Exam Recall Series (INI-CET May '23) - Community Medicine

Hello, namaskar, and a very warm welcome to the community medicine recall questions from INISET May 2023 exam. Well, a small disclaimer over here: this video does not intend to replicate directly the questions asked in the exam. Rather, the motive over here is just to discuss the topics and the pattern of community medicine questions which are asked in the May 2023 exam.

I duly understand that many students reported that the questions from pharmacology and community medicine were slightly—no, not slightly—they were on a very difficult trend, or a very tougher pattern. But I would say that there were few questions which were very easy, and there were very few questions which were very difficult.

Also, let's dive into the analytics of the questions. Total, I would say that there were 15 questions which were like pure hardcore community medicine questions. Of course, there were three or four more questions which have been discussed in other subjects by different faculties. But talking about these 15 pure community medicine questions, these were selected from broadly five topics: we had topics from epidemiology, biostats, communicable disease, non-communicable disease, vaccines, and some other miscellaneous topics.

Now, out of the 15 questions, can you believe that eight questions—roughly, we can say that 55% of the total questions—were from medical research? Which actually happens all the time in INISET or the AIIMS exam. So this has been happening even since the last 10-15 years. I have—we have all seen the same pattern happening in INISET exams.

Diving a little bit more deeper into the type of questions, or you can say the difficulty or the differential index of the questions, I have shaded these topics or these questions into green zone, yellow zone, and red zone. Green zones are definitely the questions which are very easy; the yellow are slightly moderate; and the red zone, I have taken them as beyond syllabus or like out of box questions.

If a student was not able to answer the green zone questions—that is the eight questions which are very easy questions—then I would say that the person or the student is not actually in the race to qualify into the INISET exam. At least you should be answering the very easy, and these are the very frequently asked questions; these are the most frequently asked questions. Okay.

Next is these beyond questions, the red category. Red category—I don't know why they asked, but actually I would say that that's okay for them to ask, because they—the INISET people or the AIIMS people—they have been doing this since so many years. When I was preparing for PG entrance at that time, 2005 and 6, they asked about Bland-Altman analysis; nobody knew about that. Kolmogorov-Smirnov test; nobody knew about that. And next years and next two to three years, they ask questions on sample size calculations using the relative procedure and absolute errors, alpha errors, beta errors—I don't know what all. So AIIMS people, INISET people, they do ask these types of beyond syllabus questions, which have been the trend.

But the red question, the red category, I would say that reasonably, like the majority of—not even majority, I would say 99.9 percent—of the students would not have answered these three questions which are beyond syllabus correctly. Of course, a rarity could be there, but exceptions could always be there. But questions on confidence bands in Kaplan-Meier—this is a reasonably difficult question. Kappa stats calculation—see, Kappa stats has been asked previously; it's a previously asked topic—but the calculation from Kappa stats, that is something new for this exam. Then they asked a question on immunodeficient states within, like the immunodeficiency—primary immunodeficiency, secondary—what vaccines are contraindicated and so on. So this I would say is very difficult. So these three questions, they do not make your rank; they do not decide the rank.

Now, the main differential questions or the main rank deciding questions were these yellow category questions. Why did I shade them yellow, not green, not red? Because these yellow category shaded questions—these are the topics—better listen careful—these are the topics which had been asked previously, and they have been consistently asked previously, very—like in last few exams—they asked a lot of questions on study designs and their protocols, guidelines like CONSORT, what is CARE protocol, what are STARDs, and so on. Next, we had questions on normalcy of data, Kolmogorov-Smirnov test, I told you, Dixon, QQ test. This has been asked. So how to find out normalcy of data? It is a very previously asked, a repeated topic. Snake bite—last few exams, out of eight, out of 10 exams in last three years, I would say they have been taking up questions from snake bite. Next is high fat intake. I don't know, like these years, like a lot of fancy diets are there: Atkins diet, DASH diet—not DASH diet—uh, Atkins diet, then you have these keto diets, paleo diet. All these diets are there, so it is like competitive for, and INISET people to ask the questions from these related topics as well.

So let's dive into the questions and very quickly cruise through what is the best possible correct answer and a very quick discussion about these topics. First, I'm going to take up is epidemiology. So the question number one goes: 30,000 females were followed up for a period of 10 years for development of breast cancer. 1200 females developed cancer and were given additional questionnaires for assessing the possible risk factors. Additionally, 2000 females from the same study were used as control, and they were also given the questionnaire. What type of study is this? Beta. This is a clear-cut example; this is a very pure, clear-cut example of a nested case-control. The nested topic of nested case-control has been duly discussed in the video number 8 of edition 6 Marrow module, Marrow video modules at this timestamp, and it is also there in my textbook, in my book, Conceptual Review of PSM, CBS Publishers, page number four.

So why do we call this as a nested case-control? For example, let us say you are planning to do a study. You select some group of individuals, maybe let us say, for example, 1000 people. You are selecting 1000 people. You select based on any—any—you take up them randomly or using any sampling method, you select 1000 people. Then you do a baseline assessment of these. You find out how many are smokers, non-smokers, how many have some comorbidity, how many have obesity or no obesity. You find out how many are literate, non-literate, socioeconomic—whatever you want to do—you do a baseline assessment, and then you do a follow-up. For example, let us say you selected some people in 2010, and you wanted to do a follow-up for 10 years, just for example, I'm saying. And of course, after 10 years, by 2020, you find out how many people were disease or non-disease. So of course, it's a 10-year study, and there is a follow-up. So this becomes a cohort study, very simple. But now let us take the numbers; let us talk a little bit about the numbers. If you say 1000 people are there, let us say by 2020, 50 people you found out that there were disease, and remaining 950 people, they were non-disease, right? So out of these 50 people, these 50 people did not get the disease directly in 2020. Somebody must have got them in 2011, somebody in 2012 and 13, five people, four people, so on. So common sense: every year, if you take up these cases, and let us say 10 people got in 2011, okay? So how many did not get? Out of 1000, 10 people got the disease; how many did not get? 990. Out of these 990, you take up some people who are controls; they will act as controls, and these 10 people will act as cases, and you can compare the cases with the controls within the same cohort. This phenomena, this concept, this funda, it is called as nesting; it is called as nesting of the case-control study within or the cohort. It is nested case-control study. So nested case-control, please remember, nested case-control is a type of cohort study. It is a type of cohort study where you take up a group of cohort, and then people get disease; you check them for any risk factors, and you check the controls for risk factors within the same cohort. You're doing case-control, case-control within the cohort. This is a nested case-control. If the question—there's nothing called as case-cohort study. This—this is just a filler, I would say. Cross-control cohort is nothing there. Then you have retrospective cohort. When do you say a retrospective cohort? So let us say that the question was like this: 300 females were selected from previous medical records. The—the key note of question over here, the key—the key point over here is previous medical records were taken in the year 2015 who visited PHC in Village XYZ for family planning services and adopted their either IUDs or OCPs. So you took females in 2013 with OCPs or IUDs, and then the selected subjects were interviewed today in 2023 and assessed for liver function, other metabolic functions, cancer, whatever. So what is this example of? This is a typical example of a retrospective cohort study where you go back in time, you select the cases, you select the individuals, and you check them today. So this is not a nested case-control. This example over here, this is a typical type of a nested case-control study. Please note that nested case-control, the advantages has been asked very frequently in your exam. The advantages are it is used in uh in in maybe in investigating applications which are investigations which are expensive investigations or they are rare investigations where you do not want to spend a lot of money. So if you select 30,000 females, rather than doing mammography on all 30,000, why not to do mammography only on 1200 and on 2000 females just to find out the level, right? So this is where the nested case-control studies would come into play, their advantages for expensive investigations or rare investigations.

