II. Types: Randomized Study
- Randomized Clinical Trial
- In randomized trials, participants are assigned by chance to different groups
- Researchers and participants are typically "blinded" to the assigned group
- Randomized Controlled Trial (RCT)
- In randomized controlled trials, a control group is compared to one or more treatment groups
- As with other randomized trials, researchers and participants are typically "blinded" to group assignment
- Randomized Controlled Trials, when performed well, are the most reliable level of Research Study
III. Types: Observational Studies
- Observational Study
- Individuals are observed or specific outcomes are measured
- No intervention is performed to attempt to alter the outcome of those studied
- Cohort study
- Cross-section or population is selected (e.g. nurses)
- Group selection is based on specific characteristics that researchers intend to study (e.g. nurses who smoke versus non-smokers)
- Study participants are followed, evaluating specific outcome measures at specific time points (e.g. cancer at 5, 10, 15 and 20 years)
- Cohort studies may be retrospective (e.g. historical database reviews) or prospective
- Risk of confounding factors that distort results
- Case-control study
- Two groups of participants are studied: One group with a disease, and the other without
- The 2 groups are otherwise similar (age, gender, socioeconomic factors)
- Each group is interviewed with medical medical review and other analysis
- Researchers identify factors that distinguish the 2 groups and may identify causative factors for those with the disease
- Two groups of participants are studied: One group with a disease, and the other without
- Cross Sectional Study
- A specific, random cohort of participants (e.g. 20 year old men) are studied (interviewed, surveyed) at single point in time
- Qualitative Study
- Research focuses on nonnumeric data to investigate individual or group characteristics
- Describes experience and behaviors, Perceptions and beliefs with language
IV. Types: Additional Statistical Analysis Tools
- Intention to Treat Analysis (ITT)
- Statistical analysis of study participants based on the original, randomized treatment group assignment
- Includes patients regardless of the treatment they ultimately received (or did not receive)
- ITT reduces the study bias that comes from excluding patients who are not fully compliant with study protocol
- Propensity Score matching
- Used by observational studies to simulate a Randomized Control Trial
- Randomization is simulated by applying statistical techniques to divide a large study cohort into groups
- Goal is for groups to be similar in all identified characteristics OTHER than those factors being studied
- Propensity Score is the probability of group assignment based on baseline characteristics
- Run-in Period
- Study technique in which all participants are given a Placebo after study inclusion but BEFORE randomization
- Identifies participants most suitable as study participants
- Intended to exclude certain participants (e.g. Placebo responders, noncompliance)
- Run-in Periods are intended to increase a study's power, more likely to detect a statistical difference
- Used more commonly in industry sponsored studies
- Risk of decreased external validity, and risks participant unblinding
- Study technique in which all participants are given a Placebo after study inclusion but BEFORE randomization
V. Resources
- What types of Studies are there? (InformedHealth.org)