Spotting the Study Design
We can work the type of study by looking at three issues:
Will you “assign” exposures?
If you assign exposure—which might be an intervention, like an antidepressant, a placebo, or a therapy, for example—you are doing an experimental (intervention before observation) study.
Observational (no intervention involved)
Will there be a control group? Controlled study.
Will there be an active comparator or placebo?
Some trials use an active comparator, and we call it active comparator trial. if you want to you a placebo, your study design is a placebo-control trial.
will the allocation be random?
yes > randomised
no > non-randomised/quasi-experimental
Will the allocation be swapped from time to time? Yes -> cross-over, No -> parallel
Who will know about the allocation of the treatment?
Everyone -> open-label
All except patients > single-blinded
Neither patients nor assessors -> double-blinded
Observational study designs
-> will there be a comparison group?
Yes -> analytical (if you want to test a hypothesis, choose this one)
No -> descriptive study
> Analytical study designs
Want to determine the outcome of exposure -> cohort (begins with exposure)
Exposure has occurred in the past and is tracked down till date -> retrospective cohort
Exposure will occur from now onward and will be followed over time -> prospective cohort
Want to determine the (causes of/factors-affecting the) outcome -> case-control (begins with the outcome/cases)
Want to determine/identify certain factors in a population at an instance? -> Cross-sectional surveys
comparison group present? -> analytical, otherwise observational
Want to determine the prevalence of depression in teachers? -> cross-sectional survey, observational
Want to determine whether physical punishment will lead to poor academic performance in students?
-> either retrospective cohort (punishment received traced till date, better choice, beware bias (students with poor academics more likely to be punished) or case control (academically poor students vs talented students, beware of bias). Don’t choose a prospective cohort, we can not let children be exposed to punishment while conducting research.
Want to test new drug but do not want to harm patients, choose active comparator.
Want to determine the incidence of disease in a particular group, choose prospective cohort. One sample from that group and another comparison group.
Want to determine the risk of cancer in smokers? Choose cohort.
(Point) Prevalence of depression in diabetes? Cross-sectional
10-year prevalence of depression in diabetes? Prospective cohort.
Factors associated with depression in diabetes? Case-control.
CBT versus clinical management? Open label.
ECT efficacy, bright light therapy? Sham comparator.
We should finally note that studies can incorporate several design elements. For example, the control arm of a randomised trial may also be used as a cohort study; and we may use the baseline measures of a cohort study as a cross-sectional study.
To simply describe a population (PO questions)
To quantify the relationship between factors (PICO questions)
was the intervention randomly allocated?
For observational study the main types will then depend on the timing of the measurement of outcome, so our third question is:
Q3. When were the outcomes determined?
Some time after the exposure or intervention?
cohort study (‘prospective study’)
At the same time as the exposure or intervention?
cross sectional study or survey
Before the exposure was determined?
case-control study (‘retrospective study’ based on recall of the exposure)
For example, if you want to determine the prevalence of depression in teachers, it should be descriptive, observational, cross-sectional study design.
Post a Comment