Now coming over to question number two: Which of the following is not correctly matched? Systematic reviews for PRISMA, diagnostic studies for CONSORT, observational studies for MOOSE, and case report for CARE. This topic has again been discussed in the Marrow video number 10 and Conceptual Review of Community Medicine in the chapter on epidemiology, page 187. So a lot of times this is a very frequently asked and repeated question. It is a previously a repeat question. I would say this is not a new question; it has been previously asked in your AIIMS and INISET exams previously. So a lot of times epidemiological studies are there, and within the epidemiological studies there are some guidelines or protocols for strengthening of the protocols or strengthening of the quality of reporting of these study designs. So what do you mean by CONSORT? Beta. CONSORT stands for Consolidated Standards of Reporting Trials. It is used for randomized clinical trials or the control trials. So next is STROBE. STROBE is Strengthening the Reporting of Observational Studies in Epidemiology, or similar to MOOSE or STROBE which are for observational studies in epidemiology. You have observational studies in epidemiology with meta-analysis, that is MOOSE. So if you are talking about reporting of observation studies, that is true. Observational studies within the meta-analysis, the quality of observational studies which you are going to take up as that is controlled by MOOSE guidelines. Then you have very—like it's talk of the town; it is very frequently used guideline—that has the PRISMA guidelines. These are Preferred Reporting Items for Systematic Reviews and Meta-Analyses. So a lot of questions come on these systematic reviews, meta-analysis, evidence-based medicine. Then you have some preliminary studies like case report studies; you have the CARE guidelines, STARDs, QUOROM, STAIR, SPIRIT—all these are for standards of reporting; they are usually used for standards of reporting. So based on this, if you understand this, coming back to systematic reviews is PRISMA, that is correct. Diagnostic studies, we don't use CONSORT. CONSORT is primarily for randomized clinical trials or the control trials; they are usually for treatment rise or therapeutic trials. For diagnostic studies, we have beta STARDs or STARs—STARs, Standard of Reporting of Diagnostic Accuracy—that is for diagnostic studies. So the answer over here for diagnostic studies, you should have had STARD. So this is the false answer. Observational studies, we learned about MOOSE. MOOSE stands for the meta-analysis of observational studies in epidemiology. So that is in meta-analysis, what type of observational studies you're going to take, that is correct. Case report is CARE, case report—CA—case—a—report—re. So that is also correct. So what is the best answer you're going to mark over here? That is the diagnostic studies; CONSORT is not correctly matched.

Moving over to the next chapter from biostats. Now this was the lethal questions on biostats, but of course starting with the most easy one. This is one of the easiest questions which you would ever expect from biostats in INISET exam, and this is almost—uh—frequently seven out of 10 INISET exams in the past they have asked questions on standard divisions or measures of variations. So it's a very frequently asked topic. What is this formula correspond to? So the formula they had shown over here, the denominator was n-1 probably. So you had standard deviation, variance, mean, and median. Of course, it's not the mean or the median, right? Common sense. This formula is for standard deviation. We have talked about this in the video module number 16, chapter in biostats and Conceptual Review Community Medicine. Uh, this is not variance. Some students got confused. Better—what is variance? Variance is standard deviation squared. So if you just take the same formula, that is summation x - x bar squared divided by n or n-1 and you square it, that is what is standard deviation. But if you delete the square—so the square is gone—that is you square the under root, right? So this square root is gone, that is the formula for uh variance. So what is variance? It is nothing but square of the standard deviation. So this square root will be abolished in variance. So what is this formula? The best answer you would mark over here is standard deviation.

Let's move to the question number four: Mean plus minus 1 SD corresponds to—this is one of the easiest questions which you can ever pray to get in your exam, a prestigious exam like INISET exam. So this question has been discussed in Video 17 in chapter and biostats. So mean plus minus 1 corresponds to 68.3, 95, 99, 32.18. You already know that there is a Gaussian distribution of the normal distribution curve where within plus minus 1 SD, 68.3 percent; within plus minus 2 is the 95.5 percent; and within plus minus 3 is the 99.7 percent of the population, Northern values are going to lie. So plus minus 1 SD, that is 68—that is 68.32 percent of the population is going to lie. That is what the best answer over here is 68.3.

