Randy Neil Biostats 05: USMLE Step 1 Biostatistics - Drug Ads, Abstract (Part 1)

Scan the questions first. Identify math/table/graph questions before reading the full passage.

The result is not statistically significant because “no difference” is included.

Choices directly proven by the data provided in the table, graph, or abstract.

Answer: B) 5

ARR = 0.35 − 0.15 = 0.20. NNT = 1 / 0.20 = 5.

Answer: C) Not statistically significant

For ratio measures, a CI crossing 1 includes “no difference.”

All right guys, so here's our intent to try and do the drug ads or the abstracts. It would help if you could send some, actually some examples this way because I pulled these from the USMLE website that gives us free practice questions for us. And so there's a, we got several here and we're trying to develop a kind of a process or a procedure to follow. And it helps if we had more questions that we can kind of use those to make sure that we can implement it correctly and efficiently. So so far what we're going to do is you're going to look for the math problems within, always go to the answer choices first and look for the math problems because those are ones that we can zero in on, use the tables or the charts and actually actually get those done first and right away. And then we're going to look for the other answer choices where we might have to actually read the entire abstract or drug ad and then get rid of the bad answer choices. But the main emphasis that I want to do is make sure you only use the answer choices that you can prove from the data. There's going to be answer choices that may, they may look good, they may sound good, but the end of the day, if you can't prove them from the data, then you really can't use it as an answer choice. So it's got to be a better answer choice that's going to be out there. So it's just kind of a little bit of a process. But if you can send any questions, any drug ad examples that you have to that, that Randy dot me on the number eight at gmail.com. I said we can, we can use those, make a video and just kind of run through these and put as many as these together as we can. Just so when we get to the step exams, we'll be more confident and we won't kind of fear when we see those little three questions that are connected together on that little box. And so I hope this helps. I hope you enjoyed the video and see the works.

Okay, guys. So we're going to try to solve these drug ad problems. And this is what we're going to kind of base this off of. I need to know this formula. Okay. And you know, we don't ask you to memorize too many formulas. But this is one of those formulas to where you're going to have to actually know for the test. And really it's just going to be for it tells statistical significance. Okay. It's going to come in handy. Now remember, when we talk about p values, all we ever care about is 0.05 or less for the most part, right? Because that tells that there's some significance. So, a correlating value of z in this trust me on this for now. So, the correlating value of z is going to be 1.96. So we punch in two different numbers during this little formula, which you're going to use. And you can kind of eyeball the number is that if you come up with a 1.96 or that number is greater than 1.96, that means your equivalent p value is going to be 0.05 or less, which will make, which will inform that there is significance. Okay. So we're going to use this simple test to determine a score. And if that scores at least this number or greater than it will have a p value that actually is meaningful. Okay. So, need to know that. And how we're going to attack these problems is going to be this. I want you to scan the questions and I want you to scan them for math problems because we're going to attack the math problems first because they're the ones that we're going to solve the easiest and the quickest. Number two, I'm going to say exclude the bad choices. Okay. Every question has them. There's always going to be one or two that are just obviously not it because the third step is we're going to select only what I can prove. Okay. And when I say what you can prove is that it's only what's based on the information given to you. Because there's going to be choices that sound good. But

at the end of the day, I'm only going to choose the ones that I can prove are there. Okay. So, just your basics and these are for the drug ads and all the abstracts and all that kind of stuff. I want you to know this formula because it will come in handy at some point because it will give you a score with the first score of 1.96 or greater. Use your calculator. Scan the questions first. Look for math problems. Exclude the bad choices and select only what you can prove. Okay. Here's the question that was taken basically. This is taking taken from the usml.org practice questions. You can see it's pretty long. This is an abstract. Now I'm not going to say you got to read this right off the bat. We can use scan it and then move straight to the questions. Okay. Here are the questions that came with this thing. There's three of them, right? So, they look pretty extensive. But let's just look at them real fast. This one, again, always go to the last sentence and ask what the question is. Approximately how many patients must be treated with EPCS rather than EST. Now EPCS stands for emergency poor-okable shunt and then EST stands for endoscopic sclerotherapy. So there are just two procedures, long story short. Two prevent one case of recurrent portal systemic and sub-alopathy. So when they say that how many patients do need to treat to prevent one, okay? Well, they're saying number number needed to treat. Okay. Well, we know from that very first video that the number needed to treat is actually the absolute risk reduction. Okay. And the absolute risk reduction broken down is the event rate control minus the rate treatment or backwards and forward. We don't care because it's the absolute value. So essentially, I don't care which one comes first. Just put the bigger number. Whichever one's easier to subtract. So anyways, this is a math problem. Good. Which of the following is

