Show 1141: Which Health Risks Should You Worry About?
The people's pharmacy podcast is sponsored by the brain gauge developed by neuroscientists at the university of North Carolina to study brain function across a wide range of applications including aging and traumatic brain injury, the brain gauge, translates state of the art neuroscience into easy to use methods that let you take control of your brain health now. Available for home research and clinical, applications, find out more at gauge your brain dot com. Do you ever? Get fed up with scary health headlines about the risks of coffee wine or bacon. How can you make sense of them? This is the people's pharmacy with Terry. And Joe Graydon. Dr Aaron Carroll is a pediatrician and expert on health research and policy. He'll offer us advice on making sense of health risks. We know the sun raises your risk of skin cancer. No, one says never go out in the sun. They say take proper precautions and think about how much doing because you don't want to raise your risk too much. Putting scary statistics into perspective requires more than a headline distinguishing between relative risk and absolute risk is critical to making informed decisions coming up on the people's pharmacy. How did tell which risks you should worry about first this news? In the people's pharmacy hills, headlines, popular blood pressure. Medications called ace inhibitors have been associated with an increased risk of lung cancer. Lisin appeal is the most prescribed drug in the United States over one hundred thirty million prescriptions for this, Angie. Tencent. Converting enzyme inhibitor are dispensed annually that doesn't take into account. Other ace inhibitors such as bananas Apprel capped Apprel, and now Apprel ram Apprel and Quinn Apprel such drugs are very effective at lowering blood pressure. But a new study raises questions about the safety of long-term use the investigators collected data on nearly one million hypertensive patients in the UK between nineteen eighty eight and two thousand fifteen taking an ace inhibitor was associated with a fourteen percent increased risk of lung cancer. This only became detectable after five years of us. The longer people took such drugs, the greater the risk after ten years the risk increased to thirty one percent, the authors point out that although the absolute risk of developing. Lung cancer is very small so many people are taking these medications that the number of patients affected could be quite large. It's difficult to diagnose Alzheimer's disease in this early course. In fact, a definitive diagnosis has only been available upon autopsy. Now, scientists have found markers of the disease that can be seen in the retina of the eye even before people notice serious memory loss. The noninvasive test called optical, coherence, tomography and geography could be done by an ophthalmologist and is able to distinguish between people with mild cognitive impairment, and those who progress to Alzheimer disease in Alzheimer's disease. The retina has fewer blood vessels and the inner layer is thinner. These observations were made independently by two separate teams of researchers and presented at the annual meeting of the American Academy of the Malla. Gee, it's been almost two decades since the FDA approved a pill to treat influenza that was Ozil town of ear also known as tamiflu last week the agency approved a new flu pill Zo flu can be taken by teens and adults who have had symptoms for less than two days, and it shortens the duration of flu symptoms by more than a day patients take just one pill instead of a series of pills. So it is much more convenient than other flu treatments so flu so works on a completely different violence. I'm than tamiflu in relentless so flu viruses, have not yet developed resistance side effects. Abso- flus include diarrhea bronchitis. Nausea and sign you side. His public health officials stress that antiviral drugs do not replace vaccination as. As the first line of defense against influenza people with mildly. Elevated blood pressure are usually given a prescription for an anti hypertensive medicine. However, previous studies haven't really demonstrated whether such drugs prevent cardiovascular complications in low risk patients. British researchers reviewed long-term medical records of adults with mild, hypertension. They define that. As blood pressure between one forty over ninety and one fifty nine over ninety nine without medication. More than nineteen thousand people. Taking blood pressure pills were compared to nineteen thousand other patients with similar blood pressure. But not taking medicine during the nearly six year follow up period, they found no evidence that treatment prevents cardiovascular disease or death. The medications did have side effects. However, most notably low blood pressure fainting electrolyte imbalance and acute kidney injury. The off. Authors conclude this pre specified analysis found no evidence to support guideline recommendations that encourage initiation of treatment in patients with low risk mild, hypertension, gum disease has been associated with type two diabetes osteoporosis, rheumatoid arthritis and a number of cardiovascular complications. Now, researchers report that people with poor oral health appear to have a harder time controlling their blood pressure. They review data from the US national health and examination survey and found that about half of the participants had gum disease. The worst the Pero Donald assise the harder it was to manage bloodpressure. Perio? Donald therapy. Reduced the likelihood of anti hypertensive treatment failure. And that's the hill. News from the people's fund. See this week. Welcome to the people's pharmacy. I'm Terry Graydon. I'm Joe Graydon. Do you ever get confused and frustrated by conflicting health headlines one month coffee's bad for you? The next month is going to prevent diabetes heart failure, and maybe even Parkinson's disease for years. People were advised to keep their egg consumption to a minimum. Now, we're told that eggs won't clog your arteries after all how do you deal with all those flip flops is there a way to make sense of the contradictory headlines without getting whiplash? Sometimes you may feel you need a resent us down to crack the code. Well, we have just the guide you need. Dr Aaron Carroll is a professor of pediatrics and associate dean for research mentoring at Indiana University school of medicine, he's also director of the center for pediatric and adolescent, comparative effectiveness research, he focuses on the study of information. Technology to improve pediatric care healthcare policy and healthcare reform in addition to his scholarly activities. He writes about health research and policy for the New York Times, among other outlets. His most recent book is the bad food bible how and why to eat sinful iw. Welcome back to the people's pharmacy. Dr Aaron Carol