Randy Neil Biostats 07: USMLE Step 1 - Cumulative Probability with Question

Multiply the probability of each sequential event.

Convert each death rate to survival first: survival = 1 − death .

Safety.

Postmarketing surveillance for rare or long-term adverse effects.

Answer: B) 0.72

Cumulative survival = 0.90 × 0.80 = 0.72.

Answer: C) 80%

Survival = 1 − death = 1 − 0.20 = 0.80.

Alright guys so this video is about the cumulative probability. This was sent in by one of the viewers per se from the comment section. So thank you for sending that in. It's a good question. I think it came up and I appreciate you sending those in. I think it helps everybody when we can see those questions that we don't really cover kind of classically throughout the basic stuff. So make sure you keep sending those. It really helps and I enjoy doing it. And watch this video up until the end. Actually you guys will have a homework assignment that will actually impact one of the upcoming videos and the questions that we have with that. So watch it. Watch all the way to the end and again it will be like a tip to watch some other video that it's really important. So hope you like it. Alright guys so this is going to be about the cumulative probability and how I want you to understand this. And in case you see that word on the test came up in a question. I don't want you to think of that as combined probability same thing. And to really understand the concept you got to think of it like a dice. And how you're going to do this is to think if I was to roll a dice and I say what's the chance that you're getting it to? You'd say oh that's one out of six. Right? Because each time it doesn't matter what you did the first time because when you roll it again you start back and scratch. So and then if in so one if I said okay if you're going to roll it two three times in a row what's the probability of you doing that and be one six times one six times one six and you multiply that. Okay you multiply across and that would give you the probability of rolling a rolling three two's in a row. So that's the cumulative probability of doing something. So they started using these like when they started doing drug research perhaps even drug ads but they'll talk about the cumulative

probability of lasting or of survival after some of it. So here's the question. It says a study of breast cancer patients who use drug R.I.Y.A.8 is reviewed and the data is presented in the table below and you can see the data here says time in years. Okay one year one and two three to four five years. The number of patients at the beginning of that time period so they started out at 250 patients and you can see how each year it went down some. The number of patients who passed away during that time so in the first year at the end of the first year. 50 people died and then 15, 15, 10 and so on and then this last column is the percentage of patients who the percentage of patients who passed away during that time frame. So after the in the first year there's you know 20% of the people passed away in the year one through two seven point five percent and so on okay. So it says based on this data what is the probability that a patient who uses a drug will be alive. Okay, will be alive at the end of four years. Now this right here says the percentage of people who passed away. I want the people who will be alive at the end of four years. So I got to get all the way to here. I got to go from zero all the way to four and I got to figure out now it'd be nice that they said you know someone did all the calculations for us but they don't. So here in on this question they give each individual year now I just have to get to year four because that's what they're asking. But the kicker to this one is they give me this the percentage of people who passed I need to people who are alive. Because think about this if I was going to say what's the chance of someone passing away if I asked you right now what's the percentage percent chance of someone passing away using this drug at the end of one year. That's easy that's 20 percent. And if I say what happened at the end of two years well

you could say at between year one and two seven point five percent people passed but I you'd have to do the combined you know with someone last two years on that drug you'd have to say oh well chances of them passing away would be you know this times this. Now there's one disclaimer and I'll say this in math I don't like percent I like decimals so you got to convert all this stuff to decimals. I don't know in your step exam make sure everything's a decimal. So again if I was to if I went back to that whole dice thing if I said again if I'm at the roll of two one out of six one out of six so in year one if someone's going to pass away was the chance. Point two and then a year or two how it was the chance of them passing away point oh seven five okay. But so you can see that you just got a multiply the probability of each year up and then that'll give you the the sum total of whatever length of time you're looking for so to answer this question it says. So on this data what is the probability that a patient who uses the trouble be alive so I just got to convert this so this isn't. 20 percent that because that's who passed a percent who's alive is 80 percent or. Point eight this is the percent who are who passed I need the percent who are alive and so that's going to give me so point eight it's going to be 93 right. Well 93 percent or point nine or point oh nine three. All right I'm sorry point nine three and then this one is the percent who passed so that means the percent who are alive is 91 percent. You see what they did 80 percent and years zero through one 80 percent are living years one through two it's ninety ninety three percent are living and years three to four ninety one percent are living. So we have to multiply these up and again we don't like percent we like. Desk. So I would have point eight times. Point nine three times. Point nine one and if you

multiply all that up that would give you your. Probability that a patient uses this drug will be alive. At the end of four years you just have to multiply it up it's just like the dice you got a multiply multiply multiply and. And each event is independent of the other okay now again it would have been nice that they would just gave us the the probability themselves but again this is the step exams. They're going to make you interpret the chart you have to know your take home point of this thing is that when we say. cumulative. Probability. Just start thinking you're going to have to multiply each independent event they've got to give you something it be nice that they gave you the role of the dice right you'd nail that in a heartbeat. But it's probably going to be in a chart. Be careful is that if they say a lie down here and it says passed way up here you got to convert it you just got to take the the opposite of a hundred percent. And then subtract these and then you you'll have it okay. If I would have said passed away at the end of four years then you'd just again you would have to just went. Point two times point zero seven five times point zero two nine and it would have probably been. Well it would have been this guy but it would have been the decimals. Okay. Now what I want you to do. Did as an extra one for this there's two things is kind of a homework assignment for you. I want you to understand clinical trials number one. There's four phases right. So there's phase one that basically says is it safe you know are we going to do this so it's a very limited amount of people that give the drug. Make sure no one is going to die all of a sudden. So no one on so phase one basically answers the question is it safe phase two. Does it work right we know people can tolerate it but does it actually work. Phase three says does it work. Better okay doesn't work better

and then phase four. Just gonna memorize this this is the post. Market and the face surveillance right they put it out there and then they get feedback from it. So phase one is it safe phase two does it work phase three does it work better. And then of course phase four post market. Now here's your homework I want you to look up this person on YouTube. And it's called easy ways. To remember. DNA viruses okay. You're gonna watch this video you're gonna have it memorized and at some point I'm going to make some questions about this. Now this isn't really the bias stats but this is important for the step exams. The YouTube person I don't know this person but they're called a hundred lyric and you'll see a picture. And it's called a boy with the hat. And hopefully you guys are familiar with this because it's a great easy way to memorize the DNA viruses. Okay it'll help you and it's a very simple way it's very effective. I don't want to take credit by no means what this person that worked they put in on this. But what we'll do is we'll create some questions coming out for this. So the take on point of this video. You want to probability make sure you multiply when you see it. Just make sure they don't trick you. And then understand the phase trials and then make sure you look at this guy. Look at that video and then we'll do some questions later. Hope this was helpful.