the most strongly limit-sub-generalized ability to study? Okay. So this one I have to read the problem. Okay. So that's going to be one. I'm going to wait a minute to do. The third one says, which of the following conclusions is most appropriate and based on the results presented in the table. Okay. Well, if it's based on the table and to me, that's going to be some type of math problem. So initially, I'm just going to do these two first-year choices because I can immediately go back to my abstract or drug add and go straight to the table because this is what's going to allow me to solve this problem. So I know it's kind of hard on this to go back and forth, but remember on the very first problem here, the very first problem. It said, it reads, 50-year-old man with hepatic cirrhosis comes to the emergency part of him because of a three-hour history of vomiting blood, esophageal gastrodoid and scoppy, confirms active bleeding, esophageal varicis. Based on the abstract shown, okay, this thing, which we're going to read in a second, the physician is considering emergency protocol-shunt procedure rather than in the scoppy slurthera, PST. According to the results in the abstract, approximately how many patients do you need to be treated? So it's number two in the treat between EPS, EPS, and EST. So we go back here and it's, remember what I said, it was for the recurrent porosystemic and cephalopthi, which is right here. We look at EST, EPS, okay. So again, all we're doing is doing the math problems first, which are in the table. Now we have to find the event rate. The recurrent porosystemic with this procedure is 35% with the EPS, it was 15. So we have our two numbers there, right? That's all the number needed to treat actually asked for. So we don't care which one comes first, right? So we go back there and essentially it's going to say 35% or 0.35 minus, right? That's

part of the formula, 15% and 0.15. And all we do is take that from the chart. So we do that, 0.35 minus 0.2. And however you like to do your math, you can use that calculator. It's on the exam, but 0.2 is two tenths. We do the inverse, all that kind of good stuff. And we would get a number needed to treat of five, okay? So again, the first problem was strictly math. Did we have to read this entire asked track to get that? No, we go straight to where the data supports it, okay? It's a very quickly to get going on these problems. The second math problem says, which of the following conclusions is most appropriate based on the results presenting table. Okay, so we've got to read through these and we're going to apply that whole, you know, exclude the bad choices and select only the ones that I can prove. Is it the 95% confidence interval for the difference in the survival between EPS and EPS, T, for this child, you, I don't know, class A patients includes zero years. Well, you know, when I look at this, I'm not seeing anything in the table that really talks about confidence intervals, all right? There's no, even when I scan this thing, it doesn't really talk about confidence intervals real quick. The only time I see confidence interval is right here, but it's not talking about this, I'm not talking about the child pew and forgive, I'm pronouncing that incorrectly. When it talks about those classes, the 95% confidence interval between the difference in the survival between these patients includes zero years, I really don't see anything in the chart that talks about confidence intervals. So for me, I just kind of put a little question mark saying, okay, well, I really don't know. I don't have any data to support, yes or no, and that is just not there. That EPS is more effective than EST and decreasing hospital readmissions for various affairs, so I'm leaving, okay. Well,

I know it's got to be in the chart because I told me, you know, they told that this table is going to have everything, so that EST, again, the question is that EPS is more effective than EST, that EPS is more effective than this one. Well, for the number of readmissions, so here's readmissions, okay, and so we're dealing with these two numbers right here. So the mean number of readmissions was particularly less, all right. It was less in the EPS that it was in the EST. So for that one, as far as mean number of readmissions, I'm going to say, well, yeah, the EPS was more effective because it had less number than the EST. So this one's looking like a strong, you know, I'm going to give it like a half thing there because I really like that one. At least I like it a lot better than I like answer choice A. C says the median survival after EPS is statistically significantly less for a child if you class C rather than B. Now here's where we're getting into that whole formula, okay. And remember that formula that we're going to have to remember is Z and it's A minus B over the square root of A plus B. Now all that means is we got two numbers, right? A and B, two numbers. And we're going to see are they statistically significant or not? And this goes back to taking two numbers, plugging them in and getting some number that we're hoping to say it statistically significant, it's got to have B points here or five or less. So that number when we punch it in and it comes out the other end, it's got to be 1.96 or greater. So let's look at this. The median survival after EPS is statistically significant, significantly less for class C than class B. Okay, so well, let's look at this. It was for EPC, EPC, yes, which is right there and it was comparing class B. So the 6.19 with class C, which is 5.3. So there's my two numbers, okay. So all I know, all I care about are two numbers at

