PRACTICE 3

Question 1

What does incidence measure?
A. Total cases at a specific time
B. New cases over a period of time
C. Risk difference
D. Attributable risk

Answer 1

Question 2

Prevalence is affected by:
A. Only incidence
B. Duration of disease only
C. Both incidence and duration
D. None of the above

Answer 2

Question 3

In a study, 1000 people are followed for 5 years. If 50 new cases occur, what is the incidence rate per 1000 person-years?
A. 5
B. 10
C. 25
D. 50

Answer 3

Question 4

Which measure is best for identifying risk?
A. Odds ratio
B. Relative risk
C. Prevalence ratio
D. Sensitivity

Answer 4

Question 5

Relative risk = 2 means:
A. No association
B. Half the risk
C. Twice the risk
D. Risk cannot be calculated

Answer 5

Question 6

A study that follows individuals over time to see who develops the disease is:
A. Case-control
B. Cross-sectional
C. Cohort
D. Experimental

Answer 6

Question 7

The main advantage of a randomized controlled trial is:
A. Lower cost
B. Naturalistic observation
C. Control of confounders
D. Faster results

Answer 7

Question 8

In a case-control study, the measure of association is:
A. Risk ratio
B. Incidence rate
C. Odds ratio
D. Prevalence ratio

Answer 8

Question 9

Which study design is most prone to recall bias?
A. Cohort
B. Cross-sectional
C. Case-control
D. Randomized controlled trial

Answer 9

Question 10

A cross-sectional study provides:
A. Incidence rate
B. Temporal causality
C. Prevalence
D. Risk difference

Answer 10

Question 11

Sensitivity measures:
A. True negatives
B. True positives
C. False negatives
D. Predictive value

Answer 11

Question 12

Specificity measures:
A. Ability to detect disease
B. Proportion of true negatives
C. False positives
D. Overall accuracy

Answer 12

Question 13

A test has 90% sensitivity and 95% specificity. What does 90% sensitivity mean?
A. 90% of healthy people test negative
B. 90% of sick people test positive
C. 10% false positives
D. 5% false negatives

Answer 13

Question 14

Given:
True Positives = 80
False Negatives = 20

What is sensitivity?
A. 60%
B. 80%
C. 90%
D. 100%

Answer 14

Question 15

Given:
True Negatives = 95
False Positives = 5

What is specificity?
A. 85%
B. 90%
C. 95%
D. 98%

Answer 15

Question 16

Positive Predictive Value depends on:
A. Sensitivity only
B. Prevalence
C. Specificity only
D. None of the above

Answer 16

Question 17

Negative Predictive Value increases when:
A. Prevalence increases
B. Prevalence decreases
C. Specificity decreases
D. None of the above

Answer 17

Question 18

Likelihood Ratio (+) is calculated as:
A. Sensitivity / (1 − Specificity)
B. (1 − Sensitivity) / Specificity
C. Specificity / Sensitivity
D. Sensitivity × Specificity

Answer 18

Question 19

If a test has LR+ = 10, what does it indicate?
A. Not useful
B. Moderate usefulness
C. Strong rule-in
D. Strong rule-out

Answer 19

Question 20

A low LR− indicates:
A. Poor ruling-out capability
B. High false positive rate
C. Good for ruling out disease
D. High prevalence

Answer 20

Question 21

The highest level of evidence in EBM is:
A. Case series
B. Expert opinion
C. Randomized controlled trial (RCT)
D. Systematic review/meta-analysis

Answer 21

Question 22

The number needed to treat (NNT) is:
A. 1 / absolute risk reduction
B. 1 / relative risk
C. 1 / incidence
D. 1 / odds ratio

Answer 22

Question 23

NNT = 5 means:
A. 5 people need to be harmed
B. Treating 5 prevents 1 adverse outcome
C. 1 in 5 people benefit
D. 5% efficacy

Answer 23

Question 24

Which of the following is NOT part of the 5 A’s of EBM?
A. Ask
B. Acquire
C. Assess
D. Administer

