The flashcards below were created by user
dhamalajeewan
on FreezingBlue Flashcards.

Define epidemiology?
The study of distribution and determinants of healthrelated states or events in specified populations, and the application of this study to the control of health problems.

What are the components of Epidemiological triad? [AI 94,UP 00,IOM 11]

Classify epidemiological studies?
 1. Observational study
 a. Descriptive studies
 b. Analytical studies: ecological, crosssectional study, casecontrol study, cohort study
 2. Experimental study/Intervention study [IOM 11]
 a. Randomized controlled trials/Controlled Clinical trial: patients as unit of study
 b. Field trials [AI 09] – with healthy people as unit of study
 c. Community trials

What are cross sectional studies?
 It is based on a single examination of a crosssection of population at one point in time – the results of which can be projected on the whole population provided the sampling has been done correctly.
 Prevalence is given by cross sectional study [IOM 2067] so it is also also known as ‘prevalence study’.
 In this study, exposure and effect are measured at the same time. [IOM 2065]
Cohort study doesnot measure prevalance. [IOM 2064]

What are cohort (longitudinal) studies?
 Cohort study is known by a variety of names like prospective study, longitudinal study, incidence study and forward looking study. In longitudinal studies, observation are repeated in the same population over a prolonged period of time by means of followup examinations. Longitudinal studies are useful:
  To study the natural history of disease and its future outcome.
  For identifying risk factors of disease. [UP 97]
  For finding out incidence rate [IOM 10] or rate of occurrence of new cases of disease in the community.

What are different types of cohort studies?
 a. Prospective (Concurrent) [AI 04] cohort:
  outcome has not yet occurred at the time the investigation begins
  begins in present and continues into future prospectively from "cause to effect"
 b. Retrospective (Historical) cohort:
  outcome have all occurred before the start of investigation
  investigator goes back in time to select his study groups ("cause") but again traces them "forward" through time up to the "effect".
  not actually a retrospective study but intact a "prospective study in retrospect" a "nonconcurrent prospective study"

What are the basic steps of conducting randomized controlled trial?
 Drawing up a protocol.
 Selecting reference and experimental populations.
 Randomization.
 Manipulation and intervention.
 Followup
 Assessment of outcome.

A study is designed to evaluate an antihypertensive drug in two groups of patients. What is this study called as? [IOM 98]
 Randomized controlled trial or controlled clinical trial.
 It is the most appropriate study to find out the effectiveness of a drug over the other. [IOM 00,01,04]

What are the factors determining sample size?
  Design of study
  Plan of statistical analysis
  Accuracy of measurements to be made
  Degree of precision required
  Degree of significance required
  Ratio of cases to controls in a case control study or prevalence of disease in cohort study

For calculation of a sample size for a prevalence study all of the following are necessary except: [AI 03]
A. Prevalence of the disease in population
B. Significance level
C. Desired precision
D. Power of the study
D. Power of the study
 Power of the study is the ability of a statistical test to detect a null hypothesis that is false. (this multiple choice question has been scrambled)

All of the following are true in a randomized control trial except [AI 06]
A) Baseline characteristics of intervention and control groups should be similar
B) Investigator's bias is minimized by double blinding
C) The sample size required depends on the hypothesis
D) The drop outs from the trial should be excluded from the analysis
C) The sample size required depends on the hypothesis
Sample size is an important, independent predictor of the precision of results while confirming or refuting an aetiological hypothesis. A small sample size lacks precision whatever the hypothesis. It is believed that sample size should not be less than 30 (whatever the hypothesis). (this multiple choice question has been scrambled)

What are the units of randomization?
 a. Individual randomization:
  is done when intervention is to be delivered at the level of individual
  E.g. Effect of drug, vaccine, surgery, etc.
 b. Group of individual/cluster:
  is done when intervention is to be delivered at the level of an organizational unit.
  E.g. health education [AI 12] health in screening programs, etc.