this point. And let's just take this right here. So when it said class B, that was what? 6.19, class C is going to be 5.3. Now this scene might seem like overkill, especially if you already say what did you say, choice B was the best answer? Yes, but we're just kind of trying to do for completeness here. And this is a very simple test that if you get used to doing it, it can show you whether or two numbers are statistically significant or not. Okay, so anyways, if we did this A minus B, so 6.19 minus 5.30, and we're going to get 0.89, okay. So 0.89 goes on top and then A plus B, we're essentially going to get, A plus B is going to be 9 for 11.49, and you got to take the square root of that. Of course, you can get a calculator and punch all the numbers, but essentially you're going to get 3.39. And again, no one, you're not going to be able to do this in your head. If you would, you probably not going to be watching this video. And so if you put 0.89, you've got to have by 3.39, you are essentially going to get a number called 0.26, okay. Now, all we said that you had to remember is 1.96. Is that number greater than that? Any answer is no, right? It's some number a lot lower, so the p value is going to be too big, so there is no statistical significant. It's not statistically significant, that correlation between the two, the difference between them is not statistically significant. So answer choice D is going to say the randomization procedure was ineffective in decreasing the bias in the study. It's hard to say when we talk about biases in this, it really, I can't pull that from the chart and up here if I scan, I really don't see anything that's talking about bias when I go through here. So I'm going to say on that one, I don't know, it's possible but I don't know. So when I do my answer choices for this one, the only one that I can prove, remember, the only one

that I can prove is going to be answer choice B. So again, it's seem like a long time for these problems, but the more you do these and if you have that little process of scan the questions, do the math problems first, okay? Do the math problems first. And then you come down to the very last one and says, which of the following most strongly limits the generalized ability of the study? Okay, now generalized, generalized ability is just how much this study applies across the board, right across the entire population. So we got to say, was it because the allocation was because you know, now here's a deal. We got to, at this point, I'm not sure what all this means without actually scan and reading it. So I'd say, okay, okay, in the patient with cirrhosis, acute bleeding barricies, okay, and it's comparing control of bleeding and survival versus versus those two procedures, okay? My method, it was randomized, controlled. Allocation was concealed. You know, allocation is essentially, you know, how you disperse the patients out. So it was blinded, it was only blinded by the provider there and who evaluate the patient. So it led and applied the patients for blinded, followed up 17 years, this, there's the end, okay? And again, just kind of scan through here and eventually it almost becomes too much. So then for me, I go back to this, I go back to the problem. Which of the following most strongly limits it? The patient was concealed, okay. Well, maybe, do I know? Well, they didn't really tell me much about the allocation and anything other than that it was, you know, it wasn't given. The EPCS is available only at special, specialty centers. So with that limit, it going across the entire population, if it was only performed at a certain center, okay? Sure, maybe. The follow-up period was too short. I don't know, the thing said the follow-up period on this one was 17 years. So

I don't know, it seems a bit like kind of long to me. So to me, it's just not a great, it's a, it's a maybe, but it's not a very strong one, whatsoever. The patients were not blinded. Okay. Well, with that impact, generalizability. Let's think about this. This study was really just to, it was between two procedures, right? It was just between this, you know, emergency, poor-ocabled, shunt and the, you know, square of therapy. So, you know, both, it wasn't like there was a placebo involved. So the fact that the patients weren't blinded isn't as impactful in an objective test or study versus a subject, right? Now, it feels a subjective, where you had subjective answers, then, yes, the patient's not being blinded to what they were doing or something, if it was a subject to response, then it would be a very impactful, but the fact is this was more of a, you know, more of a procedural study, not as impactful. Now, if the examiner, just as a side note, if the examiner was not blinded, say the examiner to kind of knew, what kind of bias was that? We'd say that would be a confirmation bias, right? That's if the examiner wasn't blinded. And then what we call it if they, you know, if you're taking, like, a survey to a certain group of people or if it's like, it's a survey at an election poll where there's certain types of people there, you know, that would be more of a selection bias. But we're not going to get too kind of hung up on this gap because the patients were not blinded when it becomes like when it's an objective. Essentially, it's a procedural thing. It's not as impactful. So, I'm going to eliminate that one. And then unmeasured confounders were not controlled by the study. When I skim through this, I really didn't see too much that was explaining confounders or anything like that. Something that maybe relates to the outcome, you have to kind of tease it out. I just