Let's move to the next question: All of the following are true for central tendency—this was, I would say, a googly question. This—what I would say, a googly question. This was a rank deciding, differential question. So which of the following statements are true for central tendency except? But of course, this question is—I would categorize this as a mild to moderate question, not very moderate or very hard, but yes, this was a tricky question. Arithmetic mean is affected by extreme values—common sense; it has so many times discussed this that mean is most affected by extreme values. This is true. Mode is most affected by extreme values—answer is false; this is wrong answer. Mode is in fact not affected by extreme values. Median and mode, they are not affected. So mode is false. Mode is most affected; this is the false answer. Median is calculated by overall observation—common sense. What is median? Median is the central value after you arrange in the ascending or the descending order. So after you arrange an ascending descending order, how will you arrange an ascending descending? That is only by observation. So you arrange an ascending descending, and then you find the middle value, that is the median. So option C is correct. A is correct. Mode can be used as qualitative data descriptor—common sense. You already know that the data could be of two types: you can have quantitative data, or we can have a qualitative data. For quantitative data, we can use the mean, the median, or the mode; anyone can be used, but mean is the best and the—and the easiest measure of quantitative data. If there are extreme values, in case of—ICO—in case of extreme values—in case of extreme values, we tend to use median and the last is mode. But in qualitative data, it can be of two types: we can have ordinal data, or we can have a—a nominal data, right? Ordinal data, we can use either median or we can also use the mode. But in nominal data, we you tend to use the mode only. So for qualitative data, we usually tend to use the mode, and it is a qualitative data descriptor. And this is what is uh has been asked in your exam. So option D is also true. So what is the best answer you would like to mark over here? The best answer you would like to mark is mode is not affected by extreme value; that is the wrong answer.

Question number six: All the following may be used to find the normalcy of data except. I would say that this question is not an easy question; it is not a difficult question also. If you have uh if you do the question with some presence of mind, and if you would have watched all the videos, I have discussed this in Video 17, Video 22, Video 19, at different times we have talked about these—how to say about—what is normalcy of data. So first is—I think many of times you would have seen this normal distribution curve, right? I talked to you about the normal distribution curve. Many of you already know it that within plus minus 2 is the anything which falls—that is taken as zone of normal. So you remember I talked about zone of normalcies. So that zone of normalcies, that is within plus minus 2 SD. What is this curve called as? This curve is beta. What is a plot histogram, box and whisker? I hope everybody knows about box and whisker. You see that the box and whisker, the within the box, you have the box and the upper and the lower quartile. This is called as the interquartile range. This is also a previously asked question in INISET, and I hope everybody remembers the IQR, that is the interquartile range, that is the value of the box. The box also tells us about the data distribution; it also tells us whether it is a normally distributed, yes or no. So box and whisker, yes, this is true. Plot histogram, yes, it will show us. Shapiro-Wilk test or Mantel-Haenszel test—these are probably the words which you might not have heard about them. Let me just have a quick word about the Shapiro-Wilk test. Shapiro-Wilk test—this—let us say this is the value, okay? Let us say that this is what is a normal distribution curve, and let us say you get me the blood glucose levels from your community. If you get me the blood glucose levels, they come out some things like this. If you can see the light blue shaded area, let us say this is what the researcher give me the data, and this is what is the original or the main or the—or the—you can say a standard normal distribution curve looks like—that is the red color. Now how—when and how and by what means or by using what criteria will I say that this researcher had given me a data which is a normally distributed? Is it like if the normal distribution is this, what if the researcher has given me this, or the researcher has given me this, or it came out—when to say a normal distribution, or do you think it always has to be on the same line only then I will say normal distribution? So what they do is they check the P values or they check the significant values from this—from the original and from the researcher data. Just listen very carefully. We are assuming that the researcher has taken the data, has taken the sample from the normal distributed sample or from the universe. So the—the researcher should have a non-significant P value. Non-significant means that there should not be any significant. So in a Shapiro-Wilk test, it is to be tested for normalcy based on the null hypothesis that the sample is drawn from a normal population—that is based on the P value. If the P-value is more than 0.05, what do you think is more than 0.05? That is non-significant. If the P-value is non-significant, that means—why work on a Shapiro-Wilk test? It means what? It means that the—that the researcher data, that the study data has a normal distribution curve; has a normal distribution curve. That means this P-value is a non-significant P-value; that is there is no difference.

You remember what is null hypothesis? Try to interrelate everything now. Now, in case the Shapiro-Wilk test tells us that the p-value came out to be 0.05, less than 0.05; if the p-value beta comes out to be less than 0.05 on a Shapiro-Wilk test, it means that whatever sample you had taken, that sample is significantly different from a normally distributed sample. Listen again, I'll say it again: if the p-value is less than 0.05 on a Shapiro-Wilk test, that means your study is significantly, significantly different from a normal distribution curve, and therefore it does not show the normal pattern. So that is what a Shapiro-Wilk test is. I know it was super easy for me to just say that, "Better Shapiro-Wilk test this for normalcy," but what if in the next INI-SET they ask something in detail about the Shapiro-Wilk test? So I don't know, like, because I have said this, probably there is a chance that they won't ask you at least this, but uh, just for satisfaction, you should be knowing what a Shapiro-Wilk test is.

Now, what do you think is this Mantel-Haenszel situation or the Mantel-Haenszel test? It is used for stratified or matched categorical data. So let us say there was a group of population whose age was more, or they were smoking, so they were different confounders in them. You remember in the confounding chapter, I talked about stratification—how to remove unknown confounders. Unknown confounders are removed by randomization, standardization, stratification, and regression. Now, once you stratify the group and then you see what population is getting the disease or not getting the disease, that analysis is done by Mantel-Haenszel tests. I know that this is way beyond the MBBS curriculum. In fact, I learned about this in PG during my PG curriculum only, in my third year. I clarified this from a very favorite sir of mine, a statistician, and he taught me how to do the Mantel-Haenszel test, and that's how I came to know. But uh, I don't know why AIIMS people ask this, but always it has been like this. But I would also like to tell you that don't bang your head on learning about in detail about Mantel-Haenszel tests. There is a very high chance that this question or this topic will not be repeated ever again in INI-SET. So these are some outlier questions which definitely come in your exam.

So anyways, if you know about this, if you are always able to rule out at least two options and you are left with doubt in two options, uh, personally speaking, I would say that go for the answer if you are doing very well; if you're scoring way more than 150, 160; if you have done at least 140, 150 questions in your INI-SET, I would say that in the remaining questions, if you are doubtful in two options, at least you give it a good try. But in case you have not scored very nicely, then you don't take a risk at all. But anyways, that was the question I wanted to discuss. The best answer over here is, "All of the following may be used to find the normalcy." So yes, all three can be used. Which of the following cannot be used? Not be used as Mantel-Haenszel test; that is, to check for confounding, or it is used in paired or matched categorical data. It is a test of significance for categorical data.

