General H&S

Question 1

What is a significant sensitivity score?

Answer 1

Near 1 means Low false negative rate and ability to OUT out disease

Sensitivity is particularly useful for diseases with low prevalence.

Question 2

What is the desired specificity value for ruling in a disease?

Answer 2

Near 1, indicating low positive rate.

Specificity is a confirmation test to rule IN disease.

Best used as a confirmatory test after positive screening tests.

Question 3

How is the odds ratio calculated?

Answer 3

(event occurring in experimental group/event not occurring in experimental group) / (event occurring in control group/event not occurring in control)

This calculation helps compare the likelihood of an event occurring in two groups.

Question 4

What does an attributable risk greater than 0 indicate?

Answer 4

Reduction in risk

Question 5

What does an attributable risk less than 0 indicate?

Answer 5

Increase in risk

Question 6

What is late look bias?

Answer 6

Information gathered at inappropriate time

Question 7

What is the pygmalion effect?

Answer 7

Researcher’s belief in the efficacy of a treatment changes the outcome of the treatment

Question 8

What does the Hawthorne effect refer to?

Answer 8

Group being studied changes its behavior due to the knowledge of being studied

Question 9

In a positively skewed graph, how does the mean compare to the median?

Answer 9

On “patient” right so mean is skewed to be high
Mean is greater than median
median is greater than mode

Question 10

In a negatively skewed graph, how does the mode compare to the median?

Answer 10

Mode is greatest, then median, then mean

Question 11

What does a T-test check?

Answer 11

Differences in the means of 2 groups

Question 12

What does ANOVA check?

Answer 12

Differences between the means of three or more groups

Question 13

What is an ordinal variable?

Answer 13

A variable with inherent ordering, such as mild, moderate, and severe

Question 14

What is a discrete variable?

Answer 14

A variable that can only take certain specific values

Question 15

What does the chi-square (X2) test check?

Answer 15

Significant differences between 2 proportions of categorical variables, one that was obtained and one from the expected values.

Question 16

What is the purpose of a meta-analysis?

Answer 16

To assess the clinical effectiveness of healthcare interventions by combining data from 2 or more RCTs

Validity is based on the quality of the systematic review it is based on.

Question 17

What does a good meta-analysis include?

Answer 17

Sensitivity analysis, exploration of heterogeneity, and complete coverage of all relevant studies

Question 18

What is included in sensitivity analyses?

Answer 18

Exploring the effect of excluding various study categories and consistency of results across different subgroups

Question 19

Define precision in the context of measurements.

Answer 19

How close a set of measurements are to each other

Question 20

Define accuracy in the context of measurements.

Answer 20

How close a value is to the true mean

Question 21

In a forest plot, what do squares toward the left indicate?

Answer 21

Treatment is beneficial

Question 22

In a forest plot, what do squares toward the right indicate?

Answer 22

Treatment is less effective

Question 23

What does the size of the square in a forest plot represent?

Answer 23

Proportionality to the precision of the study and sample size

Question 24

What does the line of no effect represent in odds ratio?

Answer 24

An odds ratio of 1, meaning risk and benefit are equal

Question 25

What does a very low p-value in a test for heterogeneity indicate?

Answer 25

Studies are very different

Question 26

What does a large p-value over 0.1 in a test for heterogeneity indicate?

Answer 26

Studies are quite homogenous and measuring the same thing

Question 27

What is preferred over p-value for understanding effect sizes?

Answer 27

Confidence interval

Question 28

What must be assessed before interpreting confidence intervals?

Answer 28

Bias

Question 29

What determines the importance of a study?

Answer 29

Size of effect, not statistical significance

Question 30

What does a funnel plot check for?

Answer 30

Existence of publication bias in systematic reviews and meta-analyses

Question 31

What indicates potential publication bias on a funnel plot?

Answer 31

If one side of the plot has no dots

Question 32

What does Eager’s regression test for?

Answer 32

Used in meta analysis to assess for publication bias Examines Small studies and their effect sizes compared to expected effective size

Question 33

What does the Cox model do?

Answer 33

Isolates the effects of treatment from other variables and narrows the confidence interval

Question 34

What is censored survival time?

Answer 34

When the exact time of an event is not observed due to the end of follow-up or observation

Question 35

What is the mode?

