 While running for president of the United States, former New York mayor Rudy Giuliani ran a campaign ad contrasting his chance of surviving prostate cancer in the US 82%. With the same chance of surviving prostate cancer in England, only 44% under socialized medicine, where they don't do routine PSA testing for prostate cancer. To Giuliani, this meant that he was lucky to be living in New York rather than old York, because his chances of surviving prostate cancer seemed to be twice as high here in the US. Yet despite this impressive difference in this five-year survival rate, the mortality rate, the rate at which men were dying of prostate cancer, was about the same in the US and the UK. Wait, what? PSA testing increased survival from 40 to 40, 82%. How is that not evidence that screening saved lives? For two reasons—lead time bias and overdiagnosis bias. I've talked about overdiagnosis, where a cancer is picked up that would otherwise have never caused a problem. Without screening, let's say out of a thousand people with progressive cancer, only 400 are alive five years later. So without screening, five-year survival, only 40%. But let's say with screening, an additional 2,000 cancers are overdiagnosed, meaning you picked up cancers that would have never caused a problem or even would have disappeared on their own. Since the cancer was harmless, five years later, of course they're all still alive, assuming their unnecessary cancer treatment didn't kill them, and all of a sudden you just doubled the five-year survival rate, even though in either case the same number of people died from cancer. And that's one way how changes in survival rates with screening may not correlate with changes in actual cancer death rates. The other is lead time bias. This is how it works. Imagine a group of patients in whom cancer was diagnosed because of symptoms at age 67 years, all of whom die at age 70. Each patient survives only three years, so the five-year survival for this group is 0%. Now imagine that same group undergoes screening. Screening tests, by definition, lead to earlier diagnosis. Suppose that with screening, cancer is diagnosed in all patients at age 60 years, and imagine in this case they nevertheless all still die at age 70 years. In this scenario, each patient survives 10 years, so the five-year survival rate for this group is 100%. Survival just went from zero to 100%, call the newspapers. With this new screening test, now cancer patients are living three times longer, 10 years instead of three, it's a miracle. Whereas all that really happened in this case was that the person was treated as a cancer patient for an additional seven years, which, if anything, probably just diminished their quality of life. So that's the second way. How changes in survival rates with screening may not correlate with changes in actual cancer death rates, and in fact the correlation is zero. There is no correlation at all between increases in survival rates and decreases in mortality rates. That's why if there were an Oscar for misleading statistics, using survival statistics to judge the benefit of screening, we would win a Lifetime Achievement Award hands down. There is no way to disentangle the lead time bias and the overdiagnosis bias from screening and survival data. That's why these statistics are meaningless when it comes to screening. Yet that's what you see in the ads and the leaflets for most of the cancer charities. That's what you hear coming from the government. And prestigious cancer centers like MD Anderson have tried to hoodwink the public like that. If you've never heard of lead time bias, don't worry, you're not alone. Your doctor may not have either. 54 of 65 physicians surveyed said they did not know what lead time bias was, and then when they asked the remaining 11, okay, what is it? Only two were actually correct. So at this point in the video, already you may know more about this than 97% of doctors. To be fair, though, maybe they don't recognize the term, but understand the concept? Nope. The majority of primary care physicians did not know which screening statistics provide reliable evidence on whether screening works. They were three times more likely to say they would definitely recommend a cancer screening test based on irrelevant evidence compared to a test that actually decreased cancer mortality by 20%. If physicians don't even understand key cancer statistics, how are they going to effectively counsel their patients? Statistically illiterate physicians are doomed to rely on their statistically illiterate conclusions, or on local customs, or on industry representatives and their information.