|
Characteristic | Randomized controlled trials (RCT) | Real-world study (RCT) | Cohort study | Case-control study |
|
Purpose | Focused on efficacy | Diverse research purposes, including efficacy studies | Test etiological hypotheses, evaluate preventive effects, and study the natural history of disease | Explore the cause of disease |
|
Study population | Ideal world crowd and strict standards of inclusion and exclusion | Real-world population and broader inclusion and exclusion criteria | Exposure to study factors and control groups | People with and without the disease |
|
Sample size | Calculated according to statistical formula, the sample size is limited | Calculated based on the real world or statistical formula, the sample size can be large or small | Calculated according to statistical formula, the sample size is required to be large | Calculated based on the study design or statistical formula, the sample size is small |
|
Research time | Shorter (mostly end with the assessment of outcome) | Short term or long term (to obtain all treatments and long-term clinical outcomes as endpoints) | Longer research time, long follow-up time | Shorter (depends on the purpose) |
|
Outcome | Internal validity | High external validity | Internal authenticity is not as good as RCT, and external authenticity is not as good as RWS | Both internal and external authenticity are deficient |
|
Design | Randomization, control, prospective study | Random or nonrandom sampling, prospective or retrospective study | Prospective or retrospective or ambispective cohort study | Control, contrast |
|
Implementation scenario | Highly standardized environment | Medical institutions, communities, homes, etc. | Medical institutions and communities | Medical institutions and communities |
|
Data | Standardization, strict specification of the collection process | Diverse sources and high heterogeneity | Different confounding factors and bias | Different confounding factors and bias, heterogeneity |
|