thank you so much about me back. Dr Carol, you know, you are the one person we go to all the time when we have a question about some new research. Who was a study that was published several weeks, maybe several months ago about the the horrors of alcohol even one drink is too many, right? Wondering shorten your life dramatically, and then, you know, millions of people go, oh my goodness. And I even heard on television physicians, you know, MD types saying, oh, yeah, alcohol, it's really bad for you. And so you've kind of put it into perspective. Help us understand statistics probability. And what those studies mean for us? So this study got a lot of media attention, and it was portrayed in very much the way that you're describing it with headlines like there's no safe amount of alcohol, and, you know, lots of news stories arguing that even one drink is too much. So it's important to understand that this was a population based study, which is trying to get across a population base message, and that's perfectly reasonable and this. Has certainly pretty much the largest study of its type ever done was a meta analysis or a study of studies that granted together all of the observational studies that exist. It had looked at least twenty three different alcohol related problems that could come from that they gathered together hundreds of sources to estimate how much people might be drinking worldwide and put it all together. And basically they found that very high levels of drinking. You're very likely I have many of these problems, and there's a pretty consistent dose response, or at least it gets worse as you drink more and the lowest point was zero. And that's how they came out with those headlines, but there's a lot of things we have to consider what studies like this? The first is that it's Oster facial data. It can easily be confounded. There could be unmeasured factors that are contributing to the harm people that drink also smoke people to drink off tend to be poor. There could also be genetic differences or obesity differences, all of these would matter, and none of them could come into the play when they were actually looking at the study because the only thing they data for was basically age sex and location, and that's not the researchers fault. That's probably all they could do. And if you're going to model population level wide effects, that's fine. But they and a lot of the media then carried this individual level risks. And if you've got to make a claim that even one drink a day is really dangerous. Well, it's it's important to understand first of all what's the magnitude of that risk. And this is going to get a little into the numbers, but it's important to understand. So for every hundred thousand people who drink one drink a day nine hundred eighteen will probably have one of the Twenty-three related alcohol problems in any year, but of one hundred thousand people who drink nothing nine hundred and fourteen would experience one of those problems. This means that of one hundred thousand people ninety nine thousand eighty two will not be affected by drinking drink nine hundred fourteen will have an issue. No matter what they do only four in one hundred thousand people who drink drink a day might have one of those Twenty-three related. Problems that is an incredibly incredibly small risk. And no one should be assured that we've proven causal data from this study, even at true drinks a day. The number of people who experience a problem goes up to nine hundred seventy seven hundred thousand but again, nine hundred eighteen of them would have a problem with no matter what even at five drinks a day. We're only into the low thousands of people who might have a problem out of one hundred thousand and no one would argue that five drinks a day is a good idea that is too much. And so I'll Kahal ISM is terrible. It really is and drinking too much is really really bad for you. But the actual potential harm from very low level drinking or or even light to moderate drinking is very very small does not hit everyone equally and to take these kinds of studies. And then they huge sumptious about how it's going to affect individuals is really really going too far. Now, Dr Carol you said we can. Make causal inferences from this observational data. That's that is kind of a problem, isn't it? It is. And with some things like smoking. You will hear I've mostly tobacco. And she, but you'll hear some experts at what we've never proven that smoking causes cancer because we have no randomized controlled trials. But the odds ratios and the damage the numbers of people who have these are so great. And so large that at some point we say, okay, we're not going to do a randomized controlled trial, but we can pretty much prove it these are incredibly small numbers, and we could do a randomized controlled trial of alcohol. In fact, one was really in the works until some articles in the New York Times which actually reported on ethical concerns in the ways it was being pitched to funded by industry got it shut down. But that doesn't mean that we couldn't do a trial of light to moderate drinking, and that it wouldn't be in the public interest and wouldn't be worth funding. It absolutely would be. I'd also point out. We do have some randomized control. Field trials of alcohol where people were randomized to water or red wine or white wine, for instance, and it has been shown to have positive effects with respect to perhaps the prevention of diabetes markers or lowering of them and also with respect to some markers that would show cardiovascular risks. So if anything there's a little bit of evidence that light to moderate drinking might be beneficial in that respect, we won't know the true causal effect until we do a real randomized controlled trial. We probably should if we really want to close the door that, but until we do that to keep making huge claims from vast huge observational studies that you'll very very tiny rest. We're cheating statistical significance without necessarily achieving clinical significance and doing another meta analysis is not gonna do us. Any more good? We have about as much knowledge as we can get out of observational studies. If we want new knowledge, we're going to need a big randomized controlled trial. Thoughts are Carol. We probably should describe the difference between all epidemiological studies case control studies observational studies and the gold standard are CT's randomized control trials could you give a quick overview of the difference between these German Y R C teaser so much more important. So this study is a collection of studies that we would call mostly cohort studies, which is basically they get together a huge bunch of people. And then they check and see whether they have disease, and they also ask them if they have been drinking, and if they're drinking how much if they gathered them together and they follow them forward. That's