Research Design Flashcards
What are the 6 types of research designs
randomized control trials (RCT) cohort studies case-control studies cross-sectional (prevalence) studies case reports/series systematic reviews/meta-analysis
what is a randomized control trial?
subjects are randomly assigned to different intervention groups, including control
groups are studied prospectively and a rigorous comparison of outcomes is done
what are RCTs used for?
to determine effects of treatment
how are RCTs useful? ie what are their advantages?
- use of two groups minimizes bias and confounding variables
- most powerful study type to establish cause and effect relationships
discuss randomization methods:
pure random selection
systematic randomization
stratified randomization
means that patients are randomly assigned to different groups to equalize effects of extraneous variables
- pure random selection = equal probability of anyone in a study being chosen
- systematic randomization = selecting in a systematic way (every third infant or every other woman)
- stratified randomization = divides groups into homogeneous subgroups first (male v female) and then proceeds with random sampling
discuss double blinding
RCT tests
- both researcher and patient are unaware of which group they are in
- minimizes bias, but not always possible
discuss intention to treat (ITT) analysis
RCT tests
- all patients are initially assigned to treatment group are analyzed in that group, regardless of whether they received their treatment
- preserves the value of randomization, better predicts outcomes in practice
discuss interim analysis
- used to limit patient risk by evaluating outcome data at points during the study period
- -> if it looks like a group is experiencing a comparative excess of negative outcomes, a trial may be terminated early
negative effects can be a dramatic benefit of one group or adverse effects of one group
describe the difference between parallel and cross over designs
parallel design: groups stick to their designated treatment
cross over: each group receives both with a washout period in between
what are some benefits of cross-over studies
- each group is its own control, minimizing effects of differences between groups
- increases functional study size
what are some weaknesses of cross-over studies
- introduces the potential for contamination or carry over effects between interventions
- more than doubles the time required
- only useful for short term outcomes
what are the weakness of RCTs
expensive, labor intensive
ethical or practical considerations may prohibit use (depriving treatment to a group)
what is a [prospective] cohort study
focuses on factors relating to development of a disease
-a group that does not have the disease of interest is selected and observed for an extended people of time noting any exposure to risk factors and observing for development of disease
what kind of study?
observing a group of women for development of breast cancer, documenting exposure to potential risk factors such as hormones, high fat diet, etc
prospective cohort study
what are the strengths of cohort studes
- prospective decreases biases
- can establish risk of disease
- strongly suggests a cause effect relationship
what are the weaknesses of cohort studies
- time consuming, expensive
- impractical for rare diseases
what is an inception cohort?
subjections followed starting soon after development of disease of interest
-used primarily to obtain information on prognosis
what is a historical cohort?
information regarding cohort is obtained from previously collected or existing historical data - sometimes called retrospective
what is a case-control study
- a group with a disease or outcome (case) is paired with a group without the disease or outcome (control)
- past exposure of both groups are investigated to identify factors which are associated with disease or control
what kind of study?
chart audit of patients admitted for stroke to investigate the association of smoking and stroke
case-control study
what are the strengths of case-control studies
- quick, cheap
- few ethical problems
- excellent for study of rare diseases
what are the weaknesses of case-control studies
- significant bias due to retrospective design
- cannot determine prevalence rates or risk of disease
- weak indicator of causality, only demonstrates association
what is a cross-sectional study
surveys a population for simultaneous presence of a disease and potential risk factors
-snapshots of a population, not longitudinal
what kind of study?
prevalence of diabetes in a community is compared to prevalence of diabetes in another community which differs in potential risk factors such as obesity and exercise levels
cross-sectional study
