Interpreting medical literature

Question 0

Follow Up

Answer 0

Pick a group without outcome, watch their behavior over time and watch for outcome

Question 1

Case Control Design

Answer 1

Case: has outcome, Control: does not have outcome

Ask both groups about their past factors, and compare accordingly

Question 2

Cross-Sectional design

Answer 2

Start with a large population, ask about outcome and factors in the immediate past (“slice in time”)

Question 3

Retrospective study AKA

Answer 3

Case-Control

Question 4

Cohort AKA

Answer 4

Follow-Up

Question 5

Prospective AKA

Answer 5

Follow-Up

Question 6

Retrospective study timeframe

Answer 6

Begin and end in the present, watch the past

Question 7

Prospective study timeframe

Answer 7

Begin in the present and march forward, collecting data about a population whose outcome lies in the future

Question 8

Retrospective follow-up AKA

Answer 8

Historical prospective, retrospective cohort

Question 9

Retrospective follow-up, defined

Answer 9

Behaves like a follow-up, but uses data carefully collected in the past. (think alcohol and BP example from book)

Question 10

diagnostic bias

Answer 10

patients with symptoms for the condition are diagnosed because of the recognized factor

Question 11

reporting bias

Answer 11

doctor more likely to report a condition if it is associated with the popular factor

Question 12

subject bias

Answer 12

people with diseases tend to have a distorted view of past events

Question 13

nested case-control

Answer 13

draw your cases from an already existing group participating in a long-term study, gives you data from the past and possibly lab samples

Question 14

voluntary response bias

Answer 14

subjects who think they have been exposed to a toxin are more likely to return a mailed in questionnaire

Question 15

ecologic study

Answer 15

using data on groups instead of individuals, ex: california cigarette sales vs heart disease prevalence

Question 16

surveillance bias

Answer 16

scrutinizing one group more than another, see infant abuse highlight

Question 17

Internally valid

Answer 17

Thin the confines of the study, results appear to be accurate, and the interpretation of the investigators is supported

Question 18

External validity

Answer 18

Do results generalize to the real world? Think of studies where a majority of subjects are excluded

Question 19

Convenience allocation

Answer 19

Put everyone on the north side of the room in group A

Question 20

validity in data collection

Answer 20

the degree to which a measurement reflects a true value

Question 21

reliability in data measurement

Answer 21

reproducibility of measurements

Question 22

social desirability bias

Answer 22

patients and subjects want to answer in a way they perceive to be “right”

Question 23

Hawthorne effect

Answer 23

subjects behave differently when they know they are being watched

Question 24

inference

Answer 24

generalization made about a large group based on data from a sample population

Question 25

type one error

Answer 25

rejecting a true null hypothesis

Question 26

type two error

Answer 26

accepting a false null hypothesis

Question 27

Sensitivity

Answer 27

Ability of a test to single out people who have the disease

Question 28

Specificity

Answer 28

Ability of a test to classify people who do not have the illness as negative

Question 29

Predictive value

Answer 29

Frequency with which a positive test actually signifies disease

Question 30

Efficiency

Answer 30

Overall ability of a test to classify patients correctly

Question 31

Likelihood ratio

Answer 31

Ratio of true positives to false positives