a that's a prospective cohort trial. If they asked him about things that have happened in the past that's a retrospective cohort Trump and what they do is. They can sort of identifying what we call. Either odds ratios relative risks where they can say people who have been drinking are at higher risk or more likely or less likely to have. Diseases or problems than people who don't drink the problem with observational. Studies is that they can be what we call confounded where there can be a link or a relationship between alcohol consumption and bad outcomes. But there could be something else in between that is the cause of that some of the things I've already mentioned already, for instance, people who drink tend to be more likely to smoke. We know smoking causes all kinds of health problems. And it could be that people who drink it's not the drinking that's causing the promise the smoking. It could be the people who are drinking are poor. And that often tends to be the case. And when we do these kinds of studies therefore, it looks like it turns out that it's the poverty, which is much more risky. In fact in a previous study like this. They found that alcohol was associated with worse outcomes. But when they broke it down, they found that actually beer consumption was was associated with worse outcomes. But wine and spirits were associated with slightly better outcomes. Now. No one is arguing that wine and spirits are good for you. It's just that people who drink wine and spirits tend to be wealthier. And again, poverty is associated with a huge number of health problems. So without the ability to control in some statistical for fashion for all these confounding measures, you can wind up with a result that there's an association, but that's not the cause it's not it might not be the drinking that is causing the bad outcomes. It might just be the people who drink also tend to have other issues, and those things are what are causing the bad outcomes. The only way to be really sure this. In fact, one of the few ways to sort of get a 'cause -ality is to do a randomized controlled trial where we take people, and we randomly assign them to drink or not because of that we can be assured or more assured that there's not some factor that's associated with their choosing to drink or not where we're randomize ING where making it by chance. And if we just randomized people to drink. Or not. And then we see that there's a relationship between drinking and some outcome. We can be more assured that the drinking is the cause of that outcome because we've not allowed the other factors to confound the results or to prejudge or to change whether or not people are going to drink or not so to really get it. 'cause -ality to really figure out does lighter moderate drinking caused these kinds of health problems. We'd need to do a randomized controlled trial. You've been listening to Dr Aaron Carroll. He's professor of pediatrics and associate dean for research mentoring at Indiana University school of medicine, he's also director of the center for pediatric and atlas and comparative effectiveness research his books include the bad food bible how and why to eat sin Fily, he spoke with us from his office, which is next to a busy highway after the break. We'll discuss why. Nutritional studies can be so confusing. Some studies about how to get children to eat more healthily sounded good, but aren't based on sound science part of the problem is that studies that find no difference. The no hypothesis are much less likely to get published. So how would you know about those results, and how does that affect our understanding of drugs like antidepressants safe and effective? Are they? You're listening to the people's pharmacy with Joe and Terry Graydon. The people's pharmacy podcast is sponsored in part by Kaya -biotics, K A Y A -biotics offers the first probiotics which are both certified organic and Hypo allergenic I'll probiotics are produced in Germany under laboratory conditions with high quality ingredients and under strict regulatory oversight. The three available formulas are created for very specific purposes such as strengthening the immune system, fighting east infections and helping with weight loss to learn more about Kaya -biotics, probiotics and the important topic of gut health you can visit their website Kaya, -biotics dot com. That's K A Y A -biotics dot com. Use the discount code people for ten dollars off your first purchase. Welcome back to the people's pharmacy. Ontari graydon. I'm Joe Graydon. If you would like a purchase a CD of this show, you can call eight hundred seven three two two three three four. Today's show is one thousand one hundred forty one that phone number again, eight hundred seven three to twenty three thirty four or you can find it online at people's pharmacy dot com. You can also download the podcast from I tunes today. We're taking a look at health risks, which ones do you really need to worry about. And how would you know, most of us are easily confused when it comes to the statistics we encounter in articles about medical research, drug companies are very good at using statistics to their own advantage. How can you defend yourself to help us better understand risks and benefits? We're talking with Dr Aaron Carroll. He's professor of pediatric. Trix and associate dean for research mentoring at Indiana University school of medicine, he's also director of the center for pediatric and adolescent, comparative effectiveness research, he's written three books debunking, medical myths. The most recent is the bad food bible how and why to eat sin fully. Dr Carol we've been talking about the problems of observational studies, and this may help to explain I there's so much confusion about nutrition. Studies people get all kinds of upset if we tell them one week that coffee is bad for you. And then three months later, we tell them no coffee's good for you. Or we say don't eat butter, eat margarine. And then we turn around and say, oops, we were wrong. Of course, we would never say those things. But that's what the media tends to do like, oh, don't eat butter and then ten twenty thirty years later. It's like, ooh, margarine wasn't so good after all, and then there's the whole saturated fat cholesterol story. And so when it comes to nutrition there is so much confusion in large measure because of these observational studies. I would say I go further and say, it's not even just because of the observational studies. It's also because of the way we interpret research and the way that we talk about it and covered in the media. So we can start with the fact that I think you're correct that a lot of these studies are finding associations and those associations are often statistically significant and on necessarily clinically significant. And from that, we extrapolate there must be a causal pathway. And that one thing these things must 'cause one of the