Study Designs Overview and Concepts

Study Notes

Study design refers to the approach used in a study.
It considers whether the study is well-designed, and if so, the presence of potential errors like random error, bias, and confounding.
It also specifies the methodological approach used in the study.

Study Purpose: Different types of studies, e.g., descriptive and analytical.
Time Orientation: Prospective, starting in the present and going forward, or retrospective, starting and ending in the present but looking back in time.
Investigator Orientation:
- Experimental trials: Researchers impose an intervention.
- Quasi-experimental trials: Researchers compare groups but do not fully control the intervention.
- Observational trials: Researchers observe and record data without manipulating any variables.

A case series is a collection of cases with a similar exposure.
Clinicians often use case series for quantifying the incidence of a reaction after a drug launch.
Case series are for describing a disease or patient characteristics and providing possible causes, but is not useful for causation.

These studies examine trends in exposures (presumed causes) and outcomes (presumed effects) over time.
They can be regional or global.
Correlation analysis suggests a relationship but does not prove causation.

Case-control studies compare groups with and without a disease.
They look back to find differences in exposures.
Useful for studying rare diseases.
They are often retrospective.
Example: Women with venous thromboembolism versus those without it, looking back at different contraceptiv.
Critical assumptions: Representative cases and controls, similar information gathering.

Cohort studies follow groups of people sharing a common characteristic.
They track outcomes/exposures over time.
Prospective cohort studies recruit groups according to exposure and follow them over time, while retrospective cohort studies start with outcomes and work backward.
Useful for studying the natural history of disease and exposures.

Researchers assemble a cohort on the basis of exposure (e.g., exposure to a certain drug)
They follow the cohort over time to observe the effect of the exposure on the outcome (e.g., the rate of side-effects).

Researchers start with a group of people who already have an outcome (e.g., a specific disease)
They look back to see if they were exposed to a particular factor (e.g., a medication) at some point in time.

Examine relationships between different factors (exposures and outcomes) at one point in time,
Measures all variables for each study subject at a defined moment in time.
It's descriptive rather than analytical.
Useful for prevalence studies and exploring relationships in relatively short periods.

RCTs are experimental studies in which researchers manipulate the exposure (intervention).
Participants are randomly allocated to either an intervention or a control group.
They are considered the gold standard for testing causality.
Vital to determining whether an intervention causes an outcome, particularly with new treatments.
Strengths: Random assignment, minimizing confounding influences.
Critical assumptions to minimise bias:
- Representitive cases
- Comparable controls without the disease
- Same information collection method.
Blinding techniques:
- Single-blind: Either the participants or the researchers are unaware of who is in which group.
- Double-blind: Neither the participants nor the researchers are aware of who is in which group.
- Triple-blind: The participants, researchers, and analysts are unaware of who is in which group.

Sample size determination for RCTs requires planning.
Essential for calculating the minimum number of participants needed to achieve sufficient power.
Power is the ability of a study to correctly detect a true effect if it exists.