Flashcards in Epidemiology 7/8 Deck (17):

1

## what does the P value mean

### probability of getting a result as extreme or more extreme if the null hypothesis is true (i.e.. no difference)

2

## when do we reject null hypothesis

### P less than or equal to 0.05

3

## T/F: P values measure how likely study hypothesis is true

### FALSE

4

## What is a confidence interval

### a range of values that could be plausibly generated by the sample

5

## When you want to be more "confident" what occurs with the confidence interval?

### It gets wider so you can capture the true value more often

6

## define sensitivity

### how well does a test correctly identify those with a disease (TP/all with disease)

7

## specificity

### how well a test correctly identifies those WITHOUT a disease (TN/all without disease)

8

## PPV

### of those that test positive, what proportion actually have disease

9

## NPV

### of those who test negative, what proportion actually don't have the disease

10

## What happens to sensitivity/specificity/PPV/NPV when prevalence changes

###
Sensitivity/specificity don't change

-PPV increases with increasing prevalence, decreases with decreasing prevalence

-NPV decreases with increasing prevalence and increases with decreasing prevalence

11

## Receiver Operating characteristic

###
describes intrinsic accuracy of a test considering various threshold changes

- larger area under curve indicates greater discriminating ability of test

12

## number needed to treat/screen (NNT/NNS)

###
what is the impact of the treatment

- number of ppl you need to treat/screen to prevent 1 bad outcome

13

## What does a higher NNT mean

### increased number needed to treat = less effective test

14

## Case-control study

### select ppl with/without disease or outcome. Compared with controls regarding past exposures

15

## Cohort study

### select ppl with exposure/not but are initially free of the outcome. Followed and assess incidence of outcome in both exposure groups

16

## ODDS Ration

###
Odds of disease in exposed/odds of dz in non-exposed

(ad/bc)

17