What is crossover study?
 In Randomized controlled trial, one group is treated and the other group remains untreated. But for ethical reasons, no group involved can remain untreated.
 In crossover study, no group involved remain untreated.
 All subjects receive intervention but at different times (eg AZT trials). For example, Group A receives AZT for 3 months, Group B is control. For the second 3 month, Group A receives AZT and Group A is control.

What are various types of blinding trials?
 Single blind trial : The trial is so planned that participant is not aware whether he belongs to the study group or control group
 Double blind trial : The trial is so planned that neither the doctor nor the participant is aware of the group allocation and the treatment received [TN 91,AI 96,IOM 11,10,AI 06]
 Triple trial : This goes one step further. The participant , investigator and the person analyzing the data are all "blind".

What are the various types of probability samples?
 Probability samples are samples in which the researcher can specify the probability of anyone element in the population being included. There are four basic kinds of Probability samples.
  Simple Random samples
  Stratified Random samples
  Cluster samples
  Systemic samples

What is simple random sampling?
 This is the most basic (simplest) kind of probability sample.
 Elements are selected at random from the entire sampling frame without any stratification or segregation of the sampling frame into subgroups or strata with similar characteristics.
 Requires a complete list of items within the sampling frame. [IOM 06]
 Suitable for small homogenous populations (large hetergenous populations will require stratification).
 Every element in the population has an equal probability of being included. [AI 09]

What is stratified random sampling?
A random sample of a population in which the population is first divided into distinct subpopulations or strata,[IOM 2068] and random samples are then taken separately from each stratum. An example of stratified sampling is population of a city is grouped in many groups according to their characters, and the sample is taken from each group. [IOM 11]

A village is divided into five relevant subgroups for the purpose of a survey. Individuals from each subgroup are then selected randomly for the purpose of a study. This type of sampling is termed as [AI 11]
A) Systematic sampling
B) Cluster sampling
C) Simple random sampling
D) Stratified sampling
D) Stratified sampling (this multiple choice question has been scrambled)

A region is divided into 50 villages for the purpose of a survey. 10 villages are then selected randomly for the purpose of a study. This type of sampling is termed as [AI 11]
A) Simple random sampling
B) Stratified sampling
C) Cluster sampling
D) Systematic sampling
C) Cluster sampling
In cluster sampling, target population is identified.
Target population is divided into naturally occuring subpopulations or clusters (e.g regional population is divided into villages).
Next, rando sample of clusters is selected (5 villages out of 50 villages are selected for the purpose of study) by simple random sampling or systemic random sampling. Individuals within the selected clusters may be further selected in total or by random sampling.
Stratification of population is done based on naturally occurring clusters. It is not done on the basis of one or more important characteristics. (this multiple choice question has been scrambled)

What is systematic random sampling?
 A common way of selecting members for a sample population using systematic sampling is simply to divide the total number of units in the general population by the desired number of units for the sample population. The result of the division serves as the marker for selecting n^{th} sample [IOM 01] from within the general population.
 For example, if you wanted to select a random group of 1,000 people from a population of 50,000 using systematic sampling, you would simply select every 50th person, since 50,000/1,000 = 50.

All of the following are true about cluster sampling except [AI 07]
A) The sample size may vary according to study design
B) Samples are similar to those in Simple Random Sampling
C) Is a Rapid and simple method
D) It is a type of probability sample
B) Samples are similar to those in Simple Random Sampling
Samples in cluster sampling are randomly selected groups, unlike samples in simple random sampling which are randomly selected individuals. (this multiple choice question has been scrambled)

"Design effect" is associated with which of the following sampling techniques: [AI 12]
A. Systemic sampling
B. Simple random sampling
C. Stratified sampling
D. Cluster sampling
D. Cluster sampling
 The design effect (difference) refers to the difference in precision of the estimates produced by a complex design (like a cluster sample) relative to a simple random sample.
 It is a ratio of variance of a statistic calculated from a cluster sample (or any complex sample) to that of the same statistic calculated from a simple random sample of the same size.
 The "design factor" refers to square root of design effect. (this multiple choice question has been scrambled)

What is confounding factor?
 A "confounding factor" is one that although associated with "exposure" under investigation, itself, independently of any such association, a "risk factor for the disease.
 For example, in the study of the role of alcohol in the aetiology of oesophageal cancer, smoking is a confounding factor because:
 i. it is associated with the consumption of alcohol and
 ii. It is an independent risk factor for oesophageal cancer.
 In these conditions, the effects of alcohol consumption can be determined only if the influence of smoking is neutralised by matching.

