Biostat/Epi Flashcards Preview

Step > Biostat/Epi > Flashcards

Flashcards in Biostat/Epi Deck (20):
1

What is an attack rate?

# of people who contract and illness/ # of people at risk of contracting the illness
IE # of diarrhea cases after eating potato sald/ # of people that ate the potato salad

2

What is the median?

Value in the precise center of a dataset. If there is an even number of values, average the two middle values

3

How do you calculate relative risk?

risk of outcome in the exposed/risk of the outcome in the unexposed
So if 120 out of 400 become symptomatic in the exposed and 100 out of 300 in the unexposed:
(120/400)/(100/300)

4

What type of study measures exposure and outcome simultaneously at a particular point in time (ie snapshot study)

Cross-sectional study
ie. Occurence if a protein in pts with HTN vs those without HTN. Samples taken when pts visited their PCP for routine care

5

What is an important control measure to prevent rheumatic fever?

Early penicillin tx for a case suspicious of group A strep pharyngitis

6

What is the mode?

Most frequently observed data point
Tends to be resistant to outliers

7

What is odds ratio and how do you calculate?

Odds that an outcome occurs based on an exposure
OR = (event in exposed)(no event in unexposed)/ (no event in exposed)(even in unexposed)

8

Encouraging pt to increase control over their health (ie, exercise more even though you dont have HTN or lipidemia yet) is?

Health promotion
Primary prevention
Improve diet, exercise regularly, smoking cessation, lose weight if needed.

9

What is secondary prevention?

Interrupting dz process before pt becomes symptomatic (ie cancer screening)

10

What is tertiary prevention?

Treating an established condition

11

Precision of a test is also referred to as its?

Reliability
Reliable = reproducible
NOT the same as accuracy which is a measure of correctness

12

What is the purpose of matching patients in case control studies?

Control confounding
Matching variables (age race habits) to control confounders. Cases and controls are then selected based on the matching variables so that both groups have a similar distribution in accordance with the variables.

13

How do you calculate sensitivity?

TP/(TP+FN)
TP/all the people that have the dz

14

How do you calculate specificity?

TN/(TN+FP)
TN/ all the people that don't actually have the dz

15

How do you calculate relative risk?

Risk in group 1/ Risk in group 2

16

How do you calculate number needed to harm?

1/attributable risk
Attributable risk = the adverse event rates in the treatment and placebo groups
(60/80) - (38/76) 25%
NNH = 1/25

17

What is type 1 error?

Falsely concluding that there is a difference
alpha

18

What is a type 2 error?

Falsely concluding that there is no difference
Beta error
1-B = power, strengthen with sample size

19

How do you calculate negative predictive value?

TN/TN + FN
TN/all negative tests

20

To ensure that investigators will not miss a difference between drug B and TAU, which value should be maximized?

Power (1-B)