Question number seven: I know this question is again a very troublesome question, and uh, this question also would have the slightest possibility of ever being repeated again, but all said and done, let me just discuss this. In Kaplan-Meier survival curve bands used for confidence intervals, I had to research on this question to find out the best answer because there were two options which are very similar. So, option A is Hall-Wellner band, which is the most commonly used confidence interval plans. Hall-Wellner bands, equal precision. I was actually confused between A and D. So what is Hall-Wellner band, Hall-Wellner bands, unequal precision bands, and so on? So again, there is a very slight possibility of this question being asked, but let me just talk about Kaplan-Meier survival analysis over here. But if I just teach you Kaplan-Meier survival analysis, if I just teach you Kappa stats today, if I just teach you Shapiro-Wilk test today—it is, let us say, the 15th of May—so if you are talking about these tests today, on day one, your memory—whatever I say, listen carefully, whatever I say—it will be staying in your memory. So day one, day two, day three, let us say these are the days over here, and this is the percentage which is retained in my gray matter or the white matter or the yellow matter or whatever matter you have in your brain. So let us say that this is the level of the of the of the of the retainment of the memory. Memory retained is 100 at day one. Day two, day three, I taught you how to do Kappa stats, I taught you Shapiro-Wilk test, I taught you Mantel-Haenszel test, and so on. Day three, suddenly the memory will decrease by 10, so you have 90 remaining, then it suddenly decreases again, then more decrease, then more. That's how the memory is going to decay. By two weeks, end of two weeks, you will probably be left out. That's how it normally happens. Where are you going to remember all this, right? So, so that's how the memory retains. So whenever we are talking of prognosis, whenever we are talking about any retainment or how the person is going to respond, that is usually prediction. Predictions are usually drawn by prognosis, usually drawn by something called as the Kaplan-Meier survival analysis. Survival analysis, Kaplan-Meier survival analysis, which are always in a step-ladder pattern. That is how usually the survival analysis will be.

A very typical example, I'm going to show you from this article. The article was about the age of diagnosis, age at diagnosis, and the breast cancer survival in Iran, from a beautiful journal, International Journal of Breast Cancer. You had, like, what age group is going to survive more? So common sense, I think that age more than 35—if the age is less than 35 years, the chances of survival beyond 60 months, that is five years, was very high, like something around 0.8, 0.9, so 80 to 90 percent chance that the people less than 35 years, if they are detected with breast cancer, they are going to survive at least till five years, right? Ninety percent chance they are going to survive. But the age more than 65 years, the chance is 0.2 percent. That is how we understand this. So this is what a Kaplan-Meier survival curve is. Now, whenever we are talking of Kaplan-Meier survival curve, let us say this was the time in months or days or weeks, and this was the chance of probability. This is the curve that I was talking of. Let us say I show you this Kaplan-Meier. Now, now this is the Kaplan-Meier curve that you're seeing right now. Let us say this is with treatment number A, and this is—this curve is with treatment number B. So there were two treatments which are given, the blue treatment and the red treatment. Which one do you think is a better treatment? Common sense: the A treatment is better, much better than the B treatment, because the survival chances of survival with A is way, way more than the chance of survival with the standard or the B treatment, common sense. But now let us say that at this particular time, let us say this is corresponding to 28 weeks, for example, 28 weeks. So if this is 28 weeks, the chance of survival is 50 percent. But do you think that the person is going to survive exactly 50? But common sense, I think there is always a plus-minus a range should be there. So the person at this junction, at this junction, at this particular time, the person may survive from 0.6 to 0.4. So at this junction, the person is surviving from 0.6 to 0.4 probability chance. The median chance is 0.5. Do you understand this band is formed? Do you understand there is a, there is a path which is formed, on the median of that path is what is the curve, but there is always a range. This range, beta, this range, try to understand this range. It corresponds to 95, usually it corresponds to 95 percent confidence intervals, and these are what we call as confidence bands. These are confidence bands. So there are various statistical methods to find out the confidence bands. The different types of statistical methods like we can have the likelihood ratios, we can have log equal precision—that is what I was confused about, was it unequal precision or equal precision? It has equal precision. Hall-Wellner bands, these are the most commonly—and usually we use the Hall-Wellner bands for Kaplan-Meier survival analysis. We can also have the point-wise bands. So you can see that they try to compare different types of bands over here. So usually we use the logarithmic bands, better logarithmic bands, like the Gill bands, Hall-Wellner bands, Nair bands. See, why do we use the bands? Let us say that the patient who is coming at 12 weeks, so the chance of survival is something around 0.75, something like that. But the chance of survival, especially in talking about cancers, don't you think that is common sense, that is imperative, that the chance of survival will depend on background mortality rates for that particular population, healthcare access, do we have treatment for cancer, yes or no, population behavior, their health-seeking behavior, distribution of risk factors among the population, genetic influence, cause-specific death rates. People detected with cancer will die of cancer more, people or in countries where there is no cancer detection, people don't die of cancer because there is no cancer detection, and many other factors. So therefore, whenever we cannot say that this is the prognostic line, we cannot say that this is the prognostic line, we can never say that this is the prognostic line, we always will have to shade that area that probably it could be plus-minus this much. So what has the confidence band? These are the estimates of the standard errors of survival probability at each point and possibly the upper and lower limit of confidence interval.