didn't get that when I skim through it. So that's kind of a less than maybe. So the top two of my choices that I go is between the allocation was concealed or the EPCS is only available as a special, specialty centers. And if I was to choose one of the between these two, the one that's a little more specific is the fact that this thing was got a specialty center, which, again, is probably a certain type of person there. Perhaps, there are the only ones that could, you know, forward certain type insurance. So it's more precise. And there's not enough information to really get me excited about the allocation piece. So the only choice, the best choice, is going to be answer choice B. But go back to our strategy here, okay? I said, look, for these statistics, drug ads, you can't get, they ask a lot of correlation. It's things are statistically significant. Now, there's different formulas for this, but this one's going to work for us, okay? This doesn't deal with all that mean, standard deviation and kind of stuff. It's just going to do two numbers and see how they, how they correlate. If the number spits out is 1.96 or greater, then we're going to say, yes, okay, because that means P value is going to be small. So we like that number, we don't care about any of these other guys, okay? If the most part I'm out here about any of these, all I want you to know is basically 1.96 or even even think I just said number two, okay? If the number comes out to a greater, it means it's significantly different. There's a significant set. So then you're going to scan questions, only you the math problems exclude the bad choices, so like only what you can prove. And again, that worked for this one. I mean, worst case scenario, if you would just held with the do the math problems first, you could have easily gotten two of these correct and then you can worry about this one and maybe even

had some extra time, but always do the math problems first, even if that you have the best chance of actually getting those correct. So now when it comes to, there's other one. Now here's a drug add and again, this was taken straight from the US and the Lee practice problem website. And so you can see it's a lot. Again, we're not getting buried down in this. We're just going to scan it, look at the charge real fast, and then go straight to the questions. All right? What's our strategy? Are there either of these math problems? Okay, when I look at this one, it says which of the following is the most appropriate response from the patient's request, okay? That's not a math problem, so I'm going to put it on pause. This one says which of the following interpretations can be made correctly from the graph. Whenever there's a graph, to me, that's going to be a math problem because it's something I can pinpoint. I don't have to read the entire thing and get bogged down. I can go straight to something and see if I can get the answer. Okay, so which of the following interpretations can be made correctly from the graph on blood pressure reduction in the advertisement? Okay. And see if we can fit them both both right, they are. Hopefully that's on the screen for you. This says I'm going to try it like this. Try this. Sorry about the delay. See if I can get it just so we can see these things. Okay. Which of the following interpretations can be made correctly from the graph on blood pressure reduction in the advertisement? Is it a blood pressure reduction from the three doses of S-Sepro cannot be compared to the reduction with placebo because the number of patients on the active drugs are higher than the number of patients on placebo. And so when I look at this, I'm like, okay, what are they trying to say here? That the NN for placebo is 159 N for the 1mg over 400. So all these

were over 400. The dose of this drug was 1mg 2.5 and 5. And there's your systolic and diastolic. So we're just reading this graph. Graph says how many points less did it go down when you took placebo? Of course, when you're, I guess it was seated or something. It went down systolic, almost, you know, a little over three, diastolic over four. But when you took this medication, systolic went down nine and went down diastolic 10 on that dose and so on. So can you infer or can you, from this that the blood pressure reduction of the three doses cannot be compared? You know, I don't know. So they're saying is it significant or not? It's hard to, on that one it's really hard to say. So for me, when I look at that, I'm like, all right, that's just the system. There's nothing, there's no correlation between those. I'm just going to put a little, you know, kind of a question mark right there with that one. Because maybe there's a better answer that I can prove later on. Is it B doubling the highest dose of the medication? Will decrease diastolic pressure from baseline by at least 15 points? So when you, at the, I'm saying, doubling the highest dose. So, you know, the highest dose is, is five milligrams. You know, we can't really say that it's going to decrease it. We don't know. But if you're even to double the dose at the five milligrams, it's only going to go down, you know, the systolic is only going to go down by 11. And the diastolic is at their, at their keen. So I really can't say that even doubling it or even at the highest dose is going to get us, it's going to get us to this 15 mark. So I just, you know, I just don't like that one whatsoever because I can't prove it. Is it see the highest dose of the seprose should be used because it offers the greatest benefit? Now, from the graph, because remember, it's saying, look at the graph and determine this. Well, we can't

