Biostats

Question 1

Incidence vs Prevalence?

Answer 1

new cases in given time/(total population NOT infected)

P = Incidence * Duration of disease

Question 2

Prevalence study called?

Answer 2

Cross-sectional study

Question 3

Effect of INCREASED disease prevalence on PPV and NPV?

Answer 3

PPV increases; NPV decreases

Question 4

Only study that can determine incidence?

Answer 4

Cohort study

Question 5

Specificity vs Sensitivity?

Answer 5

Of all people with a disease, fraction who test positive

Of all people without disease, fraction who test negative

a/(a+c) vs d/(b+d)

Question 6

PPV vs NPV

Answer 6

Of all people who test positive, fraction who have disease

Of all people who test negative, fraction without disease /(a+b) vs d/(c+d)

Question 7

+LR vs -LR?

Answer 7

probability that a patient with a + test actually has the disease

Probabiliy that a patient with a - test, actually doesn’t have the disease

sen/(1-specif) vs (1-sen)/specif

Question 8

RR vs OR?

Answer 8

a/(a+b)

Question 9

Attributable risk percent vs Relative Risk

Atrributable risk percent?

Answer 9

incidence in exposed - incidence in unexposed

(incidence in exposed)/(incidence in unexposed)

(RR-1)/RR

Question 10

Odds ratio?

Answer 10

(odds that a diseased person was exposed)/(odds that a non-disease person is (exposed)

(a/c)/(b/d) = ad/bc

Question 11

Outcome of a case-control study?

Answer 11

odds ratio

Question 12

Type of bias:

1. sample population is different from actual population?
2. different measuring methods
3. additional variables
4. difference in retrospection
5. Prolonged survival with early detection
6. Test detects slow varients of disease

Answer 12

1. selection
2. measurement
3. Confounding
4. recall
5. Lead-time
6. Length time

Question 13

Type I vs Type II error (each is related with?)

Answer 13

Assuming there is a difference when none exists (p-value)

Assuming there is no difference when there is one (Power = 1 - beta)

Question 14

Goals of primary, secondary and tertiary prevention

Answer 14

Prevention

Early detection

Survial

Question 15

Hazard Ratio

Answer 15

If 1, more likely to occur in treatment group

Question 16

Test reliability vs accuracy

Answer 16

Get same value if repeat test

vs

close to gold standard

Question 17

Case control vs retrospective cohort studies?

Answer 17

Determine outcome then assess risk factors

vs

Determine risk factors and assess outcome

Question 18

Selection Biases

Answer 18

#Attrition (lose people)
#Prevalence (forgot about the dead people)
#Berkson (only hospital patients - not applicable)
#non-response (people dont respond to surveys)
#Ascertainment (nonrandom selection)

Question 19

Observational Biases

Answer 19

#Recall - sick people more willing to report exposures
#Observer - preconceived expectations
#Reporting - subjects over/under report due to social stigma
#Surveillance/Detection - risk factor itself causes increased monitoring

Question 20

Use of test:

1. Chi-squared
2. Two-sample t/z test
3. ANOVA

Answer 20

1. compare proportions
2. compare two means
3. compare 3+ means

Question 21

Hawthorne effect?

Answer 21

People change behavior because they are being observed

Question 22

Study Designs:

1. Cluster
2. Cross-over
3. Factorial
4. Parallel-group

Answer 22

1. randomization at level of groups (not individuals)
2. 2 treatment groups, which switch at a point
3. 2+ interventions further divided into sub-groups
4. 2 randomized groups (placebo vs treatment)