What are various important types of Bias and their solution?
 Selection bias – Randomization [KERELA 94, IOM 09]
 Confounding factor bias  Matching [MANIPAL 01,IOM 10,04,06]
 Interviewers bias  reduced by allowing equal time to interview case and controls [AI 08], eliminated by double blinding.

What is Berkson’s bias?
  Using the hospital medical records to represent the general population
  The bias arises because of the different rates of admission to hospitals [AIIMS 2001] for people of different disease.

The major purpose of Randomization in clinical trials is to [AI 06,07]
A) Ensure that the groups are comparable on baseline characteristics
B) Facilitate double blinding
C) Help ensure that the study subjects are representative of general population
D) Reduce selection bias in allocation of treatment
D) Reduce selection bias in allocation of treatment
Randomization is an attempt to allow comparability but does not ensure it. Randomization definitely reduces selection bias. (this multiple choice question has been scrambled)

How can you decrease sampling error?
 Sampling error is type I error (alpha error). [AI 01]
 Sampling error is the error due to chance.
 It can be reduced by increasing the sample size.
 In census, all the individuals in study population are included in the study, so it cannot have sampling error. [IOM 07]

What is sentinel surveillance?
 No routine notification system can identify all cases of infections or disease. A method for identifying the missing cases [UP 97, IOM 08] and thereby supplementing the notified cases is required. This is known as sentinel surveillance.
 Sentinel data is extrapolated to the entire population to estimate the disease prevalence in the total population.

The most important function of sentinel surveillance is: [AI 02]
A. To notify disease
B. To determine the trend of disease in a a population
C. Tp fine the total amount of disease in a population
D. To plan effective control measures
C. Tp fine the total amount of disease in a population (this multiple choice question has been scrambled)

What are the different kinds of time trend in disease occurrence?
 Shortterm fluctuations
 Periodic fluctuations
 Long term also called as Secular trend [AI 94, UP 98]

What is incidence?
The individuals developing new disease during period of time. [AI 02,01, IOM 00, 04] [Cases that come In] Incidence rate can be calculated only over a period of time, not as a single point. For example, 80/1000 patients observed for 8 years gives incidence. [IOM 10]

What is prevalance? [AI 98]
 Prevalence – the individuals with existing disease. [All cases] [IOM 01]
 Prevalence is defined as total number of all individuals who have an attribute or disease at a particular time (or during a particular period) divided by the population at risk of having the attribute or disease at this point in time or midway through the period.
 Prevalence rate may be point or period prevalence.
 Although it is referred to as a rate, the prevalence rate is really ratio. [AIIMS 05].
 Duration of disease affects it.

What is the relation between incidence, prevalene and duration?
 Prevalence = Incidence X Duration. [IOM 96]
 [Note that we cannot calculate incidence by dividing prevalence by duration , neither we can calculate duration by dividing prevalence by incidence. This relation can only be used to remember the direction of relation like the prevalence increases as incidence increase, or as duration goes down prevalence goes down ]
If the prevalence is very low as compared to the incidence for a disease, it is because the disease is very fatal and/or easily curable. [AIIMS 05]

During initial 8 years of study, out of 4000 patients, 45 patients developed CHD. What type of rate or ratio will define it best? [IOM 11]
Incidence rate.