really, you know, we really can't say that. That it gives, that it gives the greatest, you know, because these are actually, and we look at these, these are pretty tightly correlated. So we can't say that it offers the greatest benefit because we also don't know what side effects we can't infer that from, from this. So I really can't say that one because it's not like it's overwhelming giving us the greatest benefit at the highest dose. And, is it D? There is no clinically important difference in blood pressure reduction between the three groups. Okay. Well, how do I know that a systolic of say a decrease in nine is significantly different than a systolic say of 11. Okay. Or let me see, let's just do the, you know, the diastolic. And this one, diastolic is 11. And we can say the diastolic on this one, you know, one of these is 12. Okay. So 11 and 12. That's my two numbers, right? Now, what do we say? What do we talk about that formula? Z, right? How do I tell if there's, if something is statistically significant or not, the difference? So we say Z and members A, I'm sorry. A minus B over the square root of A plus B. Now, we've got our two numbers, right? It's 11 and 12. And can we, can we be 10 and 11? I mean 10 and 12, doesn't matter. If it's any simpler, we could, we could do 10 and 12. Just to give you more of a spread. Okay, say it was there, well, if we're comparing diastolic, we can go 10 and 12. So 12 minus 10 or square root of 22. And so I get that. And then the square root of 22, I was to try to look on a calculator here and try to get the square root of 22. If I don't know it, then I would just go to what's my next thing that I can do is for, you know, five would be 25. And so it's like this, if I was, I ball that, I know you got to calculate it on your exam, but if I go to square root of 25, the five, the square root of six, of 16 and four. And so the

square root of 22 is going to be up there, closer to five. So I'd give it like a 4.75 or even a five. The nevertheless, 2 divided by five, it's going to be 0.4. Now what did we say? We said it's got to be at least what? At least 1.96 or greater to be significance. So there is no clinically important difference in blood pressure reduction between the three groups, right? Because even at their worst extremes, between, you know, say a 10 and a 12, you're still getting an insignificant difference. It's not, it doesn't overcome this. So right now, I'm liking that one because it's because I can prove that. I can prove that there was no difference between that. And then E says the significance of the drug effect versus placebo cannot be determined comes as low p value. Well, where's this whole p value thing at? So when I read through this chart, because that's where they tell me to focus, there's a p value buried in here. You probably can't see it, but the p value there says it's greater than 0.5. So we should know from basics statistics that you need a p value less than 0.05 to be statistically significant. So the significance of the drug versus placebo cannot be determined because of the p value. Well, there's a high p value, not a low one. So we're going to eliminate that one. So that if all these answer choices, and we went straight to the math problem, the only one that I can prove on this is going to be answer choice D. Now that was, it's same like a lot. But the key is on this, you gotta keep doing these problems to where you start working it. And you can start actually using this statistic rule, you know, this whole Z thing. And you're, honestly, I didn't use this too too much until I started doing all this kind of stuff and trying to learn this strategy because this is one that's going to help us through these types of problems. You solve that. You look for this

number or greater. And if it is, then there is a correlation. Okay, if it's less and there's not. So back to the last problem that it had there and just by reading it, it says, which of the following is the most appropriate response to the patient's request for the medication. So the answer choices are, without even reading the rest of this, express how should be prescribed because she can get it free. Okay. Well, let's do this. Let's figure out about this later. First, she's 65 comes for blood pressure medication. History is poorly controlled. She has hypertension, psoriasis, psoria, arthritis. Previously treated with methotrexate. Alcohol use disorder, elevated liver enzymes, treating on multiple medications, vital signs of this. She's got this physical, this closes thick, scaly, plaques on the scalp. So she's got the whole psoriasis thing going all over the place. This severe spider in GMI, now when you, when you think of that, what do you think about it? It goes back to this whole liver thing, right? She's probably got some chronic stuff going on. abdomen distended and a fluid wave is noted. All right, so she's retaining a lot of fluid, liver is probably not functioning too, too you're here. Two plus edema. She'd like to try a new drug called this thing. Okay. So, we start looking through the answer choices, which of the following is the most appropriate? It's, that's appropriate for prescribed because she can get it for free. Come on, we all know that that's probably not going to be a good answer choice. Should not be prescribed because it's a worse or her psoriasis? Well, we don't know what the reason to think. Should not be prescribed as similar to other medications. We don't know. Should not be prescribed because the patient has severe liver disease. Okay. It's always something to worry about. Especially, should only be used for hypertensive emergencies. All

