 very good morning, welcome back to my classes. We will be continuing a session on Epidemiology. So today we have a small topic that is measurement in Epidemiology, how do you measure things in Epidemiology. So let us see what are the tools of measurement. So we had covered Epidemiology in detail, all the study designs that is descriptive, analytical and experimental. So once we start the study that is Epidemiology, how do you measure it. So we have a lot of things coming across in Epidemiology that is mortality, which means death, mobilities, means like cases, disability, disease attributes, a lot of things. So today we will be seeing the mobility measures, that is what really important as an epidemiologist. Remaining are also important but we will be seeing mobility measures. Mobility means mostly the cases, mortality is the death. So tools of measurement. So tools we know every professional has tool, if we go to a doctor he has tooled as BP apparatus thermometer. So using this he measure our blood pressure and temperature. Then we will come to a diagnosis. So likewise an epidemiologist is also having tools to measure the mortality or mobility whatever that is. So the basic tools are proportion rate and ratio. So the proportion rate and ratio, so the proportion is nothing but percentage. It is like we are calculating the number of people in a group of people and multiplying it with 100. So there will be a numerator, which is a part of denominator because the cases will be from within the population. So numerator and denominator are connected and there will be a multiplier of 100 and there will not be any time factor. Why it is important? We will come to know once we see the rate. So if it is a yellow circle among the total means 1 by 3 here it is 2 by 6. So we just multiply it with 100. So here is an example which I used to show in my lectures. So what proportion of this class are multi-fans? So the multi-fans divided by the total class into 100 will get the proportion. So similar the Mohanlal fans, so the total number of fans divided by total number of students in the class into 100 will give the proportion of that particular thing. Okay. So we will come to the real life example. What proportion of the population is suffering from diabetes? We get the data of diabetes patients and divide from total population will get the proportion of diabetes patients. So second one we are seeing is rate. Here the time factor comes. Okay. So the time factor is only comes in rate. So we have seen cricket matches. So usually we have seen over rate, run rate. So how much runs is code per hour? How much hours is ball per hour? So all these are rate. So always there will be a time factor. So numerator is part of denominator. Definitely. Obviously it will be a part of denominator and there will be a multiplier. Usually we multiply rate with 1000 and proportion with 100. We can do it with 100,000. It doesn't matter. But usually we do it with 1000. It has a time dimension. This is the most important thing. Rate is always expressed in time dimension. We will get to have a better idea once we see the examples. Where proportion is just the percentage of people affected with something among the total population. There is no time factor involved. So rate is we commonly say death rate over rate. So death rate how do we calculate death rate is number of deaths in one year by mid-year population and 1000. So this mid-year population we usually take in statistics the mid-year population the population which is present on the July 1st that is a mid-year because we take 6 months to June and 6 months July to December. Probably July 1st will be the mid-year population. So the population which is present on the mid-year day will be taken as mid-year population. So here numerator denominator and there will be a time factor. So the run rate is another example. So rate involves time dimension. The last one is ratio. Ratio is like male to female ratio, husband to wife ratio, doctors to population ratio. Here the striking feature is numerator and denominator are two different entities. Proportion and rate is expressing the same factor but in ratio the numerator and denominator are two different things. Males and females, doctors and patients, husbands and wives, students and teachers. They are two different entities. So numerator is not a part of denominator. So suppose sex ratio that is male to female. In Kerala the male to female ratio is 1000 to 1084. So this male and female are two different entities unlike the rate and proportion. The doctor population ratio there is one doctor for every 7500 patients. So these two are two different things. So numerator is not a part of or not a component of denominator. It is two random quantities. That is rate. That is ratio. So one teacher, five children, male to female ratio. This is doctor population ratio. This one is to 145. This is percentage. That is doctors shortfall at PSE level. This is 27 percentage of doctor is not present in primary health centers in UP. 34 percentage is not present. That will divide number of people who are supposed to be there and divided by total number who are posted. So these are proportion and this is ratio. This to this PHC, this proportion and rate. This is rate. Okay infant