test 1

Question 1

studies that document and communicate

Answer 1

descriptive studies

Question 2

studies that compare groups

Answer 2

explanatory studies

Question 3

randomization by authors does not occur it what type of study

Answer 3

observational

Question 4

strengths of observational studies

Answer 4

• study human exposure without inferring from animals

- no ethical conflicts

Question 5

weaknesses of observational studies

Answer 5

• exposures are not random

- natural experiments are usually not reproducible

Question 6

main source of bias in observational studies

Answer 6

confounding bias

Question 7

features of cohort study

Answer 7

• investigators don’t choose allocation, “natural”
• subjects identified by exposure
• forward in time
• compares to patients without exposure

Question 8

approach to doing cohort study

Answer 8

• identify population
• exclude individuals with outcome of interest
• classify by exposure
• follow forward in time
• compare exposed vs unexposed

Question 9

other names used for cohort study

Answer 9

follow-up (one to know)

prospective

Question 10

strengths of cohort studies

Answer 10

• best study design
• strongest causal link of observational designs
• you can calculate prevalence and incidence

Question 11

weaknesses of cohort studies

Answer 11

• can be expensive
• not as efficient for rare outcomes
• must account for variable follow up time

Question 12

defining features of case-control studies

Answer 12

• investigators don’t choose allocation, “natural”
• subjects identified by disease status
• direction is backward in time
• contrast between pts with vs without the outcome

Question 13

approach to case-control study

Answer 13

• identify pts with disease
• identify controls w/o disease
• go back in time to compare exposures

Question 14

strengths of case-control studies

Answer 14

• more efficient for studying rare outcomes

- can add new risk factors to evaluate at any time

Question 15

weaknesses of case-control studies

Answer 15

• no denominator

- selection of controls must be relevant

Question 16

cross-sectional defining features

Answer 16

• investigators don’t control treatment allocation

- classification of exposure and outcome are done at the same time without regard to temporal sequence

Question 17

4 key questions to determining study design

Answer 17

1. are there comparisons made
2. do investigators control allocation groups
3. are exposure and outcome measured at the same or different times
4. did investigators compare/contrast by exposure or outcome

Question 18

equation for risk ratio

Answer 18

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

Question 19

equation for rate ratio

Answer 19

(a/PY(e)) / (c/PY(u))

exposed
unexposed

Question 20

when should we adjust estimates

Answer 20

for any factors that vary between the groups and that affect the relationship the exposure and outcome

Question 21

selection bias

Answer 21

error in patient selection that makes the sample population not representative of the general population

Question 22

information bias

Answer 22

error in measurement of exposure or outcome that distorts the true estimate of effect

Question 23

confounding

Answer 23

error that results from misattribution of causal effects

Question 24

two broad types of error in studies

Answer 24

random

systemic

Question 25

random error is dealt with using

Answer 25

statistics

Question 26

systematic error is dealt with by

Answer 26

reducing bias

Question 27

error that diminishes as sample size increases

Answer 27

random

Question 28

error that does not diminish as sample size increases

Answer 28

systematic

Question 29

the 2 categories of information bias

Answer 29

differential

nondifferential

Question 30

examples of differential information bias

Answer 30

recall bias

diagnostic bias

Question 31

example of nondifferential information bias

Answer 31

if a person drinks red wine and smokes gets emphysema you could say red wine leads to emphysema, even though it doesn’t

Question 32

3 things needed to be a confounder

Answer 32

1. must be associated with exposure
2. must be a cause of the outcome
3. must not be on the causal pathway

Question 33

surrogate confounder

Answer 33

factor that serve as indicators of the true underlying confounding factor

Question 34

residual confounding

Answer 34

when confounder is identified and controlled for, but its not completely characterized by the study

Question 35

unmeasured confounding

Answer 35

when the factor is unmeasured or unknown

Question 36

recall bias

Answer 36

misclassification of the exposure due to faulty memory

Question 37

diagnostic bias

Answer 37

misclassification of the outcome due to differential diagnosis depending on exposure

Question 38

channeling bias

Answer 38

• confounding

- differences in disease severity lead to patients receiving different treatments

Question 39

study design methods to deal with confounding

Answer 39

• restriction
• stratification
• matching

Question 40

differential bias

Answer 40

error in classification is different between groups in a way that is associated with the study variable

Question 41

nondifferential bias

Answer 41

error in classification that is not different between groups in a way associated with outcome

Question 42

Common sources of confounding bias

Answer 42

• demographics (age, gender, insurance status, etc)
• disease severity
• comorbid conditions
• unmeasured confounders