So what did you learn today? Today you learned how to understand the Kaplan-Meier, and you talked a little bit about the confidence bands. The confidence bands could be point-wise bands, we can have logarithmic bands like the Gill bands, Hall-Wellner bands, we have the equal precision bands, and we can also use the likelihood ratios, and different types of bands can be used in this Kaplan-Meier survival analysis. So what is the best answer you would like to mark? Hall-Wellner or unequal? It is not unequal; this is false. Hall-Wellner is definitely the best answer; that is the most commonly used. What is this Holm method? I think you must have heard it in ophthalmology where they are doing a Hershberg test to just check for the strabismus or to check for exotropia, isotropy of the eye. So it is, I think, a filler. I, I've never heard of any statistical test called as Hershberg test, Holm test. Let us talk about Holm test. Let us say that uh, you you think of joining a coaching institute. Okay, you wanted to appear for the INI-SET exam or the NEET exam. You want to join the coaching institute. First thing that will come to your mind is, "Do I really need coaching? I am a very intelligent person. Do I really need coaching, or I think I can actually do the studies myself?" So you will go to your friends and you ask them, "Do I need coaching? What are you doing? Do you do coaching? Do you want to go to coaching? Yes or no," and so on. So you get different people saying different answers to you. Then okay, you say that, "Okay, I'll join coaching." Then you'll have to decide which coaching institute to join. You go to your student, your friend number one; they will say, "Go to institute number A. The faculties are very good, but they have low online presence. They will have less doubts or more better GTS and authenticity of the answers, reliability of the answers, validity of the questions, completeness of the syllabus," so many parameters are there. So you ask your friend one, he says that this institute is better. Friend two says that this institute B is better. Friend three says that C is better, and at the end you are the person, and you are in a big mess. You don't know what to do actually. So what has happened over here? I'm trying to solve this using statistical methods. What has happened over here? You consulted different friends, and they gave their own hypothesis. Whenever you have multiple hypotheses, listen carefully, whenever you have multiple hypotheses, there is a problem of multiplicity. You tend to get—listen carefully—you tend to get positive results when you see so many hypotheses. If I just give you two things that this could be either this or this, it's easy for you to choose. If I give you 20 options, then usually you get messed up, and there is a high chance—listen carefully—there is a high chance of false-positive error that you tend to find the answer because you have so many things. Then don't you think that you say that, "Okay, this raster is good, this is good, this is good. Okay, I'll use this raster." So you tend to find, although that that option may not be the valid option, so you tend to find a positive result whenever I give you so many options. That is called as probability of obtaining the type 1 error will increase, and for that, that phenomena, that concept, we use the Holm-Bonferroni test. The Holm-Bonferroni test is used for checking multiple hypotheses and to decrease the type 1 errors which may arise whenever we are assessing statistically, when we are assessing multiple hypotheses. For your exam, remember Holm-Bonferroni test; that is for analysis of multiple hypotheses to decrease the type 1 error or the type 1 bias which may arise in this type of study. So technically, Holm-Bonferroni test is not a method for confidence interval; this is also false. So which is the best answer? Hall-Wellner, of course. Again, I would say uh, with the good amount of confidence that there is a very less chance that these topics or these questions will be repeated again, but yes, Kaplan-Meier, I would say Meier has been frequently asked in your exam. When I was preparing at that time also, there was a question on Kaplan-Meier survival analysis, and again this year you again get a question, of course a detailed question on Kaplan-Meier survival analysis. Let's go to question number eight now.

Question number eight: Calculate the degree of agreement as Kappa stats for the following data set comparing the IGRA and the Mantoux test. So now you want to calculate the degree of agreement. So Kappa stats, I would again say that Kappa stats has been a previously asked question. It has been discussed in video number 19. Of course, I did not discuss how to solve Kappa stats, how to find out the Kappa stats, but uh, at least the formula was told. But this time in your AIIMS exam, they did ask you how to calculate Kappa stats, and majority, 99 percent students would not have attempted this. In fact, these are the questions where you would save time by not attempting them. So anyways, intelligent question, intelligent students, or maybe way beyond, or maybe the alien type of students would have attempted it, and they would have got the correct answer. Let me just talk to you how to solve this question on Kappa stats. We had been given two tests, IGRA test and Mantoux test, TB suspected, no TB suspected, TB suspected, and no TB suspected. So 80, 20, 40, 60, these numbers were given; they could be jumbled up, so don't worry about which number was on which side. At least look at what the concept was. So whenever we are going to talk about Kappa stats, the first thing for you to do is you should know the formula of Kappa stats. The formula for Kappa stats: Kappa will tell you like if I am an observer, if you are an observer, if we are checking one question paper, how many marks will you give? How many marks will I give? And do we agree upon that? That is what Kappa stats is. Okay. So Kappa stats is observed agreement minus the expected agreement divided by total agreement minus the expected agreement. So that is observed minus expected. I will write it as AO, expected I am going to write it as AE. Total agreement minus this is the total agreement, agreement total minus the expected agreement, that is AE. This is what is the formula for Kappa stats, and that is Kappa stats is useful to check the level of agreement between two observers or two researchers. So they had given the IGRA test over here, and 80 and 20, 40 and 60. You know that if you talk about the IGRA test, if you talk about the IGRA test, how many times did the IGRA test say yes, TB? Can I say that the IGRA test gave me a result 50 percent of the time they said yes, TB, and IGRA said 50 percent of the times no TB, 100 people yes TB, and 100 people no TB? Cannot do like that. Talking about the Mantoux test, the Mantoux test, can I say that how many times did Mantoux say yes and how many times did it say no? The yes, Mantoux test told me 120 times, and no told me 80 times. Can I say like this? Now I am going to find the observed agreement or the expected agreement. See, observed agreement where they both are agreeing, the IGRA and the Mantoux are agreeing. So let us say that—not let us say—from the data it is seen that 80 percent, 80 times the IGRA said yes, TB, and Mantoux also said yes, TB. So 80, they both agreed. Again, at 60 times or in 60 cases, in 60 individuals, the IGRA said no TB, and the Mantoux also said no TB. So this is what is the observed agreement. What is the observed level of agreement? 80 plus 60, that is 140. That is the observed level of agreement. What is the total agreement? Total agreement would have been 200 if 200 people everybody saying yes TB or no TB. So total agreement is 200. What is expected agreement? How do you find out expected agreement? How many times did they both say yes or how many times did they both say no? You just add them up; that is what is expected agreement. So how many times did IGRA say yes? IGRA said yes 100 times, am I correct? Said yes 100 times. That is the 100 times IGRA said yes, multiplied by 120 times the Mantoux said yes. So multiplied by 120 divided by total, that is 200. So, so how much is the answer? 0 0 cut. So 120 by 2, that comes out to be 60. How many times did they say no? So can you see that no TB was said by IGRA 100 times, so it is 100 multiplied by—no TB was said by Mantoux 80 times, so 80 divided by 200 again, 0 0 cut. So 80 by 2, that is 40. So if you just add 60 and 40, that gives me the expected agreement, that is 100. That is the expected agreement. Now it is super easy for us to calculate the Kappa stats. Of course, with due respect, with due apologies, everything uh, it's it's a it's a kind of very difficult to calculate Kappa stats in the exam scenario without pen and paper when you are too much under pressure, but that's the way to calculate Kappa stats. So how do you calculate Kappa stats? You already know Kappa is given by observed agreement minus expected over total minus expected. That is how much did we get? 140, 100. So we got 140, we got 140 minus 100 divided by total, that is 200 minus 100. So answer is 40 over 100, that is 0.4 is the best answer to find out Kappa stats, and that is what it means over here, how to find out the Kappa stats. So what is the best answer over here? You are going to mark answer is D. Option 0.4 is the best answer you are going to mark. So 0.4 Kappa, it means what? It means generally speaking, 0. anything which is less than 0.5, it usually means less agreement. Anything more than 0.7, we always say it is a strong agreement. More than 0.8, it is a near-perfect agreement. Moderate agreement is 0.4 to 0.6. So what is this agreement? This is a fair type of agreement. There may be some agreement, yes or no, but this best answer over here is it is a fair agreement, and that is what the next level of exam questions or the INI-SET pattern questions can be asked in our next upcoming exams. Now, coming over to the next chapter about communicable disease and non-communicable disease, question number nine: The slogan coined by WHO, do it right…

It has been discussed on social media platforms. It's also available from my website, psmsimplified.com. It's discussed in the book also. So, do it right—that is the slogan for first aid for snake bite. It includes all except, so you have to find out all except; you have to find out the wrong one.

