hey everyone this is Ryan and in this video we'll be talking about disease frequency so in the last video we talked about the accuracy of screening and diagnosis methods and how reliability and validity were the two main measurements that we used to calculate the accuracy and it's just important to note that reliability and validity are measurements that are important for all studies all good studies should report the reliability and validity of the methods that they use so disease frequency has three major study methods associated with it cross-sectional surveys are concerned with prevalence of disease in the population prospective cohort studies are concerned with the incidence of disease in a population and case studies are more concerned with the rare conditions in an individual so cross-sectional surveys take a random sample from a sampling frame and the sampling frame is just some source of demographic or population information that gives us information about the population that were concerned about that we want to find more information about so cross-sectional surveys get their name from B from taking a cross section in time which just means a snapshot or looking at the present time and how much of a certain population has disease so we can find this out through questionnaires or interviews and the target population refers to the whole population that we want to find more information about the target population could be say the entire US population and it would be impossible to get everyone in the population to answer the questionnaires or answer interviews so instead we take a sample from that target population so this would be the sample population which we can ask questionnaires too and hopefully this will be a good representation of the entire target population so the case definition just refers to some some categorization of what the disease actually is and what it isn't so the cases would be people who have disease and Nan cases would be people who don't have disease so prevalence is the frequency of disease for the population in the present so again that that sample population is going to provide a best estimate for the larger target population and prevalence can be calculated by taking the amount of diseased people in that population and dividing by the total number of people in that population and this can be reported as a percentage or as a decimal as in ten percent of the study population has caries in the in the present and just important to know that the prevalence prevalence is something that refers to populations where as disease is something that refers to people or individuals now severity I just think of as a improved prevalence instead of just saying yes or no you have caries or you don't you look at an average value of some diseased index of some disease index which could be a mean or a threshold and one such index is the D MFT or decayed missing filled teeth index you can also use the lowercase letters for primary teeth or the S means just refers to surfaces any of the five surfaces of a tooth and and these are all a little more clinically meaningful than just a simple yes or no do you have or do you not have caries answer so the formula for calculating severity are accomplished we don't need to know it what it can look like as a reported value is say 1.8 teeth with carries per child and we can extrapolate that out and say 18 affected teeth per 10 children or 180 affected teeth per 100 children whatever is most meaningful as a as an outcome that can be applied in the clinic so anytime we take a random sample from the target population and make an estimate about that target population we have sampling variability or sampling error this is just the nature of taking estimates and we'll see this for each of the studies because we can't ask the entire target population we can only be say 95% confident of the answer that we got so the 95% confidence interval is a likely range of the true or target population estimate so this is usually some sort of curve and we have a some mean value and then some standard deviations away from that main value that give us that likely range now the brother or sister to 95% confidence intervals are the P the p-value or the type 1 error is interesting to note that we usually use 95% for a confidence interval and we usually use point zero five as the threshold for p-value and 0.95 added two point zero five is equal to one so they lead to the same conclusion in this case of p-value you want a low value if you're two you have point zero two then there's just a two percent probability that the difference in results the statistic that's a statistically significant difference is due to chance so you want that number to be low because you want your results to be meaningful you don't want them to just be because of chance that you got a you got a cool result you want to be you want that number to be low if you increase the sample population you're getting more of that target population so likely you'll shrink the confidence interval you'll shrink these standard deviations and you'll get a smaller more a smaller curve a shrunken curve horizontally so you'll be more confident that you're at the range of your estimate it is smaller so with cross-sectional surveys you have convenience sampling which is also referred to as consecutive patients which just refers to non random sampling maybe you just talk to people that were easy to talk to and this can result in obvious bias incomplete data can also can also occur and this is from non-response or differing rates of response between relevant groups so if you're taking the confidence interval from one study and then comparing that with the confidence interval from another study say one was done for males and one was done for females and there is a statistically significant result because these confidence intervals did not overlap you'd want those the data from each group to be equal you'd want to have similar similar sizes for the sample population so prospective cohort studies are a little bit better than the the cross-sectional surveys we just talked about and we'll also see prospective cohort studies in the next video on etiology where they're really useful so prospective cohort studies take a random sample from a healthy at-risk population and this is slightly different from the cross-sectional surveys where we just asked a population we didn't care if they were healthy or diseased we wanted to actually find out about how many were diseased so this time we just we specifically asked people who we know are healthy and we know they're at risk or they're they're not immune to getting disease to getting this disease so we we take one group we don't separate people we just look at one random sample and we take a baseline or initial assessment of these disease-free patients or their teeth and then we wait some amount of time and then we follow up with them we take a second assessment to see if anyone got got the disease if there are any new events of disease and this is known as incidence the rate of developing new disease over time so the incidence rate calculation is similar to prevalence that we sort of take the disease the number of diseased and divide that by the total except this time we we just are concerned with the new disease developed in a given time how many of those people in the at-risk population got the disease so this could be anywhere from zero to a hundred percent 100% would mean everyone you ask got the disease so the disease which would be really bad and this can also be reported as a decimal and this is sort of talking about the risk or probability of the average person developing this disease and again when we're taking an estimate about the target population we have sampling variability so we'll see these two calculations and then case studies there there's no statistical data here I was just interesting findings some rare unknown case or condition or something we don't know a whole lot about a doctor comes across a cool case decides to report a case study all right so the next video we'll continue our discussion talking about etiology as the third big thing that were concerned with so I hope this video is helpful and I'll see you next time