22 new cases of TB recorded from 1st Jan to 31st June, with total population of community on 1st June being 1,65,000. If total registered cases of TB were 220, what is the incidence per 10 lakh population? [AI 10]
A. 220
B. 133
C. 22
D. 13.3
B. 133
Incidence (per 10 lakh):
= New cases/Total population X 10,00000
= 22/1,65,000 X 10,00000
= 133 (this multiple choice question has been scrambled)

Which of the following will cause an increase in the prevelance of disease? [AI 09]
A) Immigration of healthy person
B) Longer duration of disease
C) Increased cure rate of disease
D) Increased death rate in disease
B) Longer duration of disease (this multiple choice question has been scrambled)

What is number needed to treat?
 It is the number of cases to treat in order to prevent one case. It is inverse of incidence rate.
 For example, if incidence is 50 per 1000, Then NNT is 1000/50 = 20.

What is the incidence of AIDS cases in 1990?
 No of New case is 4, because the cases 3, 4, 6 and 9 got the disease in 1990.
 No of Population at risk is 6, and they are 3, 4, 6, 8, 9, and 10.
 1 already had AIDS and he cannot be at risk of catching AIDS, as they already have AIDS , and same for 2, 5, and 7.

What is the prevalence of AIDS cases in 1990?

In a village with population of 1000, a new test result showed following:
a. Positive
 Dz present: 180, Dz absent: 400
b. Negative
 Dz present: 20, Dz absent: 400
What is the percent prevalence of disease? [AI 05]
 Disease prevalence is related to the actual number of diseased persons present at a given point in time, irrespective of test results.
 So,
 Prevalence
 = total diseased/total population X 100%
 = (180 + 20)/1000 X 100%
 = 20%

If prevalence of diabetes is 10%, the probability that three people selected at random from the population will have diabetes is: [AI 04]
a. 0.01
b 0.03
c. 0.001
d. 0.003
c. 0.001
 Here, prevalence = 10%
 So, probability of getting diabetes in a person = 10/100 = 0.1
 And probability of getting diabetes in all 3 persons:
 = 0.1 X 0.1 X 0.1 = 0.001

In a prospective study 6000 subjects were put on beta carotene and 4000 were not. 3 of first 6000 developed lung cancer and 2 out of second 4000 developed lung cancer. What is the interpretation of the above results? [AI 02]
A. Beta carotene is carcinogenic
B. Beta carotene and lung cancer have no relation to each other
C. Beta carotene is protective in lung cancer
D. The study design is not sufficient to draw any meaningful conclusions
B. Beta carotene and lung cancer have no relation to each other
Here incidence of lung cancer in both groups is same (3/6000 = 1/2000 and 2/4000 = 1/2000) which shows no change in incidence despite carotene intake. (this multiple choice question has been scrambled)

What are the procedure of descriptive epidemiology?
 Defining the population to be studied
 Defining the disease under study
 Describing the disease by  time, place and person [AI 99]
 Measurement of disease
 Comparing with known indices
 Formulation of an aetiological hypothesis

How do you analyse data in cross sectional study, case control study and cohort study?
 Cross Sectional study  ChiSquare to ascess association.
 Case contrOl study  Odds ratio to estimate risk
 CohoRt study  Relative risk to estimate risk.

Define relative risk? [AIIMS 1991, IOM 2065]
Incidence on exposed/Incidence on nonexposed

Define attributable risk? [AIIMS 1996]
Incidence on exposed – Incidence on nonexposed

85% of lung cancer among smokers was due to their smoking. This is an example of [AI 07]
A) Relative risk
B) Attributable risk
C) Odds ratio
D) Population attributable risk
B) Attributable risk (this multiple choice question has been scrambled)

Which gives the incidence rates, case control study or cohort study?
 Cohort study [AI 09]
 Incidence rate can be measured only by a prospective longitudinal study where the same population is observed over a period of time by means of follow up examination.
 Since, this is the only study that gives incidence rates, the relative and attributable risk can only be calculated from cohort studies. [IOM 2059]
 Case control studies cannot give incidence rate[IOM 2061] thus cannot give RR and AR. It yields only estimate of RR (odds ratio).

Calculate the relative risk and attributable risk form the table?

If the attributable risk is 5/1000, how many people should be treated to prevent 1 case?
200 – it is the inverse of attributable risk.