Go to hospital; incisions; tell sequence of events in detail after the bite; reassurance. What do you think is, do it right? So, beta, this uh, do it right. It is a slogan which has been modified; it is carry no right. So, rather than saying do it right, of course, right is the main word, but we can also add the mnemonic carry no right. Carry means always carry the victim; never allow the victim to walk, even for a short distance. No means three things: no tonic way, no electrocution, no electrotherapy, no cutting of the wound, nothing doing. So, these things should be avoided; absolutely avoided. Then is Right—R-I-G-H-T. R stands for reassurance; you have to reassure the patients because uh, mostly 75 percent of the times the snakes are non-poisonous. T I stands for immobilize; very important it is to immobilize the limb, or the arm, or leg, or foot, or whatever, just as we do in fracture. You take the upper upper joint and the lower joint and you immobilize that. So, it has to be immobilized. Please note that: do not apply any compression or tight bandages; they can be dangerous. So, tight bandages are to be avoided; just immobilize the joint. It is not pressure bandage. Uh, get to the hospital immediately—that is GH, go to the hospital. And T stands for tell the doctor what happened, or do not uh, like go with traditional medicines, traditional things. So, just tell the doctor what happened and straight away go to the doctor; do not waste time in doing traditional things or things like this. So, what is the best answer over here? Go to hospital? Yes, we will go to the hospital. Incision is absolutely no incision, no cutting, nothing doing. So, that is false. Tell sequence—that is true. Reassurance—that is true. So, what is the best answer over here? It is giving incisions—that is the wrong one.

Question number 10: A person had a dog bite. A history of vaccination is up to date for both dog and man. The person in history is significant for having splenectomy five years back. What is the best possible management for this case? So, this year the examiners did not ask about the anti-rabies vaccine schedule—pre-exposure, post-exposure, re-exposure—rather they asked about whatever antibiotic we need to give, or do we need to give antibiotics? Government of India National reviews control program guidelines—they just say that give appropriate antibiotic. But there are UK guidelines—that is the NICE guidelines 2020—which talk about which antibiotic to be given. The best answer over here is amoxiclav. Let me just dive into the NICE guidelines. These are directly taken from the NICE guidelines 2020. It is said that amoxicillin, as a drug, as an antibiotic, it is more hydrophilic and more effective than penicillin, and especially it has the ability to pass through the outer phospholipid membrane and therefore it is active against the Pasteurella multocida, which is one of the most dangerous organisms when we are talking about animal bite, against the gram-negative bacteria. So also, amoxicillin has shown to have effect against the gram-positive aerobes and the anaerobic bacteria. But this amoxicillin is much better advocated; it is the first drug of choice; it is the drug of choice in case of any animal bite. Now, in certain cases there could be penicillinase producing microorganisms, so therefore co-amoxiclav, or with clavulanate, that is the drug of choice. So, amoxiclav, amoxyclav, that becomes the drug of choice. Alternate drugs which can be used: we can use levofloxacin, we can use doxycycline, we can use ciprofloxacin, trimethopram sulfamethoxazole—that is cotrimoxazole—plus metronidazole for the and metronidazole or clindamycin for anaerobic activity. So, this is directly taken from the the NICE guidelines. If you just zoom into it, first choice oral antibiotic—that is co-amoxiclav—that has to be given minimum for three days for profile access; it can be extended to five days in case there is some infection over there. So, what is the best answer you're going to mark over here? The best answer we are going to mark for the exam is co-amoxiclav or amoxiclav, which has to be given to the patient.

Question number 11: Match the incubation period of diseases. So, this question is again one of the easiest question which you can ever think to get in a prestigious exam, in a set exam. Syphilis is uh, 21 days, 10 days, 90 days, and 50 days. What do you want to answer? Syphilis is directly going with 90 days. We should all remember syphilis can range from as low as 10 days to 90 days; average we say 20 to 25 days. SARS, that is COVID-19; SARS or COVID-19, the incubation period is usually from 5 to 6 days to as long as 14 days. Generally speaking, SARS is 10 days. Talking about chickenpox, chickenpox, we can take it to 21 days. I think everybody knows about it—21 days, from 14 to 21 days we have this chickenpox. Hepatitis A is 50 days. So, what is the best answer? You go with 1C, 2B, 3A, and 4D. 1C, 2B, 3A, 4D. That's the first answer that is you're going to mark this as the best possible answer. This question has been taken directly from the video uh, number 27 from the conceptual review of community medicine and extra 28. So, this question has been taken directly; in fact, in the videos, if you just go, I have highlighted these topics that these can be asked very frequently in your exam.

