DIT Medicine in Society

Question 1

Case control study? What is the outcome of it?

Answer 1

compares group of people with disease to a group without it. look for prior exposure.

RETROSPECTIVE

OBSERVATIONAL

Outcome is odds ratio.

Question 2

Cohort studies? What is outcome? What is outcome?

Answer 2

A cohort is a group of people with similar characteristics.

You follow people with exposures or characteristics and see how ti goes.

OBSERVAITONAL
Prospective or Retrospective

Relative Risk is outcome. How many times greater risk of bad outcome is for those exposed

Question 3

Cross sectional study? What is outcome?

Answer 3

Data of people at that point of time.

Strictly observational

Prevalence. no causality can happen.

Question 4

Twin study vs adoption study?

Answer 4

Twin compares monozytotic to dizygotic.

Adoption is to compare heritability vs environmental

Question 5

Clinical trial phases?

Answer 5

1. Is drug safe?
2. does drug work?
3. Does drug work as well or better? (often in medical journal)
4. Can it stay? (post marketing surveillance)

Question 6

What is meta analysis?

Answer 6

Combining studies to strengthen statistical significance. Pool the studies to gather to get inclusion criteria.

Question 7

Make the two by 2 table for diagnostic statistics.

Answer 7

Diseasse?
              \+      -
Test +    TP   FP
        -     FN  TN
"Disease always comes to on top"

Question 8

Sensitivity?

Answer 8

Don’t miss anything. Sensitive. If you have it, you test positive.

TP/(TP+FN)

Question 9

Specificity?

Answer 9

If you test negative, you don’t have it. Didn’t give a false positive. OPPOSITE OF sensitifvity

TN/(TN+FP)

Vertical calculation in the table (as is sensitivity)

Question 10

Positive predictive value?

Answer 10

Proportion that a tested positive is true positive. So divide out the false positives+true positive

TP/(TP+FP)

Question 11

negative predictive value?

Answer 11

Means you don’t have false positives.

TN/(TN+FN)

Horizontal tests

Question 12

How does a higher prevalence impact positive predictive value?

What about negative predictive value?

Answer 12

higher prevalence increases positive predictive value, lowers the negative

(makes sense, think about if nobody has a disease, do you believe it?)

Question 13

What is incidence X average disease duration?

Answer 13

It is about the prevalence

Question 14

Odds ratio?

Answer 14

the name. RATIO OF ODDS.
odds of people with disease was exposed over risk VS odds that the group without the disease was exposed.

(A/C)/(B/D) or (A/B)/(C/D)

Question 15

Relative risk?

Answer 15

C/(C+D)

If low prevalence, the two are about equal

Question 16

Attributable risk?

Answer 16

AR is incidence of disease in exposed - incidence of unexposed

Easy. 30% in exposed - 5% unexposed means attributable risk is 25%

Question 17

Absolute risk reduction?

Answer 17

Easy. It is seeing if intervention decreases incidence by how much

Vaccinated 5% risk and unvaccinated is 25%? then ARR is 20%

Question 18

Number needed to treat?

Answer 18

1/Absolute risk reduction

Question 19

Number needed to harm?

Answer 19

1/Attributable risk

Question 20

What is Berkson bias?

Answer 20

a type of selection bias of only looking at inpatients

Question 21

What is a crossover study?

Answer 21

Patients are their own controls. For a period of time each patient gets either placebo or treatment

Question 22

Type I error vs type II?

Answer 22

Type I is alpha, and alpha is the chance f there being alpha error (so false positive)

Type II is the other other, so you accept the null hypothesis when there was a difference you didn’t see. False negative

Question 23

What is the alpha value in statistics?

Answer 23

Alpha is the level set by investigator in which P value is judged. Usually 0.05

Reject null if below alpha

Question 24

What is power?

Answer 24

a

Question 25

Standard Error of Mean?

Answer 25

Standard deviation/square root n

Question 26

Confidence Interval?

Answer 26

Mean +/- Z SEM

Z:
1.6 (90%)
2 (95%)
2.6 (99%)

Question 27

what is correlation coefficient interpreted?

Answer 27

R of 1 is perfect correlation
R of zero is no relationship

R<0 is negative correlation

Question 28

three types of disease prevention?

Answer 28

Primary: prevent
Secondary: Screen
Tertiary: Treat

PST and PST

Question 29

Kubler Ross stages of grief?

Answer 29

DABGA

Denial
Anger
Bargaining
Grieving
Acceptance