Practical / Calculations

Question 1

Cross-sectional Study (Observational Analytic)

Answer 1

-

Question 2

Cross-sectional Study (Observational Analytic)

Identification

Answer 2

• relationship/association (analytic)
• asked at some point in time (cross-sectional)

Question 3

Cross-sectional Study (Observational Analytic)

Measure of Association

Answer 3

• interpretation is odds

Question 4

Cohort Study (Observational Analytic)

Identfication

Answer 4

• relationship (analytic)
• followed up (cohort)
• risk of incident (cohort)

Question 5

Cohort Study (Observational Analytic)

Measure of Association

Answer 5

• can measure onset of disease – therefore interpretation is risk

Question 6

Case-control Study (Observational Analytic)

Answer 6

• starting point is outcome (disease) – identify individuals with the disease of interest
• identify the control – individuals without disease of interest
• look at exposure history (prior to having disease)

Question 7

When might case-control studies be used?

Answer 7

• when outcome (case) is rare
• when investigating an outbreak, epidemic, pandemic

Question 8

Case-control Study (Observational Analytic)

Identification

Answer 8

• relationship (analytic)
• with disease (case)
• without disease (control)

Question 9

Why are case-control studies not as precise as cohort studies?

Answer 9

because you are not following them to the outcome

Question 10

Odds Ratio Calculation (cross-sectional)

Answer 10

• odds of outcome among exposed / odds of outcome among unexposed
• (a/b) / (c/d)

Question 11

Odds Ratio Results

Answer 11

OR > 1:

• association – exposed group have higher odds of outcome compared to unexposed group
• % increase = (OR - 1) x 100

OR = 1

• no association

OR < 1

• inverse association – exposed group have lower odds of outcome compared to unexposed group
• % decrease = (1 - OR) x 100

Question 12

Relative Risk/Risk Ratio Calculation

Answer 12

• risk of outcome among exposed / risk of outcome among unexposed
• (a/(a+b)) / (c/(c+d))

Question 13

Relative Risk/Risk Ratio Results

Answer 13

RR > 1

• association – exposed group have higher risk of outcome than unexposed group
• % increase = (RR - 1) x 100

RR = 1

• no association

RR < 1

• inverse association – exposed group have lower risk of outcome than unexposed group
• % decrease = (1 - RR) x 100

Question 14

Odds Ratio Calculation (case-control)

Answer 14

• odds of exposure among cases / odds of exposure among controls
• (a/c) / (b/d)

Question 15

Randomized Controlled Trial (Experimental)

Measure of Association

Answer 15

• relative risk
• same calculation as cohort studies

Question 16

Prevalence/Pre-Test Probability (Pregnancy Test)

Answer 16

• all pregnant people / all ‘possibly’ pregnant people
• (TP + FN) / total people

Question 17

Sensitivity (Pregnancy Test)

Answer 17

probability that an individual with the condition/disease has a positive test result

TP / (TP + FN)

Question 18

Specificity (Pregnancy Test)

Answer 18

probability that an individual without the condition/disease has a negative test result

TN / (TN + FP)

Question 19

Positive Predictive Value (Pregnancy Test)

Answer 19

probability that an individual with a positive test result has the condition/disease

TP / (TP + FP)

• ie. what is the probability that an individual screened from this population having a positive test result actually has the disease

Question 20

Negative Predictive Value (Pregnancy Test)

Answer 20

probability that an individual with a negative test result does not have the condition/disease

TN / (TN + FN)

Question 21

Very Specific

Answer 21

all the real things that can cause the test to be positive

Question 22

Very Sensitive

Answer 22

all the other things that can cause the test to be positive

Question 23

Likelihood Ratio

Answer 23

• calculated from knowing sensitivity and specificity of the test
• combination of prevalence and sensitivity

Question 24

Positive Likelihood Ratio

Answer 24

• LR+ = sensitivity / (1-specificity)
• > 1

Question 25

Negative Likelihood Ratio

Answer 25

- LR+ = (1-sensitivity) / specificity - < 1

Question 26

Likelihood Ratio Results

Answer 26

- LR =1: NO CHANGE in the likelihood of the disease/condition - LR >: the MORE likely the disease/condition - LR <1: the LESS likely the disease/condition

Question 27

Prevalence vs. Incidence

Answer 27

- prevalence: 'burden' or 'living with' - incidence: 'newly' or 'risk'