Question number 12 goes like: A child came for DPT vaccination at 10 weeks with a past history of fever more than 40 degrees and inconsolable cry at six weeks following immunization or following vaccination. What should be done next? Better this question is about DPT vaccination. It is a very frequently asked question. If you got this question wrong uh, then I would say you need to reassess yourself; you need to reassess your strategy, that why did you get this question wrong. You have to study it again. Uh, one of the very common things uh, students told was that sir, inconsolable cry was given; there was no timeline which was given; it was not given one hour, two hour, three hour, or what is happening, or the fever was not so high. That's okay; this is all overthinking. Keep it simple; keep it straight. The question says that the child had taken six-week vaccination. What are we giving at six weeks, beta? Under the national immunization schedule, you already know at six weeks we are giving is the pentavalent vaccine. Now, in country we are not giving DPT vaccine at six weeks, or 10 weeks, or 14 weeks; we have upgraded ourselves to pentavalent vaccine. Pentavalent contains DPT—diphtheria, pertussis, tetanus; it contains H influenza B; it contains Hepatitis B. So, pentavalent plus fractional IPV plus we also give the Rota virus vaccine plus we also give the pneumococcal conjugate vaccine plus, of course, the oral polio vaccine which is given at six weeks. So, these are all the first dose vaccination—this is the first dose of vaccine, prefers dose pentavalent, first dose fIPV, first dose rotavirus, first those PCB, first dose OPV; it is the OPV1. Okay. Now, the question says there is inconsolable cry. When do you classify a cry as an inconsolable cry? It is usually on a single day, within 24 hours; if the child is crying for more than three hours, it could be five hours, six hours, ten hours, whatever number, but of course the child is not going to cry continuously for 10 hours; the child will probably not be alive. If the child is crying continuously for 10 to 12 hours, right? If the cry is continuous, inconsolable cry, continuous cry for more than three hours in a single day, that is inconsolable cry. Even if they have not given the criteria, then why do you overthink it? So, keep it simple; keep it straight. They have given you six weeks, inconsolable cry with fever. This is a typical example that there could be chances of pertussis vaccine adverse effect, and pertussis known to cause inconsolable cry; it is known to cause fever; it may be have a associated with neurological deficits. So, this child, particularly if this child really comes to you in a clinical scenario, you have to assess all the neurological features of the child; you have to assess the neurological complaints; you have to assess for any neurological deficits in the child, and then assess whether we want to give DPT, yes or no. As far as this MCQ is concerned, we are going to avoid DPT vaccination. Yes, this is true. DT vaccine—yes, this is definitely true. Give DPT vaccine—this is false. Differ by four weeks—there's also false. So, what is the best answer? Yes, A and B both are true, but B is much better; B is common sense, much better answer. This question again has been discussed in two, three times in your video modules; of course there could be more; I just found out two times it has been discussed in multiple pages and conceptual review book also. So, this is a very frequently asked question, but I don't get it wrong all the time. Keep it simple; keep it straight. They ask you pertussis; they want to ask you for pertussis; they don't want to trick you anywhere for pertussis. Okay, just very quick, what are the vaccine contraindications? So, first is when it anytime whenever there is severe allergic reaction. But do you also know that and whenever there is allergic reaction, talking about the allergic reaction, allergic to one conjugate vaccine like the DPT vaccine which contains the pentavalent or the DPT or the H influenza B, then other conjugate vaccine like PCB should also not be given to the child. Then there is allergic to streptomycin, pneumocin, or polymixin, then IPV cannot be given. Vaccination is also taken under medical advice. If the child is allergic to egg proteins, history of intussusception, then rotavirus vaccine—this is a very frequently asked question. Immunocompromise—that was the again the inset question; we are going to talk a little bit in detail about the immunocompromised status, unconscious or the hospitalized child or moderate, mild, acute, severe illness, then again these vaccines are contraindicated. So, anyways, coming back to the question, what is the best possible answer? You are going to give the DT vaccine; avoid the pertussis component because pertussis component is associated with inconsolable cry; inconsolable cry, it is associated with neurological deficit; it is associated with neurological deficits.

Okay, let us go to the next question that is about immunodeficiency. This is a red mark area; this is a red color area because uh, at MBBS level, I, I, we obviously as faculties, as teachers, we do expect that the students should be knowing a lot, but uh, talking in detail about the level of immunodeficiency, maybe you should be knowing, or maybe this question I would categorize as a red zone question. The source says the Nelson textbook of Pediatrics, 21st Edition, page 1361, table 197.9. It is also taken from CDC guidelines. What level of immunodeficiency, what vaccines to be given or not given? So, in which of the following conditions is vaccination not contraindicated? DiGeorge syndrome, which is immunodeficiency; Viscot Aldrich, again immunodeficiency; Ataxia telangiectasia, again immunodeficiency; complement deficiency, again immunodeficiency. So, which of the following do you think is not a contraindication? Please note that live vaccines—meningococcal vaccine, dingo vaccine, any of the conjugate vaccine, any of the conjugate vaccine—they all are contraindicated along with the live vaccine in primary immunodeficiency where there is B cell immunodeficiency; example is antibody deficiencies, X-linked agammaglobulinemia, common variable immunodeficiencies, and so on. Live vaccines and I'm not talking about the conjugate vaccine, predominantly the live vaccines, they are contraindicated in primary immunodeficiency like the T cells, cell mediated immunodeficiency like the SCID. We can have the DiGeorge, Viscot Aldrich, ataxia telangiectasia, all these are primary T cells deficiencies, and all live vaccines are contraindicated. Live vaccines are also contraindicated in primary immunodeficiency where there is phagocytic dysfunction or the phagocytic function is not appropriate. There is primary immunodeficiency; example is chronic granulomatous disease, leukocyte adhesion deficiency, leukocyte addition deficiencies. It is also contraindicated in secondary immunodeficiency, live vaccines like HIV and AIDS, especially if the CD4 count is less than 200, or if the relative CD4, relative CD4, if the relative CD4 count, relative CD4 count is less than 15 percent, or if the patient is showing HIV stage three or four, then definitely live vaccines are contraindicated. But a child who is born to an HIV-positive mother, in those cases, please note, a child who is born to an HIV-positive mother, the live vaccines are not contraindicated because the mother was HIV, child we don't know his HIV. Please refer to the guideline, to the module on HIV where I have discussed a child who is exposed to HIV. But anyways, please remember that HIV, CD4 count less than 200 cells or less than 15 percent relative CD4 count, all live vaccines are contraindicated. Secondary immunodeficiency, malignant neoplasms, or if the patient is on immunosuppressive therapies or the transplant patients, in these cases live vaccines are contraindicated. Please note, persistent complement Factor B providing deficiency; no live vaccines is contraindicated. In fact, to these patients we actually give the pneumococcal conjugate vaccines or the PPV vaccines just to have profile access from the pneumovocal infections as well, but live vaccines is not contraindicated, and that was the catch the question had taken up. So, what is the best answer over here? Complement deficiency. If you knew it, very nice; I'm so proud of you. If you did not know it, I personally don't think that this is actually a rank decided question, but yes, this could be important from your clinical perspective point of view. So, the best answer over here is complement deficiency. So, option D, that is 4 only, that is the correct answer in this question.