In which type of study do we use the relative risk and the attributable risk?
 Incidence can only be calculated in cohort study, and incidence is required for relative and attributable risk.
 In the case control study, where the incidence is not available, we use the Odds Ratio [IOM 2067, 2053] . Relative and attributable risk cannot be calculated in case control study. [AI 2002]

How do you calculate ODDs ratio?
It is a cross product ratio  AD/BC [@ Amino domini/ Before Crist)
 It is to be noted that 'A' cell consists of 'risk factor' and 'diseased'.

Out of 50 total admissions, 20 girls out of which 10 needed surgery, and out of 30 boys, 20 needed surgery. What is the probability of picking a person requiring surgery? [AI 09]
3/5

Define Odds and probability? What is their difference?
 The odds in favor of an event are the ratio of the probability that an event will happen to the probability that it will not happen. For example, the odds that a randomly chosen day of the week is a Sunday are one to six, which is sometimes written 1:6
 Probability is a measure or estimation of how likely it is that something will happen or that a statement is true. For example The probability that a random day is a Sunday is oneseventh (1/7)

There are 5 pink marbles, 2 blue marbles, and 8 purple marbles. What are the odds in favor of picking a blue marble and what is the probability of picking blue marble?
 Odds  2/13
 Probability – 2/15

Calculate the Odds ratio from the table below?
 Odds ratio = AD/BC = (60 X 200)/(120 X 20) = 5.

What does odds ratio = 5 mean?
 If you have colorectal cancer, then you have 5 times more likely to have family history of colorectal caner.
 If you have family history of colorectal cancer, then you have 5 times more likely to get colorectal cancer.
 The first statement is better, because, in case control study, you start from disease and go retrospective to find the risk factor as family history.

What is difference between case control study and cohort study? [AI 12]
 Case control study//Cohort study
 Proceeds from effect to cause (retrospective study) [AI 02, TN 91, IOM 00] // Proceeds from cause to effect (prospective study)[IOM 00]
 Involves small number of subjects // Involves larger number of subjects.
 Lower cost // Higher cost
 Short time period // Longer time period
 Cannot yield information about disease other than that selected study // Can yield information about more than one disease outcome
 One disease : multiple past exposure // One exposure: multiple future disease
 Suitable for the study of rare diseases // Inappropriate when the disease or exposure under investigation is rare
 Generally yields only the estimate of RR (odds ratio) // Yield incidence rates, RR as well as AR
 Major source of bias : recall// Major source of bias : selection

What is the formula for sensitivity, specificity, positive predictive, negative predictive value, and Accuracy?
The common rule is True on top and divide by everything
 Sensitivity = True positive / (True positive +False negative) [AI 02,03,06]
 Specificity = True negative / (false positive + true negative) [AI 04,02, MANPAL 02, SGPGI 03]
 Positive predictive value [AI 06] = true Positive / (true Positive + false Positive). It is disease positive among test positive. [IOM 07]
 Negative predictive value = true negative / (true negative + false negative)
 Accuracy = (True positive + True negative) / (True positive + true negative + False positive + false negative) or (TP + TN)/ N
 [The easiest mnemonic tells you what they DON'T do:
 Sensitivity has an N in it and so it doesn't check for Negatives.
 Specificity has a P in it and so it doesn't check for Positives.
 All parameters in Positive Predictive value has 'P']

A drug company is developing a new pregnancy test kit for use on an outpatient basis. The company used the pregnancy test on 100 women who are known to be pregnant. Out of 100, 99 showed positive test. Upon using the same test on 100 nonpregnant women, 90 showed negative result. What is the sensitivity of the test? [AIIMS 2003]
Sensitivity = True positive / (True positive +False negative)
 True positive = 99
 false negative =1
 Thus, sensitivity =99%.

A new test of diabetes was carried out of the 80 people who were tested positive, it was found that acutually 40 had diabetes. Out of 9920 people who were tested negative, only 9840 did not have the disease actually. What is the sensitivity [AI 10] and specificity of this test?
 Total number of true positive = 40
 Total number of false positive = 8040 = 40
 Total number of true negatives = 9840
 Total number of false negatives = 99209840 = 80
 Sensitivity = True positive / (True positive + False negatives) × 100
 = 40 / (40+80) = 33%
 Specificity = True negatives / (True negatives + False positives)
 = 9840 /(9840+40) = 99.59%.