others when it's not true. And therefore, if we do another observational study where the associations are just different because it's a different population or something else, we can find a very different result. There's a classic study. That was published in the eye. Kim was I think it was two thousand twelve by John wanted us, which they took a cookbook, and they randomly picked fifty ingredients in the cookbook, and then they went out and looked at our their studies that show whether these fifty ingredients cause or prevent cancer, and they found research on almost all of them. I the more than forty. But what was interesting about? It was they can find studies that showed pretty much all of those ingredients both caused cancer and prevented cancer. In other words, you could find a study that said it made cancer less likely and for the same ingredient, you could find a study where cancer would be more likely. This is part of the problem with how we do nutrition research. We we isolate these individual nutrients, we try to study them all by themselves without recognizing that. Of course, they are completely confounded because they're being eaten along with tons of other foods, and in ways that that have all kinds of issues, we know that these studies are often very small they are often for very short. Periods of time. They often involve very few subjects and because of all of that together. The results are not nearly as robust or powerful as we think they should be or we would like them to be even when they're a randomized controlled trials. They're often for weeks involving tens of people and. And again with outcomes that are not long term or in ways that we care about those. They don't follow people enough to actually look at death or true health, or will you cancer? They're following biomarkers or laboratory values that fluctuate all the time and never turn out to correlate with what we actually care about. But part of the problem is also how we talk about this. Every new study is greeted as if it's an evacuation. So if there is a study with thirty people were they find out that dark chocolate is associated with some outcome that they care about they breathlessly pronounce it. Oh, we've proven that that this is true. There's a relationship dark chocolate is good for you without saying, but we have tons and tons and tons of research already in this area. Does this change our minds or is it just a tiny study in a huge sea of data? We don't do that. We don't take the large view. We don't look at all of these things together. And because of that we get misinformed and think that each new study is. Truth and we've waffle from one direction to the other direction. Instead of saying look, we've a ton of research some of it goes one way some of it goes the other way, therefore the likely answer is this doesn't make a difference at all. And if we took that sort of broader attitude, I think we'd have a better sense what's going on. But there's even another problem, then that involves publication bias studies, which are sexy studies which are going to scare people or make them think something that's really exciting are much more likely to get published and much more likely to get covered in the news. Then studies which are boring and so scientists whether or not they know it are seeking out the results that they think might get them to more likely to get some variety. And of course, they're being cherry picked out of the either the studies that are more likely to be exciting or to say, something new are more likely to get published more likely to covered more likely to get discussed more likely to get cited which gives us a false impression of what truth really is. 'cause we're seeing this. The sexy side of it or the exciting side of it, which often is not the true side of it. And so all of that together happens far too often in attrition research, which winds us with results that don't really hold up over time or give us a good sense of what we should or should not be eating. I remember learning in high school that it was just as important to pay attention to results if you do a study it's just as important to pay attention to resolve that don't arrive at the null hype. Or that I guess do arrive at the null hypothesis that say, okay, there is no difference. This. This doesn't make any difference. It may be scientifically important, but people aren't that interested in it. Right. No from across the board. In fact, I just wrote a column on this recently talking about negative results. We don't celebrate them. Like, we should we don't get excited about them. Like, we should we don't try to public. In fact, are much less likely to get published if we do publish them. We are much more likely to actually try to change, you know, even the outcomes that we're talking about our spin them as positive which can often be a problem, if they do get published where much less likely to discuss them in the media as I said, and so we don't care almost about negative results in the same way, we do positive, but that's not how science works. The scientific method is set up around trying to set up a hypothesis and then trying to see if it is true or not and finding out that it is not is just as important as finding out that it is. Unfortunately, that's not how we do stuff. Even all the way if you go far back to grants. You know, the H is looking for innovation. They want to be convinced that that something is going to be new and exciting, and it's going to lead to positive results that that they're going to care about institutions are going to get more excited academic institutions are going to be more excited more likely to promote you and to give you accolades if you're published in high profile journals and that comes again from exciting new positive results. That's what people wanna see in the media. That's what they want to see on the news and all of this together creates a scientific environment where we are pushing for exciting new counterintuitive bizarre flashy results and not sort of necessarily getting his hyped up about or excited about things which are negative. No, boring. But those kinds of buying sees the things that lead us to research. That's not reproducible and to research that doesn't necessarily reflect truth. Well, I do wanna talk to you in a moment about pharmaceutical stuff. Studies because that's as true for drug says it is for nutrition. But before we dive into the world of medications, I do want to ask you about a column you wrote for the New York Times, titled the cookie crumbles, a retracted study points to a larger truth, and I know it would be very tempting to say, I told you so within fact you did a couple of years ago pretty much saying the same thing, but you were proven correct in the more recent research that was retracted can you give us a quick overview of what happened. Sure that was a study actually that got retracted it was it was a