mortality rate. This is per thousand live birth for one year. So that period is there. Time factor is there that one year. That is denominator and numerator are part of the same thing. Here also numerator and denominator part of the same thing. But here it is the ratio. Doctor and population is two different entities. That is one is to 495 in Goa and in Kerala it is one is to 8. So this is a summary of two. So that is ratio, proportion and rate. This is rate. It is 11,000 live birth in one year. So here time factor is there in these two cases. Time factor is not. So summary, we have covered three things. Propulsion rate and ratio. Okay. So now let's go to the principles of epra-monosy. So there are four principles basically. One is the exact observation we need to strict, rigorous and accurately precisely take the observation and it should be free from error. That means correct interpretation and there should be a scientific reasonable and intelligent explanation and the construction also should be based on knowledge and technical skill. Okay. So I made an acronym. Every coffee requires sugar. That is E, C, R, S. So exact observation, correct interpretation, rational explanation and scientific construction. Okay. So this is principles of epra-monosy. In the beginning, we learned about rules of epra-monosy. Motality, I told you we are not going into detail. That is death related stuff. So we are going directly into the morbidity measures. Okay. So this epra-monosy is very vast like emotion. So only what we need to learn is based on our objective. If it is based on our exam purpose or research purpose, whatever it is, it should be based on this. Okay. So we have morbidity measures. Basically, two morbidity measures are there. Commonly used one is incidence and one is prevalence. So incidence and prevalence we already talked in case control and cohort study. In case control study, we get prevalence and in cohort study, we get incidence. So since cohort study is going in future or forward or prospective looking, this is finding new cases. Okay. So which happens in the future time. So because cohort study starts without any disease. So in future or a period of time, they may develop case. So that becomes incidence. That is occurrence of new cases, incidence and existence of new and old case that is prevalence. So incidence is always a rate. So you should mention time factor also. One week, this one in your case, so one month or one year, it depends on the time frame, but prevalence is just the proportion of people. So if in case control study, we just take the number of cases divided by the total population, we get the prevalence that includes new and old cases. But incidence, it is going future. So there will not be any old case, only new cases will be there. Privilege we are checking the background information or previous information. Until today, how many cases are present? Until today, how many cases are present? So it includes new and old cases. Okay. So today cases and yesterday's cases of previous month, previous year, everything comes in prevalence. So basically, you can say that incidence, how many people with the disease are newly diagnosed each year, which is like a video throughout the year or the follow up study or a prospective study. It goes throughout the year. Whereas the prevalence, how many people in a population currently have the disease at present? How many of them having disease? That is just like a snapshot or a picture. Okay. So that is the basic difference about incidence and prevalence. So incidence is a formula of new cases of a disease in a particular time period divided by total population at risk at the same period into 1000. Okay. So incidence, as I mentioned earlier, is a rate commonly expressed in 1000 multiplied 1000, whereas proportion or prevalence we multiply with 100. So incidence is the number of new cases. And since it is a rate, there should be always a time period. So just check an example on January 1st, 2016, there are 10 people and while reaching on January 31st, the patient out of 10 people, three people became deceased. So it is 3 by 10 over a period of one year. So 0.3 per one year. So we can calculate it by 3 by 10, that is three cases by 10 over a period of year. So this time frame is very important. Okay. So this is, we are starting without any disease, as I mentioned in the cohort study design. Okay. Over one year period, only three people got disease. So we have to mention the time frame. Okay. Incidents which will not be there in prevalence. So there are two types of incidents, that is incidence rate or incidence density or cumulative density or incidence proportion. Okay. So incidence proportion is, it will be like percentage, there will not be any time frame. But incidence rate is the true incidence or incidence density. So you won't get confused with incidence proportion which is coming in percentage or proportion percentage. Okay. So incidence rate is the actual incidence. So just an example, it's a basic thing, cumulative incidence rate. A numerator will be definitely cases, but the denominator will be different thing. In cumulative incidence, they'll take the initial population, but incidence rate, they take the person time here. I'll explain in detail. So incidence