Answer 35

The number that appears most often in a list

Question 36

How to calculate upper quartile?

Answer 36

Order the numbers in ascending order The number in the middle in the first half is

Question 37

How to calculate lower quartile?

Answer 37

Order the numbers I ascending order Find the middle of them

Question 38

In a graph with no sew, how are the values for mean, median and mode affected?

Answer 38

They are all the same in a graph without skew

Question 39

What is likelihood ratio definition?

Answer 39

Likelihood ratio is based on how a TEST is able to change the probability of having a disease. Based on test result of patient with the disorder compared to someone without the disorder.

Question 40

What is predictive value?

Answer 40

Probability of a CORRECT diagnosis in an INDIVIDUAL/ How likely a test is to correctly identify or rule out a disease, compared to those without it. It is dependent by prevalence.

Question 41

How is pre-test odds ratio found?

Answer 41

Based on the pre-test probability: Odds of occurring(pretest probability) /odds of not occurring

Question 42

How to find post test odds?

Answer 42

Multiplying the likelihood ratio by the pre-test odds

Question 43

What is the positive post test odds?

Answer 43

Positive likelihood ratio x pretest odds= positive post test odds This can be converted into post test probability by creating a denominator value by adding the post tests odds +1 E.g answer of 4, we would put this 4/5 to get a post test probability of 80%

Question 44

How to negative post test odds?

Answer 44

Negative likelihood ratio x pre-test odds To convert into post-test probability, add 1 to the value to create denomainotr. E.g value of 0.25 will be converted into 0.25/1.25

Question 45

What is a significant cut off value for positive predictive value?

Answer 45

Positive likelihood ratio must be over 10 -> this will indicate it is likely the person actually has the disease

Question 46

What is a significant cut off value for negative predictive value?

Answer 46

Less than 0.1 for negative predictive value -> this will indicate it is likely they do not have disease

Question 47

What is a significant off score for positive likelihood ratio?

Answer 47

PPV should be over 10 -> this means it is likely that they have the disease and is used to rule IN disease

Question 48

What is a significant cut off score for negative predictive value?

Answer 48

Less than 0.1 -> this means it is likely that they dont have the disease and is sed to rule out disease

Question 49

What is positive predictive value?

Answer 49

Proportion of positive tests that are actually positive out of all positive tests. It means that particular person probably has disease. TP/TP+FP

Question 50

What is negative predictive value?

Answer 50

Proportion of negative tests that are actually negative out of all negative test. It means that particular person likely doesn’t have disease. TN/TN+ FN

Question 51

What is the difference between predictive value and likelihood ratio?

Answer 51

Predictive value is based on the test’s ability to provide an accurate DIAGNOSIS from a particular eg number of positive tests out of ALL positives. Likelihood ratio is based on how a TEST is able to change the probability of having a disease. Basically CORRECTLY identify or rule out a disease eg number of true positives identified out of both

Question 52

How is sensitivity affected by prevalence?

Answer 52

No change

Question 53

How is specificity affected by prevalence?

Answer 53

No change

Question 54

How is PPV affected by prevalence?

Answer 54

Decreases due to more people included.

Question 55

How is NPV affected by prevalence?

Answer 55

Increases due to more people being tested.ego

Question 56

What is predictive value of a test useful for?

Answer 56

Useful for individuals and specific populations. It is dependent on disease prevalence.

Question 57

What is the likelihood ratio of a test useful for?

Answer 57

Useful for comparing the diagnostic power of a test in specific populations E.g high positive likelihood ratio means a value is more likely to be present in a specific population.

Question 58

What is the cause of asymmetry in funnel plot?

Answer 58

Reporting bias Publication bias like delayed publication, language or multiple publications Selective outcome reporting Selective analysis reporting Poor Methyldopa

Question 59

What is a t test based on?

Answer 59

Null hypothesis is true Evaluates whether this is accurate based on te p valu

Question 60

What is an unpaired T est?

Answer 60

Testing two different groups

Question 61

What is a paired T test?

Answer 61

Testing the same group twice

Question 62

How is chi squared testpmeasured?

Answer 62

Standard deviation x (obeserved result - expected result) ^2 divided by expected result We look at the column with the probability of 0.05 and degrees of freedom of the study.

Question 63

What is degrees of freedom?

Answer 63

Based on sample size -1 Used to determine column to select in chi squared test