study that had we had thought had taken place in basically showed that if you put Elmo stickers or stickers of characters that kids would be interested in on apples that at school. They'd be more likely to choose the apple over the cookie that they would small dry. Offers like that we could influence behavior and get kids to eat more healthily. These kinds of studies pump up in the media all the time. They're very attractive to us. We want to believe that they are true because we'd love to believe that small painless things can get us all to eat much more healthily, and to to perhaps even lose weight and not have as much of a problem with obesity, as we do all throughout the United States that study had problems because some people went back looked at the numbers, and it turned out that it didn't take place in a school as previously thought. But in a preschool or in a daycare center in a head start. If I believe those correct. So it's not surprising that we can get very small children to choose apples over cookies if you give them it almost sticker. The problem is can we get school aged kids doing the answer? Now, it turns out that a lot of studies had been done in the lab that had produced this result. And a lot of them recently in various news had were all retracted. It was just a huge set retract. From the network of journals. But this is the kind of research you hear about where if we give people smaller plates, they less if they order food when they're not hungry ahead of time. They will order less if we change how the buffet works. Are we changed the time of day when they shop? They tend to buy less calories all of these things sound great. And we love the idea because it's like, hey, I'm not having to deprive myself from not struggling and has nothing to do with going on diets. I commit these small simple changes. And I'm not going to eat as many calories, and I'm gonna lose weight and be healthy. They're all being retracted. I think the take home message from this is what we should start at the beginning. There is no free ride. If it was easy and simple for us to all eat, healthy and lose weight. We would do it. The problem is that it's hard. And especially with the way that we consume American diets today, it's very difficult to change them sometimes in ways, which allow us to eat fewer calories or eat better and lose weight and sustain that over the long term. And these simple quick fixes which make a lot of news and sell a lot of books and make from great stories and some people's careers. Don't turn out to be true. And unfortunately, in this case, it's resulted in some serious issues for for the researcher as well as as well as the lab that he worked for. But I think in general we have to acknowledge the fact that almost with almost everything and even with food. There are trade-offs. There are no quick fixes. There are no easy solutions and often the stuff that seems really sexy and novel is not nearly as true as the conservative boring moderation type stories, we'll Dr Carol you're a pediatrician. And so I am guessing that you have in the course of your career had to tell parents give parents advice about how to feed their kids. How to encourage their kids to follow a healthier diet. What do you say? What research supports it? So I think it's you got to make small steady changes part of it is trying not to snack. You know? Snacking is a problem. I think the more you eat the more eat sometimes being active can help not in the sense that exercise leads to weight loss, but with kids sometimes keeping them moving and not sitting on the couch that is often associated, you know, sitting around with a meeting because I'm I'm just bored. And so trying to change that try to change, you know, what kids eat and trying to eat more healthy diets is the same simple stuff that we would tell adults try to eat more fruits and vegetables, try to avoid processed food as much as possible not because it's full of chemicals, and because it's odd, but because food that is processed by processed. I mean, something has been done to it from ingredients before you eat, it, it makes it easy to eat more than you would otherwise like the bread makes it easy to eat wheat. Pasta makes it easy to eat flower. Those are both processed foods. You don't have to think often of just even really company produced food or industrialized food. But the more that you can stick to ingredients the more that you can get kids and families de well, balanced diets than involve lots of different foods, probably the better off, you're going to be I also advocate for trying to remove sources of added sugars in the sense that that is just empty calories and too much of our processed food in general just contain sugar because they know it will sell more. That'd sugars empty calories. It's just not necessary. It's associated with the host of issues, and it's often something that you can eliminate pretty easily without having to radically change kids diets and often by making these small changes over time. You can see decent size results. Dr caroli- promised. I would ask you about pharmaceuticals. You were talking about negative research. It's not very sexy and doesn't get published. This. Same thing can be said for pharmaceutical research. Right. Oh, absolutely. In effect, the article that I wrote on publication by Senate around antidepressants because that is such a great example to get drugs approved by the FDA pharmaceutical companies have to show that they are safe and effective, but they often just need to do that in some studies, and they only need to sort of really promote the ones where they are positive. And so some researchers actually took I think it was one hundred and four studies and what they did with actually went to the FDA websites where all the data exists. Not just the things that are published and they found that of the half or so of trials that were positive almost every single one of them was published in the peer reviewed literature, but only about half of those that were negative actually were published and so they don't have to publish the negative trials. And so we don't necessarily know how much negative date is out there will enter the positive data. But then it goes even further of the negative trials. They often change the. Outcomes to pick secondary outcomes that looked positive instead of the ones that were primary negative and that makes even some of the negative studies look positive, and sometimes they even put spin on negative results to make it sound like negative results or positive by talking about trends in the data or by citing numbers, even if they're not statistically significant. So if you take all of that together while as I said at the top about half, the studies reposited in half the size or negative. If you include the idea of we're gonna publish not are we going to cherry pick out comes in. And we're going to spin more than ninety percent of