rate goes from 0 to infinity. But whereas cumulative incidence goes from 0 to 1, because it is percentage. So maximum value is percentage is 100. So the maximum value will be 1 for cumulative incidence, but incidence rate will go to infinity. It is also known as incidence density or proportion probability. You won't get confused with this. So let's take an example where for 12 people are being followed up for 14 years. So the 12 people, 12 people are being followed up for 14 years starting from 1980 to 1994. So the first person entered the study at the year one and he was followed up for eight years. But after that, he might have left the study. Okay. So he was under risk. Okay. They all were population under risk or to develop a particular disease. All were having habit of smoking to expect outcome of lung cancer. Okay. So this person was so for eight years, all eight years, he was under risk. So the time at risk became eight. So this is known as eight person year. So in incidence, risk or incidence density, we calculate person year. That is a time frame. We calculate in this format that is population under risk. Okay. For the time frame. So this has become eight person year. The second person entered the study in the beginning and he was followed up for 10 years. But at the year of 10, at the 10th year, he developed lung cancer. So the person under risk was 10 years. Okay. So once it's developed disease, there was no risk. It became a disease. So this is 10 person year. And the third and fourth person, it was they entered the study at the beginning and they were followed up for all 14 years and they having that risk, but they never developed disease. So this became 14 person years each. This person died at four year, four person year and this person entered the study in 1981. Okay. Then they were followed up for 12 years. Actually, this is 12 not 10, 1981 to 1993, it will become 12. Okay. So that's a mistake. And second and the remaining all person entered the study in 1981. So all this time person here will calculate and will divide from the total number of case. Okay. So the 14 year, the three cases have been reported. Okay. The second person, fifth person and 11 persons were the cases among all 12 people. Okay. So we are not taking the number of people. What we're taking is person year, the population under risk, the duration of population and risk. So this 12 people were under risk for 100 person years. How we get 100 person years is different. Okay. So this is how we calculate incidence, risk or incidence density, we have to calculate person year. So this is how we calculate person year from the beginning of study until they leave or study until they develop the study, we calculate the duration. Okay. So that's how it is 8 and 10, 14, 14. Hope you're clear about this person year and calculating incidence, risk or incidence density. So the one more example, here the person is getting disease at second year. So become only one person year. Here also one person year because he left the study here. He became disease at 1991. So two person year here he left at one second year. So one person year. So this is three person year. This is five, six, five. He became deceased here. Okay. This is two actually. This is one. This is also one. Okay. So here it's a slide. It's clear. Okay. So only four people are became four people are becoming diseases. One, two, three and four. So four cases per person year we have to calculate. So this is one, one, three, one, three, five, six, five, one and one. So total 26 person years. So four by four cases by 26 person years. So four by 26 is 0.15 or 15 by 100 person years or 0.15 person years. So this is how we calculate person years. So this is incidence, risk or incidence density. Okay. So this question mark is the person is lost to follow. As I mentioned you about the attribution factor in cohort study or follow-up study. So you might have left out the study. Okay. And this is case. So once you become a case after that we won't calculate the risk because the risk was for having the disease. The risk was too for developing the disease. So he is having only one year for the risk. Second year develop the case. So the next one we've seen now is incidence, risk or incidence density. Now it's a cumulative incident. That is, I told you this is a percentage. Just an example, it's very easy. So in 2001, there were 5,572 women aged in 20 to 30 years who were sex workers based on the record of whatever 45 were HIV postage during this three-year period or four-year period. They have, it should be 2001 to 2005. So what is the cumulative incidence of HIV postage during this four years? Okay. This percentage again comes in prevalence, but the problem is prevalence. We don't mention about four years. Privilege based, we just calculate 45. That is new cases. We will calculate the total cases that will come beyond 2002. There is no time frame, but we can calculate time frame that is period prevalence, that is different thing, but that will include all the new and old case. Okay. So this is cumulative incidence. So 45 new cases by total population, that is 5572 divided will come 0.8 percentage. Okay. So incidence rate, the denominator, you can see person here. The time frame is present, but in cumulative