 Which is safer on average? Cars or motorcycles? Let's look at statistics for death and injury for 1998. 2284 motorcyclists died during that time and 49,000 were injured. During the same time, 42,000 car drivers were killed and 3 million were injured. So we can logically conclude from these figures that motorcycles are hundreds of times safer than cars. We should probably get everyone to start driving motorcycles everywhere to move our freight and for our public transportation. I'm picturing Harley-Davidson chopper buses and Kawasaki Ninja 18-wheelers. Let's take a look at another statistic. In 1998 again, the number of accidents caused by ambulances just in the New York City area was 423, resulting in three fatalities. On average, an ambulance in New York is more than twice as likely to be involved in a collision than a car. Should we classify ambulances as unsafe transportation and replace them with something safer? Perhaps a motorcycle, a horse and buggy. That's logical, right? Now, what's wrong with my thinking about motorcycles and ambulances? There's an asymmetry in the comparison, a missing component. In the case of motorcycles versus cars, the number of people riding around in cars versus the number of motorcyclists is an important factor not considered. When we adjust the comparison to the number of people and the time spent on the road in that vehicle, the numbers are very dramatically reversed. Motorcycle riding is a very high-risk activity when compared to car riding, even though your straight chances of getting hurt in a car are actually far higher. If we replaced every car in America with a motorcycle, the fatalities would most likely not go down. They would go up. Plus, there are a lot of things that a motorcycle can't do. It can't haul heavy freight. It can't handle large-scale public transportation. It can't take the injured to the hospital, not without fundamentally changing what we think of as a motorcycle. Keep this analogy in mind when we examine a very common argument used by alternative medicine proponents. The argument is some variation on. Conventional medicine is the third leading cause of death in the U.S., or Conventional medicine kills 700,000 people a year, or 250,000. It's always hard to track down where they're actually getting the numbers. I'm going to call this the deadly doctor gambit, and I'll show you why it's bad logic. A disease, disorder, or death caused by medical treatment is termed iatrogenic. The Greek root is iatros, which means healer. The category of iatrogenic causes in medical statistics is very broad, and includes medical errors, failure to treat, equipment malfunctions, infections transmitted between patients, drug reactions, or interactions, alert reactions to medicine, and actual negligence, including bed sores and hospital-derived infections. How big a problem is iatrogenesis? Very big. Figures range very widely, but it's a broad category. It probably accounts at least in part between a hundred and two hundred thousand deaths a year in the U.S. This puts the broad category of iatrogenesis behind heart disease and cancer as leading causes of death. You won't see this in most mortality charts. Why? Are the vital statistics people hiding this shameful fact? No, it's because when we break out all the different components, they don't individually add up to much. No, sacomial infections are their own category. Drug reactions another, medical mistakes another. Each of them is not in the top 20. I don't want to minimize them. This is a real problem and a major threat to patients, but I want to try to draw the real risk out as clearly as I can. Life is 100% fatal. How you die depends on lifestyle, genetics, environment, healthcare, behavior, and luck. Most Americans will die of cancer or heart disease. In fact, those two together account for 50% of deaths. Stroke, pneumonia, COPD, accidents, suicide, murder, Alzheimer's, infection, diabetes, and medical accidents make up most of the remainder. In the developing world where healthcare is often a luxury, not an expectation, the causes of death are very different. HIV-AIDS is a big one, as are respiratory infections, diarrhea, childhood disease, malaria, and stroke. Hunger and nutrition still account for the death of over six million children a year around the world. But you won't see medical mistakes as a major cause of death. That's not because their doctors are any better. It's because in the developing world we have the medical capability to keep patients alive for extremely extended times. It's not until a doctor makes a wrong decision. He goes left when he should have gone right, that you finally succumb to your terminal illness. In the developing world, they don't have the luxury of playing chess with a disease process for months or years. People are more likely to die of an acute disease like pneumonia, and less likely to die of something chronic like heart disease. The more care a person receives, in fact, the more likely we would expect their final cause of death to be iatrogenic. But all of this is completely irrelevant to refuting the deadly doctor gambit. If you go this route in a debate, you've allowed the alternative medicine proponent to successfully divert you from the real issue at hand, which is whether alternative medicine has any benefit. After all, even if doctors killed one out of every ten patients, it would have no bearing on whether alternative medicine works as promised. If the debate is over which is more effective, that's something that can be tested by clinical trials. In fact, many alternative therapies have been rigorously tested and simply can't compete with the science-based treatments. They fail to treat disease. They may also be very safe, as in the case of homeopathy. But if they're ineffective, what's the point of even using them? I want to give a specific example, one I hear a lot. Insates, or non-steroidal anti-inflammatory drugs like ibuprofen, have a very high rate of damage to the stomach lining. This is because the enzymes they inhibit, the cyclooxygenases, have dual roles. They cause pain and inflammation, but they also play a role in maintaining the pH of the gastric juices. Now, we could treat people in chronic pain with crystal healing, and they would have a mild placebo response. But in the long term, they would continue to have pain. Arguing for crystal healing by arguing against NSAIDs is a false dichotomy. There are many other ways we could respond to the chronic pain, other than NSAIDs and crystal healing. We can choose not to treat. We can treat with another placebo. We can treat with meditation, or diet, or exercise, or we can perform surgery. Each would have an associated risk and benefit. It could be that there would be a clear best treatment, one that minimized risk and maximized benefit. This is essentially what physicians do their best to determine. Looking at the risks alone is not the right way to choose a therapy. It so happens that the NSAIDs, when properly used, are very effective and safe. Crystal healing is also very safe, but not at all effective. I've left my analogy of the ambulance for the last. There are some jobs that alternative medicine just isn't capable of doing at all. No matter how poorly. There are no homeopathic cures for a rupture of appendix. No chiropractic treatment for AIDS. And no herbalism solution to advanced tissue gangrene. Just like our risky ambulance ride, there are going to be times when there is simply no rational substitute for conventional medicine, no matter how risky the existing therapy is. For example, a caesarean section is anything but natural childbirth, and yet it saves tens or even hundreds of thousands of babies and mothers from dying every year. It remains a high-risk procedure, but it's better than the ineffective alternative. So the next time you hear someone use the deadly doctor gambit, you'll recognize it for what it is, a logical fallacy, and a smoke screen for the inadequacy of alternatives to science-based medicine. Thanks for watching.