Answer 24

Question 25

PICO stands for: A. Patient, Investigation, Control, Outcome B. Patient, Intervention, Comparator, Outcome C. Population, Index, Control, Outcome D. Problem, Intervention, Complication, Outcome

Answer 25

Question 26

Selection bias can be minimized by: A. Matching B. Blinding C. Randomization D. Stratification

Answer 26

Question 27

Information bias occurs due to: A. Incomplete follow-up B. Incorrect measurement or classification C. Too small sample size D. Random error

Answer 27

Question 28

Confounding is best controlled by: A. Increasing power B. Larger sample size C. Randomization D. Using one group only

Answer 28

Question 29

Loss to follow-up introduces: A. Recall bias B. Attrition bias C. Selection bias D. Lead time bias

Answer 29

Question 30

Observer bias can be minimized by: A. Open-label study B. Cross-sectional design C. Blinding D. Stratification

Answer 30

Question 31

A p-value < 0.05 means: A. Strong association B. 5% probability the result is due to chance C. Study is perfect D. No clinical significance

Answer 31

Question 32

Type I error is: A. Accepting a false null hypothesis B. Rejecting a false null hypothesis C. Rejecting a true null hypothesis D. Accepting a true null hypothesis

Answer 32

Question 33

Type II error is: A. Rejecting a true null hypothesis B. Accepting a true null hypothesis C. Accepting a false null hypothesis D. Rejecting a false null hypothesis

Answer 33

Question 34

Confidence interval (CI) of 95% means: A. 95% probability the null hypothesis is true B. Range where the true value lies 95% of the time C. 5% chance of error D. True effect is zero

Answer 34

Question 35

Power of a study is: A. Chance of making Type I error B. Probability of detecting a true effect C. 1 − alpha D. Degree of correlation

Answer 35

Question 36

A forest plot is used in: A. Case-control studies B. Meta-analysis C. Randomized controlled trials (RCTs) D. Cohort studies

Answer 36

Question 37

Heterogeneity in meta-analysis is measured by: A. Chi-square B. p-value C. I-squared D. Standard deviation

Answer 37

Question 38

Funnel plot is used to detect: A. Confounding B. Bias C. Publication bias D. Sample size

Answer 38

Question 39

Fixed-effect model assumes: A. Varying true effect sizes B. One true effect size C. Publication bias D. Heterogeneous effects

Answer 39

Question 40

Systematic review differs from narrative review in: A. Objectivity and method B. More references C. Shorter length D. Expert opinion

Answer 40

Question 41

Absolute risk reduction is: A. Control event rate – experimental event rate B. Relative risk × 100 C. Odds ratio – 1 D. Incidence × prevalence

Answer 41

Question 42

Relative risk reduction is: A. ARR / CER B. (CER − EER) / CER C. OR / RR D. RR − OR

Answer 42

Question 43

A good screening test should be: A. Highly specific B. Low sensitivity C. High sensitivity D. High false positive rate

Answer 43

Question 44

Lead time bias results in: A. Worse outcomes B. Apparent longer survival without real benefit C. Overdiagnosis D. Better outcomes

Answer 44

Question 45

Overdiagnosis bias means: A. Early diagnosis leads to better cure B. Diagnosing more mild or non-progressive disease C. Disease disappears naturally D. Increased specificity

Answer 45

Question 46

A diagnostic test has positive predictive value (PPV) of 20%. What can you infer? A. Good for ruling in B. High prevalence C. Many false positives D. Low specificity

Answer 46

Question 47

A new treatment has relative risk (RR) = 0.5. This means: A. Doubles the risk B. Halves the risk C. No effect D. Harmful

Answer 47

Question 48

In a 2×2 table, Odds Ratio is calculated as: A. a / b B. a / c C. (a × d) / (b × c) D. (a × b) / (c × d)

Answer 48

Question 49

In hypothesis testing, alpha is: A. Power B. Type II error C. Significance level D. Confidence interval

Answer 49

Question 50

A high number needed to treat (NNT) indicates: A. Highly effective intervention B. Weak treatment effect C. High prevalence D. High specificity

Answer 50