Biostats - Research Design

Front 1

an informal study in which the experience of a group of patients or only one patient is described

Back 1

case report or case series

Front 2

studies which use groups of individuals as the unite of analysis

Back 2

ecological studes

Advantages: quick, helop with hypothesis
Disadvantages: poorly suited for causal inference. thos exposed may not be those with the disease

Front 3

two or mjore assessments are made simultaneously on the subjects in this type of study

Back 3

cross-sectional

Front 4

this study may use descriptive statistics such as age-specific disease rates that summarize the disease burden

Back 4

ecological studies

Front 5

groups selected on whether they have experienced the outcome

Back 5

case-control study

Front 6

an odds ratio greater than one indicates

Back 6

exposure is associated with increase risk of disease

Front 7

odds ratio of 1

Back 7

means no association between exposure and risk of disease

Front 8

odds ratio less than 1

Back 8

means exposure is associated with a decrease risk of disease

Front 9

study where it is imposible to compute disease risks in exposed group and unexposed goup - oods ratio is used instead

Back 9

case-control study

subjects are selected according to whether or not they have experienced the outcome or not. you cant compute the disease risks because the numbers of cases and controls have been fixed by the study design and not by nature.

Front 10

a case control study which includes all cases occurring in a defined group during a certain time period

Back 10

population based

Front 11

in clinic based case-contro studies why might the data be distorted

Back 11

patients may be reffered from a differnt population - so the computed relative risks may be distorted.

Front 12

When case and control groups are inappropriately matched

Back 12

selection bias

Front 13

formal follow up of individuals over time

Back 13

cohort (follow up) study

Front 14

a cohort study where all outcomes in the final analysis would have occured before the study even started

Back 14

retrospective cohort study

Front 15

a cohort study where cohort would be directly observed over time

Back 15

prospective cohort study

Front 16

the ascertainment of the endpoints is particularly inportant in what study

Back 16

cohort studies

Front 17

what are the study measurements in a cohort vs. case-control study

Back 17

cohort : disease outcome
Case-control: exposure history

Front 18

can you measure the risks in a cohort or a case-control study

Back 18

you can measure risks in a cohort but not a case control study

Front 19

what are the three ways you can measure relative risk in a cohort study

Back 19

risk ratio, rate ratio, odds ratio

Front 20

how do you measure relative risk in a case control study

Back 20

odds ratio.

Front 21

a cohort study in which athe investigator determines the exposure

Back 21

clinical trial

Front 22

helps ensure that the interventions under study are being tested on similar patients

Back 22

randomization

in small samples may still be imbalanced but in large samples imbalance is unusual

Front 23

blinding

Back 23

when investigators are kept ignorant of which patients are on which treatments.

Front 24

if the investigators or patients (but not both) in a study are kept ignorant of treatment assignments, the study is called

Back 24

single blind

Front 25

if if both investigator and pt is unaware of of assignments the study is called

Back 25

double blind

Front 26

historical controls

Back 26

the control group is a group of people who in the past have been treated a particular way -- they are compared with the group getting the new treatment.
poor evaluation method - groups of patients will obviously differ quite a bit

Front 27

in what study might patient outcome affect measurement of the exposure in individual subjects

Back 27

case-control

Front 28

in what tho type of studies involve follow-up of subjects over time

Back 28

cohort and randomized clinical

Front 29

in what two studies can the risk or rate of an outcome for an individual be calculated according to exposure status

Back 29

cohort and randomized clinical

Front 30

what type of study controlls for confouding affects

Back 30

randomized

Front 31

the final analysis which includes a comparison based on randomization assignments regardless of what treatment the patient actually received

Back 31

intent-to-treat analysis

includes thos who have crossed over between study treatments

Front 32

an analysis of a randomized experiment based on subjects who actually receive and comply with assigned treatments

Back 32

efficacy analysis - not based on the entire randomly-created groups, because of patients who drop-out or cross-over between study treatments

Front 33

when the results of a study can be applied to a general population, we say that the study results are

Back 33

generalizable or externally valid

Front 34

refers to the correctness of the study conclusions regarding the particular study population

Back 34

internal validity

Front 35

a descriptive study

Back 35

involves no comparisons

Front 36

if there are comparisons but the investigator only watched what would have happened anyway the study is called a

Back 36

analytic observational study

Front 37

if there are comparisons and the investigator determined the treatment of the subjects, then the study is a

Back 37

clinical trial (intervention study)

Front 38

if diseased and non-diseased groups are separately chosen for the study then what kind of observational study is it

Back 38

case-control

Front 39

if subjects with known exposure are followed up over time then what kind of observational study is it

Back 39

cohort

Front 40

if the unit of analysis is groups of people, and individuals are not questioned or measured, then what kind of observational study is it

Back 40

ecological

Front 41

if one group is investigated and disease and exposure status are determined at the same time what kind of observational study is it

Back 41

cross-sectional

Front 42

what is the concern with a descriptive study

Back 42

precision and bias

Front 43

what is the concern with a case-control study

Back 43

selection bias (do cases and controls correspond?

information bias (did the presence/absence of disease affect the data obtained)

Front 44

what is a concern with a cohort study

Back 44

follow up time (where exposed and unexposed groups followed and endpoints determined with equal intensity and w/o bias)

Front 45

what is the concern with clinical trial

Back 45

what kind of controls? bias?

Front 46

a dose-response relationship

Back 46

if a gradient of risk can be demonstrated such that greater exposure causes higher disease risk, then causality is implied

Front 47

a repeated finding suggests the possibility of causality, particularly if the result is foundin studies of different designs performed in different population group

Back 47

consistent association

Front 48

a series of analyses that show an exposure to be associated with virually every disease is suspect

Back 48

specific association

Front 49

the ability of a study to apply to populations other than the one actually studied

Back 49

external validity - also called generalizability

Front 50

response bias is also known as

Back 50

informaiton bias

Front 51

odds ratio is used to evaluate risk in what kind of study

Back 51

case control