incidence, there is no person here. It is just percentage. It is almost like prevalence because prevalence is a proportion. This is a proportion, but here we have a time frame. Okay. Time frame is important. That's why it is different from prevalence. So there are a few common example that rape, a number of reported rape cases per one lakh women in 2014 and 2013. So in Delhi, we can see this is 1813 in 2014, whereas 141 in 2013. So this is rate per one lakh women. Okay. So it's a common example. So what about incidence? It is referring only new cases and it is not influenced by duration of disease. That is like if a disease happened 10 years back or five years back, it is coming. So can we ask being only new disease, but the time frame is different. So it is always refers to particular time period and denominator is people at risk. We had seen an incidence risk of the time person here will be calculated. Next is a prevalence. It is proportion. So it is just like the old and new cases or a particular period of time. There is no time frame. It is a time frame is a pre period prevalence, but it is checking all the new cases. The true prevalence is only one point of time and total population into total population at risk into 100. So there are two types of prevalences. One is point prevalence and one is period prevalence. So just take an example on Jan 1st, 2016, there are two people are deceased. In Jan 31st, there are five people are disease category. So in Jan 1st, if we take prevalence that is 2 by 10 that is 1 by 5 or 0.2 or 20% here it will be 50%. So it is a number of cases that is 2 by total population that is 2 by 10. Here it is 5 by 10. So 20% and 50% will be the prevalence. But the period prevalence. So in point prevalence, we are taking only one time. Jan 1st or December 31st. Period prevalence, we will be considering the total cases. Okay. So the total cases throughout the time is same cases are written here. So again it will be 5 cases, 5 by 10. But the time frame we have to mention during one year period. Okay. It will be the same as point prevalence at December 31st. Because December 31st case, the two cases are carried over from the Jan 1st. So our spirit of time only 5 cases are present. So point prevalence and period prevalence is different. So period prevalence, this is point prevalence is as I told you 2 by 10 and 5 by 10. 20 and 15% H. So in period prevalence, we will be think the all number of cases throughout a one year period. Okay. This is not like incidents. Incidents is different. We will be assessing the time person. But here we will be taking only the number of cases and total number of population. Okay. So here 6 people are having disease and 3, 4, 5, 6 into 36. So 6 by 36, that is 0.6% H 6 by 6 into 6 years, 36, 6 by 36, 6 by 36. So number of cases are 6 and total population is 6 over the period of one year. So it started in Jan 1st and it ended in December 36. So the percentage is 16%. So the period prevalence is 16% H over one year period of time. Okay. So the point prevalence will be just on one day or period prevalence will be over a period of time. It can be one week, one month or one year or five years. Let it be any time frame. Doesn't matter. But we will be checking cases or a period of cases and the total population will be taken. So the denominator is different in prevalence. Denominaries will be taken total population. But in incidents, it is different. Incidents rate will take the time person whereas in cumulative incidents will be taking the population under risk. Okay. So here we take all cases. Some cases might be present before 2016. They are carrying over to 2016. These two might be became deceased in 2014. If they are cancer patient, they might have started the disease in 2010, 2013. Nobody knows. But they are still being cases. So we'll count this. But in incident cases, we'll just see at the beginning of study there's nobody's having disease and over a period of time, we'll be checking the incidents of cases. Okay. So in prevalence, it is not like that. If cases was present, even the status of case was present even before the start of the studies will also be counted. In chronic cases, chronic disease cases, it will always happen because the duration of this is very long. In chronic cases like hypertension, cancer, so such cases, these might be present before checking data. So this might be present on cases. They became deceased on 2013-14 and still they are deceased. So that also will be counted. Okay. So in incidents that will not be counted because we see in 2016 Jamfers, there should not be any case and we'll follow up for one year, two year or five year and we'll count the number of new cases. Okay. And the denominator will be time person here in incidence density. Okay. This is prevalence that will be percentage if it is a point prevalence or period prevalence. So prevalence is like it increases if the duration of disease. So I told you like cancer patient, it increases if the patient is having a longer natural history and a patient is being case that is if it is a chronic disease, the prevalence is also will be increases and if the treatment goes prolonged, it will increase and increase in incidence when people come from outside, it will increase and healthy