the literature looks positive, and that's how you get the sense that antidepressants massively work when again only half of the studies that we're actually done turned out to be positive. And so by moving or changing, how these things are published on and how they're published you can actually really affect how people think about them far more than the results would actually show. You mentioned say. Safety and effectiveness, and I think those are really important terms. But we don't really know what they mean. So. What do we even mean when we say safe ineffective? So f- effective means that in the actual clinical trial. They saw a different. So in an ideal perfect world situation, you see some kind of result from the drug more than placebo safe means that it did not have a significant level of harms or adverse events that occurred from it. But of course, you know, drugs can be have efficacy without affective nece, or how will they work in the real world and just because they have efficacy in a set small population. For study doesn't mean it's going to work in a large much larger population in the real world that can get us into trouble. You've been listening to Dr Aaron Carroll. He's professor of pediatrics and associate dean for research mentoring at Indiana University school of medicine. He's also director of the center for pediatric in adolescent, comparative effectiveness research his books include the bat food bible how and why to eat sinful. If you go to our website, WWW dot people's pharmacy dot com, you'll find a link to his New York Times say about the alcohol research, we also have a linked to the research itself in the Lancet after the break. We'll talk more about drug safety. How should we be talking about benefits and harms? So that we understand them better. What's the difference between relative and absolute risk? Why does it matter especially when it comes to medications like Lipitor, how can you as a consumer figure out which health risks, you really need to worry about Dr like the rest of us have a hard time unlearn things. Why is that a problem? You're listening to the people's pharmacy with Joe and Terry Graydon. If you do the health information, you get when you listen to the people's pharmacy consider subscribing to our Email newsletter. You'll get the latest health news and information on upcoming podcasts delivered to your inbox twice a week look for the link at people's pharmacy dot com. Welcome back to the people's pharmacy. I'm Joe Graydon. And I'm Terry Graydon to purchase a CD of today's show or any people sperm ac- broadcast. You can call eight hundred seven three two two three three four. Today's show is number one thousand one hundred forty one that number again, eight hundred seven thirty to twenty three thirty four you can also place the order at people's pharmacy dot com or you could download the free podcast from I tunes or from our web store. We invite you to consider writing a review today were trying to understand benefits and risks. You frequently read about a new drug that reduces the risk of some condition by twenty or thirty percent. But what does that really mean for you drug safety ineffectiveness seem clear those are the criteria for drug approval. But the. FDA has a lot of leeway on such definitions. How can you tell if a medication will really help our guest is Dr Aaron Carroll? He is professor of pediatrics and associate dean for research mentoring at Indiana University school of medicine, he's also director of the center for pediatric and adolescent, comparative effectiveness research, he's written three books a debunking medical miss. The most recent is the bad food bible how and why to eat simply Dr Carol you mentioned that safety is one of the criteria that the food and Drug administration uses before they approve a medicine, and yet all you have to do is turn on television these days, and you will see prescription drug commercials for consumers in which they mentioned side effects like cancer lymphoma, for example, or heart attacks strokes kidney disease. And sometimes they even say, including death, including death, and you go. Whoa. Wait a minute. How can that be safe? And by the way, people are having an absolutely fabulous time as all aside effects and being mentioned, I've counted six different smiles during one little voice over with side effect information. Well, that has to do with the way that we talk about or fail to talk about benefits and harms I think with respect almost everything in medicine. We just talk about them as if they are binary they exist. They do not exist versus trying to actually quantify how much benefit versus how much harm you might receive. So look, I I have all sort of colitis. I take medication Bioserve colitis. It has significant potential side effects. Have my blood drawn every three months to make sure that I don't have what's called a plastic anemia or the idea that my my bone was shutting down, and I'm not creating read one white blood cells that sounds. Horrific, but the benefits that I get from taking this drug are massive and the absolute risk of having a promise, very low. It is worth it to me for many of the drugs might see on TV it's possible that there's a quantifiable large benefit that people might achieve quantifiable small risk. Even if those things sounds scary or it could be the opposite. There could be a small benefit and a relatively large risk. You can't tell from those commercials. But that is how we've sort of mandated that those commercials exist. They have to by law state, no more benefit than what is actually true, but they have to state what harms exist, but they don't do them in ranking order, they don't quantify how many people might get them. And so they can panic you correctly or incorrectly, and they might not panic you incorrectly or correctly. Unfortunately, would just won't ever know from those types of commercials. Well, one of the techniques that commercial sometimes use t use the relative benefit when it comes to benefits because it sounds so much more impressive than than absolute benefit. So for example, some years ago there was an ad for Lipitor. I believe suggesting that there was a thirty percent reduction in heart attacks for people taking Lipitor, which if he looked at the data from the study was true. But it was because if you had one hundred people taking Lipitor for five years and two of them would have a heart attack and three of the ones taking the placebo would have a heart attack during that five years. So one person out of one hundred over five years that's a thirty percent reduction. It sounds a lot less impressive. When you put the. Numerator denominators in it absolutely does. And so this is definitely one of the ways that companies try to mislead us. But also the ways that the media can often mislead us because they will almost always cite the relative risk. So relative risk is exactly what it sounds like. It's