not offer any benefit at all. Hence, lead time bias can make it appear—falsely—that an ineffective treatment does in fact work simply because a person was treated earlier in the course of the disease than someone else.