people if goes from our city to outside, the denominator will go less. So the prevalence will increase. Prevalence will decreases all the cases, all the points against the shorter duration, better recovery, improved cure rate, if decreased incidence, emigration of new cases, if new cases are going out of the city and immigration, healthy people are coming into a city, all these cases prevalence decreases. So prevalence will give you the magnitude of problem. Okay. So and administrative and planning purpose, we can use it. So this is some common example, 40 percentage of Indian people are underweight population. Okay. So it will become around 30 or 40 crores. So how it came on the number of underweight divided by total 130 crore, this percentage will get. So like that, the 20 million obese women, that is 3.7% and 9.8 percentage, Indian men that is 3.7 percentage and 20 million that is 5.3 percentage. The denominator will be 130 crore. Okay. So how we calculate the prevalence? Okay. So the prevalence, let's take an example which is of incidence risk. So this is something we compared for prevalence. Okay. So here October 1st, 2004 to September 30, 2005, we are observing this downward arrow means date of onset of disease. Okay. Date of death is a positive sign and the upward arrow is recovery. If you are taking prevalence on April 1st, 2005, okay. So we have to see how many cases are present on April 1st. So 1, 2, 3, 4, 5, 6 and 7. Okay. 1, 2, 3, 4, 5, 6, 7. So 7 cases are present on April 1, that is point prevalence. Okay. So let's take total population 100. So our point prevalence on April 1 will become 7%. So if it is October 1, 1, 2, 3, 4, 5, 6. Okay. I haven't written it here. October 1, the prevalence will be 6 percentage. Okay. 6 by 100. Here it is 7 by 100. But what happened was 1 person died before April 1, 2005 and 2 person died before 2005, April 1. What happened was 3 person became ill. Okay. So that's why this change here 2 got out of the study but whereas 3 came into the study. So that is why this became 7, 7 and this became 6. Okay. On September 30, just count 1, 2, 3, 4 and 5. Because 2 people died here, new person came here. Okay. So on September 30, this is just 5 cases. 1, 2, 3, 4, 5. So 5 by 100, that is 5 percentage. Period prevalence for the period of October 1 to September 30. Okay. So how many cases? Like 1, 2, 3, 4, 5, 6, 7, 8, 9, 10. So total 10 cases were present. We are not bothering when they joined or whether they died or nothing. Whether they were there as a case will consider that is the prevalence. Okay. Whether they recovered or they died or nothing matters. So that is why incidence is very important. Prevalence is just giving a percentage. Okay. So this will give you a clear idea about prevalence. Can just see this as an onset. This is death and this is recovery. So period prevalence, you know that 10 percentage would what happened was many cases that is 1, 2, 3, 4, 5 people died during this period and 1, 2, 3, 4 people newly came into the disease. So still it is 10 cases per total population. So we kept total population as 100. So this is period prevalence. We had seen one example here. Okay. So this is just a graphical presentation and it is very precise graph to explain this point prevalence and period prevalence. So denominator will be total population. In cumulative incidence also total population, but they will consider only cases. So suppose if we take incidence of October 1 to September 13, what happens is you check the new cases 1, 2, 3, and 4, then you have to calculate the person time here. So this will be four new cases, four new cases during this period. Okay. See one new case, two, three, and four, four new cases, then we have to calculate the time person here or time month or time week. So we can calculate in any way. So that is different with the incidence. Okay. So prevalence and incidence can be expressed in this graph in this picture. Okay. So prevalence is a total case. Okay. This will include old new cases. So incidence is a new case from the tap. The new cases are being into this prevalence. Okay. So prevalence, some will be recovered or some will be cured. Some will be cured or some will be died. So we can see that here some people recovered. Some people recovered. Some people died. Here some people died. So that will change the prevalence because on the recovery and on the death, here people were changing. Okay. So our death and our recovery, the prevalence will reduce. So incidence and prevalence, there is a relationship that is prevalence will be, prevalence is always high. Okay. This is very huge. Privilege is incidence and duration. So incidence, if it is 10 cases per thousand population per year. So incidence will be always like this. Okay. This is time person here. This is time frame population and new cases. Privilege will be a total duration five years. It will be 50 per 10,000 or 5 percentage. Okay. So we can calculate if one is missing, if prevalence and duration is there, you can calculate incidence and same like voice. So we have completed the mobility measures that is prevalence and incidence. And it was proportion and rate actually. So tools of measurement were rate ratio and proportion. Okay. Thank you. I'll come up with another topic in next class.