the relative increase. So if I go from ten to twenty percent, I had ten percents before to twenty percent now I have doubled that is twice as large. If I went from twenty percent ten percent. I had a fifty percent reduction. But the absolute reduction is the difference between the two, and so I went from twenty to ten that is only a ten percent reduction, of course, ten percent reduction. Still sounds great. But fifty percent, but sounds better. But if I also go from point oh two percent, two point. Oh one percent. That is also a relative fifty percent reduction in the news story that is what you will hear v. Fifty percent reduction. It's also what you on the advertisement. Even though we only really had an absolute reduction of point, oh, one one of my favorite examples in this involves red meat where if you believe the stories that say red meat causes cancer. Even though those aren't randomized controlled trials. They will cite the fact that that they believe that processed red meat. They believe increases your lifetime risk of getting colon cancer one serving a day by eighteen percent. That sounds horrifically scary. Eighteen percent increase in getting colon cancer of our lifetime. Sounds huge. But that is a relative risk. So if we want to look at the absolute risk, we can also do that. If I went to the National Cancer website than I entered all of my data. And I'd have to pretend that I'm fifty because fifty is the youngest it goes for it. They would say that I have a lifetime cancer risk of. I think two point three percent. If I then say, I'm going to eat three extra pieces of bacon every day for the rest of my life. Which I'm not going to do my risk would go from two point three percent to two point seven percent. That is a relative risk of eighteen increase of eight relative increase of eighteen percent. But the absolute risk increase was point four percent, very low. But the eighteen percent sounds scary. The point four percent does not. And that's also if I if I choose to eat an extra three pieces of bacon every day for the rest of my life. That is not have anything to us. I said by want bacon once in a while. Which is really how they tried to scare you. But that's a massive difference between the relative risk increase, which they will often scare us with and the absolute risk increase. This happens all the time with talking about alcohol and cancer again with talking about other things we might do where we will focus on the relative risk and say it goes up by four percent by ten percent. Even by twenty percent when the absolute risk increases are very small often even less than one percent. Dr Carol our listener. I get very frustrated when they hear people talking about the dangers of alcohol the dangers of red meat the dangers of butter the dangers of this and the dangers of that. And they want to know. Well, well, Dr Carol how do I make sense of those confusing headlines where they try to scare the heck out of me based on relative risk. How can I get to that absolute risk information and not just when it comes to the risk of let's say some sort of food item. But also when it comes to medications, how can I determine how effective my medicine is going to be in the real world. Not just in some clinical trial where perhaps the data has been very carefully cherry-picked. Well, it's really hard. You can do it often by going to the actual research papers and reading them, which is what I do. But of course, I'm not expecting that everybody in the lay public is going to do that that that is what? Try to do in my Collins what I try to do in the book is try to lay out and bring the research lights that you can see it. It's hard. The media could do a much better job of trying to quantify the absolute risk changes. With all these things not just the relative risk changes. But I would also argue that we need to take a better view of risk of not only looking at one side of not just looking at harms, but also benefits. The example, I always use the number one killer in the United States have children by far is accidents car accidents, kill more children than almost anything else that we could pick. No one ever says we should not drive because so many children are killed by cars. We accept a certain number of children are going to be killed by cars because we know that this aside benefits of driving are phenomenal. And therefore we can make a logical decision that driving while increasing the absolute risk of death. And the relative IRS could death by quite a bit is worth it. We don't have that same kind of commonsense, balancing, benefits and harms in so many other things we do. Let's take the bacon example. I used a minute ago. I like bacon. It may be totally an reasonable for me to say, I'm going to take a one thousand chance that over the course of my life. I might get cancer. If I want to eat bacon every day because that's how much I love it. I'm probably taking a much lower risk because I'm not eating bacon day. But the answer is not to eat. No bacon. You have to sort of judge. What it is. We know that the sun raises your risk of skin cancer. Now, one says never go out in the sun. They say take proper precautions and think about how much doing because you don't wanna raise your risk too much. We can make balances and recognized that there's good, and there's bad and all of these things try to quantify them and measure the difference. And then determine what is the right decision for us. But that's all. Often. How news stories are not pitched and how recommendations are not done. They only focus on one side. And not the other scare you with large numbers and never make any kinds of trade offs. And I would argue that one trade off you always need to consider as a benefit is joy, you know, some things are quality of life improving, and they are more quality of life improving than the actual harm. You're accepting that is perfectly rational and reasonable. I think chocolate might fall into that category. You go perfect example for me. It's scotch. What she sticks. Dr carol. You have written that it's hard for doctors to unlearn things. Why is that a problem? So it's first of all let's knowledge, it's very hard to get human beings to change behavior. It's very hard for doctors to change behavior. There's there's some studies that say it takes fifteen years for something to sort of be proven in the medical literature, and then to finally have trickle into clinical care, but as hard as it is to get doctors to do things it's almost harder to get them to undo things part of that is because it's hard to change behavior. Part of it is because we at some level get paid to do stuff. That's not to say doctors are committing fraud or that they're trying to do extra work. It's just that. You know, they often have done things for a long time believed that they are doing good. They start to believe the causalty exists when it's just an association, they start to