High School Science Fair & Clinical Research
Podcast Transcript
Original Air Date: May 13, 2022
Introduction
Welcome to the MedEvidence podcast hosted by Dr. Michael Koren and Michelle McCormick. MedEvidence, where we help you navigate the real truth behind medical research with both a clinical and research perspective. In this podcast, we'll discuss with physicians with extensive experience in patient care and research. How do you know that something works in medicine, we conduct clinical trials to see if things work. Now let's get to the truth behind the data.
Michelle McCormick
Welcome to MedEvidence! Truth Behind the Data with Dr. Michael Koren. He is a practicing cardiologist and chief executive officer at ENCORE Research Group, which conducts trials across Florida. He has been the principal investigator in over 2000 trials and has been published in the most prestigious medical clinical journals. Dr. Koren received his medical degree cum laude from Harvard medical school. For more information on local trials, visit encoredocs.com or call 904-730-0166. Well, in this episode, MedEvidence! Truth Behind the Data with Dr. Michael Koren, we are actually going to talk about the truth behind the data. Well, Dr. Koren, in the first segment, we talked about lady-tasting tea, this experiment, and why it's so important to what you're doing now with clinical trials, but you mentioned some statistical concepts let’s dive into those a little bit.
Dr. Koren
To remind people who missed the first segment, the lady tasting tea experiment was a member of the British royalty who claims that she can tell if milk was placed first or tea was put in first when she had her mid-afternoon tea, the British love their tea and her claim was that she can tell which was put in first whether she saw it or not. A lot of people believed her, and some people were skeptical, and a very famous statistician named R. A. Fisher said I can test this to see if she's telling the truth, and this is the basis of modern clinical trials, believe it or not.
Michelle McCormick
Well, I look forward to finding the response to that.
Dr. Koren
There are three important elements of hypothesis testing in modern clinical trials. They have to be prospective, blinded, and randomized. Let’s talk about what each one is. Prospective means you are putting the thing together before you have any information, so prospective means ahead of time. So, for you to do a good research study and run a clinical trial or do a good hypothesis test. You need to set up all the rules for the test beforehand. It will be cheating if you say okay, we will make a bunch of observations, and then we will take a look at it later. And so hard science and a good hypothesis test require that you lay it all out. This is what we are going to do? This is how we are going to do it. And this is the number of observations that will be required to determine whether or not we can establish the truth.
Michelle McCormick
Yea, it sounds like a very fancy science fair project.
Dr. Koren
Well, remind me to tell you about one of my kids’ science fair projects.
Michelle McCormick
Okay, I will
Dr. Koren
So perspective is a very important, planned experiment. We had other sessions where we talked about the very famous smallpox experiment back in 1721 at the Newgate Prison in the UK that was considered the world's first clinical trial because it was a planned experiment of vaccination when the first experimental test showed that vaccination actually works in a clinical trial rather than just based on general observation which is inductive reasoning versus clinical trial which is deductive reasoning. Prospective ahead of time has to be planned. Second, Clinical trials have to be blinded. Why does it have to be blinded? People cheat.
Michelle McCormick
Well, This reminded me of Pepsi versus Coke back in the day there was the blind tasting either they have blindfolds on, or they have a barrier in front of them.
Dr. Koren
Right, so if you saw it was Pepsi or Coke, which tastes better, it’s whatever you believe, right? If you do not know what it is, it’s more credible that it is blinded. Blinding is very important. We may think we are objective but will have biases, and biases are absolutely impossible to overcome. So, you have to do something very specific to overcome biases. And the most effective way is blinding. And in science and clinical trials, we call it double blinding. So, in double-blinding, we have a situation where both people who are volunteering for the experiment and the experimenters, both are unaware of which assignment they have. So that there is no bias from either side and there is no bias between interactions. So, for example, if you are tasting Coke or Pepsi and you do not know what it is because you have blindfolds on, but I know what it is, then I could see the bias.
Michelle McCormick
I was going to say you can direct me, are you sure you don’t like this one better?
Dr. Koren
Right, if you are a Coke salesperson, you can say well, do you like the first one, that incredible fizz one, or do you like the second one? So, if both sides are blinded, there are different ways that you could influence people.
Michelle McCormick
Is that where the placebo comes in place?
Dr. Koren
Placebo is one form of control, which is not one of the three things, you are going to the fourth thing.