right. So, to answer this one, we have to actually, you know, we actually have to scan it. So, from here, we can see it's just a blood pressure. We've already looked at this one. There's meaningful reductions. There's 1,000,000,000, 2,500,500, the number. It's significant reductions in heart rate. Okay. But we're looking for this whole thing in psoriasis. And this thing about the liver, we're the two dealbreakers in it. So, it's an incident with a women. Oh, my minorities appears. It's a beta-adternetic blocking edge. So, it's a beta blocker. So, keep that in mind. But if we went back, it doesn't look like she was on a beta blocker. She was on an asian inhibitor, um, sparing a lactone. Okay. So, she wasn't on another asian, find another beta blocker. And then here, it talks about adverse events. But that really wasn't part of any that question thing that we talked about. And then, most side effects were mild, the nerve-cars discontinuation. No statement. We're observed. You know, drugs that inhibit, um, drugs that inhibit 2d6. Know this for all your step exams, all these inhibitors. I mean, I want to stop in the middle of this. I wanted to do it. But when in doubt, if you don't know inhibitors, always go with 2d6. I'm really, I'm sorry, 2d6, really, really common. But in this one, it's a 2c9. Okay. And then down here, important safety information. Patients treated with this medication. I've got sudden discontinuation. It's counterinucated in Brady Cardio Heart Block, first degree. And severe herpatic impairment. Okay. Right? She had the fluid wave, all that kind of good stuff. She would be using diabetics as beta blockers may, some manifestations I've been listening to. Okay. All right. So, back to our best answer to choices. It said nothing about gizlaking worsensuriasis that I read. She never describes it. It has similar to other medications. It's not on any other

beta blockers, so I'm not too worried about that one. Especially, should not be prescribed patients severe liver disease. Well, it did mention that. Okay. And it's spresso, should only be used for hypertensive emergencies. I didn't really talk about that. So the best choice. The only one that I can prove is going to be this liver disease. Now, again, if you don't read the questions first, I mean, you're going to waste a minute. It means problems are long, and that's why I don't like them. But there's no sense, at least in my opinion, to read all this stuff, unless you're really zeroed in on what you're trying to focus in on. So, always read the questions first. And what are you going to do? You're going to look for the math problems. You're going to exclude all your bad choices, and you're only going to select the ones that you can prove, because sometimes everything sounds good, but you're only going to do the ones you can prove. And you're going to know this formula right here. Now, I think there was one other question ahead for you guys. And here it is right here. Last one, okay. When you see something like this, the reason I bring this in, because these drug ads, it's all I've telling you most of the time, it's all about the box. It's all about what's in here, okay? Now, when you see confidence interval, the one thing you want me to look for is when it crosses one, I want you to think that it's no good, right? Because that means that there was really no difference between treatment and placebo, right? At some point it was equal. That's what the one means if it crosses that one in the conference interval, right? So for me, without even reading this problem, I'd go down here to switch to the data as correct, okay, while I'm looking at this, I don't like any of these problems, automatic eliminate the ones where the confidence interval crosses zero, because that data

is irrelevant to me, I can't use it, okay? Yeah, this one seems doesn't cross one, so catches my eye. What about p values? I only care about p values that are what less than 0.05. So offhand, you know, this is a long problem, I'm zeroed in on this one right here, which is talking about this. So look at the question. Based on these data, which of the following is the most appropriate conclusion regarding the five potential risk factors for the transmission of HCV? All factors appear to have effective to have an effect. You can't really say that, right? Some of the information was garbage. Feet of electric monitoring was protective. You know, again, a cross here, I can't, I can't, I don't like any of the data what talks about that. Infant HIV infection was the only significant associate or factor, okay, that, and it looks like compared to all this data, that looks like it's a possibility. The terminal, criminalized was, significantly, most significantly, no, again, bad data, okay. So there, and then no factor was significantly associated, no, can't say that, because there was one that was associated. It's going to be the HIV, and it's usually C, okay. So the point of using this one, and again, all these problems I pulled were from that USMLD.org. This is a stuff they want you to know, but if you can attack these problems just by ruling out the garbage stuff. But I'd say, again, it's going to go back to this scan. Look for the math problems first. Exclude your bad choices. It only select the ones that you can prove, okay. Without a doubt, you can save from this information from this data. I know that this one would be a good choice, because I can prove the result. And then know this formula because it's going to save us when it comes to comparing, you know, if I got to compare two numbers, I can see it there actually statistically, a significant or not. And that's going

to come important, especially when we're trying to compare graphs and stuff like that. Okay. So guys, I hope this was helpful. I mean, it was a bit long, a lot longer than I was hoping. But again, if you just supply these to the drug ads, I think we can kind of get through this. And again, focus on the math problems, so the ones we're going to definitely get right. Alright, guys, the only questions make sure you send some problems to me with what we can solve. Otherwise, we'll see you in the next video.