Coming to the other miscellaneous topics, just one or two questions. High fat intake—that was interesting question because it is from the recent advances because we are undergoing, we are seeing epidemic of obesity and metabolic syndrome. So, high fat intake, according to FDA, is this question has not been discussed in the Indian guidelines from the Indian perspective, but of course it's slightly important from biochemistry point of views. So, we do have known about the keto diets; you must have talked about the keto diets. Keto diets are predominantly high fat intake. That's so I've taken two resources, two articles, very good articles; I would suggest that if you get time at least go and see uh, high fat and obesogenic diets, and there is another ketogenic diet and weight loss. So, these are two interesting articles for read through. So, how to talk about keto diets? You already know that keto diet, it is beneficial in terms of weight loss because keto, that is high diet, so there is low carbohydrate intake, low carb carbohydrate intake of less than 30 grams a day. So, there is less insulin secretions, and there is a positive metabolic effect that there is low body weight, there is oxidative stress decreases, chronic inflammation decreases, and so on. So, ketogenic diet does have some metabolic effects. When do you say a diet is a ketogenic diet? Diet is a ketogenic diet when the fat intake is more. So, usually we say that the fat intake, if anytime it is more than 50 percent of the total calorie intake, you already know from the recommended dietary allowance, beta, what is the ideal fat intake? 10 to 30 percent. If it is 30 to 50, it is moderately high fat, but more than 50 it is taken as a keto diet, high fat diet. So, coming to a table over here, this table is a very frequently asked table and it is really important for your exam. What is the recommended dietary guidelines? Recommended dietary guidelines for carbohydrates—that is 50 to 80 percent of the diet of the calorie intake should be carbohydrates; 10 to 15 percent should be protein; and 10 to 30 percent should be fat. This is a very frequently asked topic in your exam. Now, when do you say a keto diet? A keto diet we can have is as a standard keto diet, which is very difficult, which is 10 percent of carbs and 70 to 80 percent of fats, and remaining 15 to 20 will be proteins. So, very high amount of keto of fat—that is a very difficult diet; that is a standard keto diet. Mostly people what they do is a targeted keto diet. What is a targeted keto diet? Where you increase the carbs to at least 10 to 15 percent; proteins are usually kept at 20, and the fat intake is 60 to 70 percent, not 70 or more; it is 60 to 70 percent—that is a targeted keto diet. This is the most common diet which all the dietitians would also be given to anyone who's taking a keto diet. But many people are also taking a cyclical keto diet where they keep a keto diet, a standard keto diet for four to five days, and they keep the regular diet where they take in excess of like normal carbohydrate, the recommended dietary allowance for two to three days. So, in the four to five days what the standard keto diet? 10 percent of carbs, which is very low carbs; protein is 15, 20; fats is 72, 75, which is on a higher side, and that is what they normally do with the keto diet. There is also a special subtype of a keto diet where they keep a high protein keto diet where the fat intake is 60, lower side, but the protein is on the higher side, 35 protein. So, predominantly the fat, the diet is made up of fats and proteins only; there is very low or minimum carbohydrates. So, these are different types of keto diets which we can have. Based on this, if you understand, coming back to the story, we have is something called as high fat—that is a keto diet, low calorie diet. So, because the person is taking high fat, the the calorie is usually kept around 1000 calories, less than 800, it has taken as very low calorie. Okay. So, 800 to 1000 is kept as a decently okay type of calorie diet; protein is 250; carbs is 150; and the fats we already learned is 500 to 600 out of the total total calorie; 500 to 600 is what is from the fat. So, based on this, if you understand, diet should contain 80 fat—that is kind of a wrong answer; it's plus or minus; yes, it can contain, but it's okay; 80 is not the criteria; more than 60 would have been the better answer. That should contain 400 calorie—that is wrong; 400 is way too much. 800 calorie from fat—that is also way too much; 800 calories. 80 fat and diet should contain 1000 calories—that is what is the diet and uh, that is what is the recommendation for a ketogenic diet or a keto diet where we tend to keep the calorie intake of half like around 1000 calories, and out of this 1000 calories, 50, more than 60 percent is what is the fat intake, and that would constitute the keto diet. So, what is the best answer you would like to mark over here? That should contain at least a total of thousand kilocalories for for a good keto diet to happen.

Coming to question number 15: CDC has classified 30 organisms into three categories of bioterrorism agent. Which is the false statement? So, common sense, I don't know, like this is one of the easiest question you would ever get. So, three categories, when the question is saying three categories and they are getting category A, B, C, and D. So, obviously which is the wrong one? Category D is not there; category D is not there. So, which is the false statement? Category D is not there. We have category A, B, and C. It is discussed in video 99; it is also there in community medicine, conceptual review of PSM, and so on. So, you have category A, B, and C under bioterrorism agents. The category A—these are easily transmittable and from person to person; they can transmit, and this is what is the highest priority, highest priority because these are diseases which can happen anytime; example is anthrax, botulism, plague, smallpox, tularemia, Ebola, and they need special action for public health impairment, disruption preparedness; there could be social disruption, high mortality, and people may die. Category B is of moderate intensity, or they they uh, there is moderate priority; these infections are easy to disseminate; they have low morbidity and low mortality, okay, lower than the category A; example for this could be any food or waterborne; example could be waterborne; could be called; we can have foodborne illnesses, brucellosis, Q fever, typhus fever, all come under the category B, tier bioterrorism agents which are moderate priority. Category C is low priority; it is not low in the sense that the infections are not important; it is because the infections do not exist as epidemics or outbreaks in the in nature, and but they can be artificially produced and disseminated. So, there is ease of production and dissemination, and there is a potential—the key word is potential—these infections, if they spread, they may cause high mortality and morbidity, but they are just a potential of infection; example is all the emerging infections like…

Nipa virus, antivirus; so these are the categories: category A, B, and C. Until now, the WHO, the CDC—they have not categorized COVID-19 under any category. But probably it could be category B, but we don't know that.

The best answer over here, you are going to mark—there's a simple, direct MCQ—the best answer you're going to mark is option number D: Category D does not exist in nature.

And so, with that, thank you so much for joining in the video and uh, discussing all the community medicine recall questions. So I'd like to reiterate this point: Yes, there were some difficult questions, but there were some easy questions. You have to focus on the green and the yellow zones. You have to focus predominantly on the green and the yellow zones. Don't get these topics wrong; these are very frequently asked in your exam. And in the next INSA set exam, these topics are again going to be repeated. All the very best, and take care.