Michelle McCormick
I jumped ahead
Dr. Koren
So, for example, in the lady testing tea, whether the tea was first or milk first, the placebo is that you are just trying to see if someone is telling the truth. So, you do not need a placebo in the hypothesis testing, but we often have one. And the third thing is randomized. So randomized means that when you are doing your observation when you are going through an experiment. There has to be a random order of things. For example, in the lady tasting tea, if you didn’t randomly mix up the milk first or tea first, you can figure out a pattern, so let’s say that you did five in a row that was all milk first, and then you switched it, and the next five the tea is first. Well, the human brain may be able to perceive a slight difference, the same for a few times, then a different, and so that may actually help the person along the way, so randomization makes it much more difficult for you to find a pattern. So, for example, let’s say that you are testing a blood pressure drug and you know that the blood pressure drug should work better in people who are overweight if it was a random assignment of patients of all the overweight people in one arm and all the people in the other arm and that may include the results, but if you have a random number generator then you are coming in we don’t know if you are overweight or not. People will be assigned randomly. Any possible differences will sort themselves out by chance, and the number gets big enough that you are almost certain they will sort themselves out by chance.
So those are the three key elements of good hypothesis testing, and that was what R. A. Fisher was able to teach us and to use this incredible and powerful concept to establish the truth. Now you have the fourth one, which is placebo control. When you say placebo, it means the control is inactive, when we do medical drug testing or a medical product evaluation, we often use a placebo. For example, a new drug is coming on the market and we want to see whether or not it is effective the best way to do that is to put it in a form where you can’t really tell what it is and compare it with something that is inactive and give it one or the other in a prospective or randomized fashion to a group of patients and see what the outcome measurement is to see if it is different. For example, if you are testing a blood pressure drug against a placebo, you will have a pill or an injection that looks identical whether it is active or not then you would give it to a bunch of people, and then you measure the blood pressure and if its an effective treatment the people who have the active product will have lower blood pressures after some time after the intervention than the people that were on the placebo but there are other forms of control. So controlled experiment doesn’t necessarily mean a placebo so for example, if you want to show that drug A is better than drug B, but you really are testing drug A, that is your permanent hypothesis, then drug B is called a controlled substance, and that is comparative. Now we often do that when there is extensive care when for example, in a lot of blood pressure studies, you wouldn’t want not to treat patients for an extended amount of time because it will be unethical because we know blood pressure lowering saves lives, prevent stroke and does other things that are positive for the human body. you may get away with a short-term drug versus a placebo. So, in that case, you need to use some controlled agent, so that way everybody is getting something so you can see the new agent is an incremental improvement compared to the controlled agent. That is a key part that we do in development. That is a key point that we make every day.
Michelle McCormick
What is the history of clinical trials? When did they first begin?
Dr. Koren
So, there is a lot of debate on that, getting in the faith thing, in the bible, there were choices people had to make, and there is a passage about Daniel in the bible and the Babylonians. So we sometimes talk about that for the first experiment it is really not a clinical trial because they didn’t have all those elements such as prospective, blinded, or randomized.
Michelle McCormick
Right, which grew over time, I am guessing?
Dr. Koren
But basically, Daniel said that if the Babylonians eat rich food and I eat basic bread and water and healthy foods, then I will be stronger than you, and in 10 days, that was true. His hypothesis was that he chose one diet, and there was a control; basically, the outcome he picked he did better. So that is an example, but we don’t consider that modern-day clinical trial.
Michelle McCormick
You mentioned smallpox was that the first modern clinical trial?
Dr. Koren
Right, and even that is not perfect by any means, but back in 1721, there was this theory based on observations from Africa that if you had smallpox, you could protect yourself from severe complications of smallpox by purposely infecting yourself in the leg or other parts of the body with small amounts of the smallpox virus in your body. Smallpox was very devastating for hundreds and hundreds of years, and the mortality rates during the smallpox pandemic were 33 % about a third of people in the community would die of this infection. So obviously, people were scared to death when they learned that a smallpox pandemic was coming their way. And what people observed is that the smallpox pandemic survivors are much less likely to get sick during the next round of the pandemic. And so, they said, why is this? Ultimately, they figured out that if you expose yourself to this virus in a certain way, you can protect yourself. if you get smallpox in the respiratory means and you are not protected, you can get really sick and die, but if you got injected with smallpox, they found that you had a small local infection that seems you can protect yourself. This was an observation in the middle east in Africa, and during the pandemic of 1721, the British royalty was so worried about this that they took these crazy ideas from the middle east and said let’s, try it. So the researchers went to prison because, at that point, it is considered ethical that if you want to get out of jail, then you would be part of the study, and we will set it up and give consent, that we actually practice every day, but in that case, it was not the consent of the patient but the consent of the king, but it was prospective, and it was consent-based, and it was an organized trial where they took six people and gave them these surgeries cut their leg and put smallpox in those wounds made sure they had an infection and saw how these people compared to the prison population in terms of whether they contracted smallpox or died from it. And all six people in the experiment seem to be immune to the effects of smallpox. It was considered the first clinical trial because they looked at a group, and they had a control group and a comparative group.
To continue listening to the MedEvidence Clinical Trial series, listen to June 22, 2022 episode on the History of Clinical Trials.
Michelle McCormack
I am your host, Michelle McCormick, and we want to thank Dr. Michael Koren for his clinical and research perspective behind the science in this episode of MedEvidence the truth behind the data.