The area shaded in the graph below obtained from a new test for diabetes, represents: [AI 11]
A. True positives
B. True negatives
C. False positives
D. False negatives
C. False negatives
(this multiple choice question has been scrambled)

What is the effect of prevelance on sensitivity, specificity, Negative and positive predictive value?
 There is no effect of prevelance on sensitivity and specificity.
 Positive predictive values – increase with increase of prevalence
 Negative predictive values – decreases with increase of prevalence.
More false positive cases in a community signify that the disease has low prevalence. [AI 02]

Which type of screening test do you select to gain best positive predictive value – high specificity or high sensitivity?
 For best positive predictive value – select with high specificity
 For best negative predictive value – select with high sensitivity.
 SPIN and SNOUT are commonly used mnemonics which says:
 A highly SPecific test, when Positive, rules IN disease (SPPIN).  specificity is about healthy people. (Non diseased)
A highly Se Nsitive test, when Negative rules OUT disease (SNNOUT).  sensitivity is about disease. (to remember the association of sensitivity and the diseased people, remember that the diseased people are more sensitive)

The diagnostic power of a test to correctly exclude the disease is reflected by: [AI 05]
A. Negative predictivity
B. Positive predictivity
C. Sensitivity
D. Specificity
A. Negative predictivity (this multiple choice question has been scrambled)

For the diagnosis of SLE 6 tests are done of which 4 came as positive and 2 negative. To determine the probability of SLE at this point, we need to know: [AI 12]
A. Relative risk of SLE
B. Incidence of SLE and predictive value of each test
C. Incidence and prevalence of SLE
D. Prior probability, sensitivity and specificity
D. Prior probability, sensitivity and specificity
The question aims to establish the post test probability of having the disease (SLE), which is determined from the known sensitivity and specificity of the test and the estimated pretest probability of the disease. (this multiple choice question has been scrambled)

ELISA in comparision to Western blot technique is [AI 93]
A) Less sensitive and less specific
B) More sensitive and less specific
C) More sensitive and more specific
D) Less sensitive and more specific
B) More sensitive and less specific
Screening test should have high sensitivity. [AI 04]
 ELISA is sensitive test for HIV, so used as screening test for HIV.
 Western blot is a specific test for HIV, so used as confirmatory test. (this multiple choice question has been scrambled)

ELISA is performed on a population with low prevalence, What should be the result of performing double screening ELISA tests? [AI 01]
A. Increased specificity and positive predictive value
B. Increased specificity and negative predictive value
C. Increased sensitivity and positive predictive value
D. Increased sensitivity and negative predictive value
A. Increased specificity and positive predictive value
The specificity of ELISA increases by carrying out the tests in row using different HIV marker. (this multiple choice question has been scrambled)

A test has 90% sensitivity and 80% specificity in diagnosing the disease. The prevelance of the disease is 20%. What is positive predictive value?
 Let us suppose there are 1000 total population
 Then, the people having disease is 20%  200
 People not having disease is 800
 Sensitivity is 90%  so the test showing positive is 180
 Specificity is 80 %  so the test showing negative is 80% of 800, 640
 Drawing table,
 Positive predictive value is  180/(180 + 160) > 50%

The parameters of sensitivity and specificity are used for assessing: [AI 03]
A. Discriminant validity
B. Content validity
C. Construct validity
D. Criterion validity
D. Criterion validity
Reference: to search self. (this multiple choice question has been scrambled)

What is reliability and validity?
Reliability (Repeatability) – the test must give consistent results, when repeated more than once on the same individual or material, under the same conditions.
 Validity (accuracy) – the test should be able to separate or distinguish those who have the disease from those who do not. It is the measure of true positive and true negative. [IOM2061]
 It has two components  sensitivity and specificity.