believe that the thing they are doing causes good. It is very hard to learn that behavior. The example, I used a reason column. Mm was talking about recommendations for how tightly we should control people's glucose levels when they're very sick. And intensive care units for a period of time. We thought we should really be on it and tightly controlled their glucose. And while we thought that the number of doctors were actually doing that increased steadily. But slowly, but then new reserved came out and said, that's a bad idea causes harm, and there's no benefit you should stop immediately. And it didn't trickle down it sort of just stayed constant because it's hard for them to unlearn behavior. I also point to a campaign called choosing wisely from the American board of internal medicine, which asks specialty groups to identify five or ten practices that their specialty. Does that all the evidence says you should not do this? I mean, basically, it's just directives don't do something. There are something like six hundred different recommendations as of this moment on the website of things that doctors should not do. Do that. We still do all of the time. It's very hard to change behavior and get doctors to stop doing stuff. How do we change that we could try to have perhaps different incentives, and in the way that we pay for things to try to get them to change, but it's very hard. And unfortunately, those actions don't do good. They don't have a quantifiable benefit they do have a quantifiable harm. They also have a very quantifiable cost and this is pure waste. It's a significant part of the healthcare system. Probably the single biggest bit of modifiable savings. We could get at and it would probably help us to do good. But it's very very hard to get physicians and not just positions but lots built, but certainly physicians to stop doing things, Dr Carol your profession. Your specialty has come under scrutiny over the last couple of decades for all those tonsillectomies that were performed back in the. Fifties and sixties, and then all those antibiotics that were prescribed for Erin factions, and then all those ear tubes. So how do we influence pediatricians to be a little more cautious? So tonsillectomies is sort of the perfect example for something that was being done that everybody thought was doing good that turned out not to him. And it was Jack Wynberg, and what later became the Dartmouth atlas that pointed all that out because he basically showed that there was huge areas of variation in the United States in the rates of tonsillectomies. There were being performed without any kinds of improvements and outcomes which proved that really didn't do any good an overtime. That's changed. When I was a kid. There was even a curious George book, which is pretty much entirely about going to hospital. Arguably about getting kids ready for tonsillectomies that don't need to happen. Antibiotics have been harder to fix the problem with antibiotics is that again, that's a good example of a mislaying of the benefits in the harms people think that antibiotics are going to cure pain and. Twenty four hours. You know, my kids and pain give me an antibiotic it never does. There's no study that's ever shown at twenty four hours will some antibiotics can cause a reduction in symptoms over the course of say two seven so that would be a benefit, but the number needed to treat is closing in on twenty on the other hand about the number needed to harm or the number of kids, you need to give an antibiotic to to give them a rash or vomiting or diarrhea is like nine. So I tell parents all the time especially when it's low risk ear infection. If I give you an antibiotic twice as likely to cause a harm as I am to give you a benefit when when portrayed in that manner. Many patients will choose not to get the antibiotic. But too often patients think there's only an upside to antibiotics, and no downside and say physicians feel the same way. All of this. I think in a lot of our conversation has been a good example of the ways that we just don't think of a whole picture when it comes to medicine everything in in health and medicine is a trade off there are harms and there are benefits and every individual decision that we make what we eat in. What medicines we take what actions are therapies, we're going to undergo should think about one of the the actual benefits and quantifiable benefits, I'm going to get what are the actual harms or quantifiable harms, I might get. And if I put them on a scale, which is more important to me, and if you take that kind of holistic outlook and think about it you're gonna make far better decisions for yourself. And we as the society would probably make far better decisions about what things we do. And do not want to do Dr Aaron Carol. Thank you so much for talking with us on the people's pharmacy today. Thank you. You've been listening to Dr Aaron Carroll professor of pediatrics and associate dean for research mentoring at Indiana University school of medicine, he's also director of the center for pediatric and adult comparative effectiveness research his research focuses on the study of information technology to improve pediatric care healthcare policy and health care reform. In addition to his scholarly activities. He writes about health research and policy for the New York Times, among other outlets. His most recent book is the bad food bible how and why to eat simply lean Seagal produced today show. How would our ski engineered Dave Graydon edits? Our interviews. The people's pharmacy is produced at the studios of North Carolina public radio W and see the people's pharmacy. Theme music is by B J Liederman to buy a CD of today's show or. Or any other people's pharmacy broadcast. You can call eight hundred seven three two two three three four. Today's show is one thousand one hundred forty one the number again, eight hundred seven thirty to twenty three thirty four online at people's pharmacy dot com when you go to our site, you can share your thoughts about today's show. How did you determine benefits and risks? If fifty people have to take medicine to help one person get a therapeutic effect. Do you think that's worth it? What about risk how do you figure out which health threats you care about in which you can ignore? Please. Share your story in the comment section for today at people's pharmacy dot com, you'll find links to Dr carols article in the New York Times. And the study we discussed you can also sign up for our free online newsletter or subscribe to the free podcast of the show. When you subscribe to the newsletter, you'll get our free guide to. Favorite home remedies in Durham, North Carolina. I'm Joe Graydon. Can't I'm Kerry grading, thanks for listening. Please join us again next. We hope you enjoyed this podcast if so please consider taking a minute to write a review on I two and thanks for listening to the people's pharmacy.