woroud

'Overdiagnosed' is, without a doubt, an important book. It convincingly shows that if we look harder for a disease, we find more "abnormalities" but we're not preventing any deaths. Instead, the more people are subjected to unneccesary medical procedures, which can be just scary, expersive and annoying but also debilitating or even lethal. The authors explains how this happens with great clarity. Let's say, as an example, that each year a hundred women find a lump in their breast. They go and see a doctor, who does an autopsy and confirmes it's breastcancer. All these women are treated. 90 get better, 10 die. Now, let's travel one year back in time. All 100.000 women in a town our having mammography to screen for early signs of breastcancer. Even though our hundred women wouldn't have lumps in their breasts yet, the mammography will show an abnormality. But they will not be the only ones with an abnormality. Research shows that up to 40 percent of all women have some sort of 'could be cancer could be nothing' in their breasts. This means that screening will not only raise a red flag for our 100 women who would have developed clinical symptoms of breastcancer (eg a lump) but also for up to 39.900 other women in their town. They will be scared, have biopsies, maybe have their breasts removed, just like our 100, but they were never destined to develop clinical symptoms. That would all be fine, if early detection of asymptomatic breastcancer would save a lot of lives. But as Gilbert Welch shows, it doesn't. Mortality rates have not gone down since wide spread breastcancer screening was introduced. So those 39.900 women were not lucky, they were overdiagnosed, because their abnormality would never have progressed into lethal cancer. Breastcancer, ofcourse, is just an example. Almost every disease or condition (high blood pressure, diabetes, high cholesterol, prostate cancer as notable examples) that we screen for in a large group of healthy individuals is prone to overdiagnosis. This is because, as Welch argues, the most serieus cases - which are also the ones to benefit most from treatment, because they are sicker and thus have more to loose and more to gain - don't need the screening: they will present with symptoms and see their doctor anyway. So screening mostly diagnoses relatively mild cases, people who otherwise feel ok. This 'mildly abnormal' group is, as in the breastcancer example above, much much larger than the 'dangerously ill' group. So what screening does is tranforming a large group of otherwise healthy people into patients by giving them an early diagnosis of a disease that may never progress to giving clinical symptoms. In other words: it makes patients out of healthy people who, even in the future, may have never gotten ill. And, to add insult to (nonexistent) injury, these now 'not symptomatic but worried' patients only really benefit from medical treatment until they do become symptomatic. But since they are also worried, and doctors are uncomfortabke doing nothing, they probably will get treatment, and suffer all the risks and side effects from it. Welch's remedy is simple: if research shows early detection has no benefits because it does not prevent suffering or death (as is the case for all examples I mentioned) don't go out and screen lots and lots of healthy people. Instead tell patients to visit their doctors as soon as a symptom occurs, find out what's wrong and treat accordingly. It will save just as many lives, but it will spare healthy but slightly 'abnormal' people ghe expense, anxiety and pain of being overdiagnosed.

2023-01-24 20:23