Study Design - Case-control study, Cross-sectional study, Cohort study (Prospective, Retrospective)

Case series studies involve a single group of individuals, usually sharing a diagnosis, treatment, or outcome, without a comparison group. They are used to describe clinical signs and symptoms but cannot establish risk factors, correlation, or causation.

The video introduces various study designs in biostatistics, including case series, case-control, cross-sectional, cohort (prospective and retrospective), ecological, twin concordance, adoption studies, and randomized clinical trials. It highlights the fundamental rule of statistics: correlation is not causation, meaning two events happening together don't necessarily imply one caused the other.

Observational studies, such as case series, case-control, cross-sectional, and cohort studies, involve observing phenomena without intervention and can only establish correlation. Interventional studies, like randomized double-blinded placebo-controlled clinical trials, involve active intervention to establish causation. The video emphasizes that only interventional studies can prove cause-effect relationships.

Cross-sectional studies ask about events happening in the present to determine the prevalence of a disease or phenomenon. They can be descriptive (measuring prevalence) or analytical (examining associations with other characteristics). These are observational studies and thus cannot establish causation, only correlation.

Case-control studies start in the present and investigate past exposures to determine the odds ratio. For example, asking lung cancer patients if they smoked in the past. These are observational studies used to identify risk factors but not direct causes.

Prospective cohort studies begin in the present and follow groups into the future to see what will happen, establishing relative risk. An example is following smokers and non-smokers to see who develops lung cancer. These are observational studies focused on future outcomes related to present exposures.

A summary reiterates that cross-sectional studies determine prevalence (present), case-control studies determine odds ratio (past), and prospective cohort studies determine relative risk (future). All are observational and establish correlation, not causation. The video then clarifies the difference between case-control and retrospective cohort studies: case-control examines past risk factors, while retrospective cohort examines past diseases.

These studies aim to distinguish between genetic and environmental influences on diseases. Adoption studies compare biological and adopted children within the same environment to assess genetic factors. Twin concordance studies compare disease frequency in monozygotic (identical) and dizygotic (fraternal) twins; higher concordance in monozygotic twins suggests a stronger genetic component.

Ecological studies compare disease rates between different populations or geographical areas, often correlating them with environmental factors. For example, comparing COPD rates in cities with varying air pollution levels. These are observational and demonstrate correlation, not necessarily causation.

These are interventional studies designed to establish causation. Patients are randomly assigned to a placebo group or a group receiving a new drug (or an old drug vs. a new drug). 'Double-blinded' means neither the subjects nor the administrators know who receives which treatment, minimizing bias. This study design helps determine if a drug's effect is statistically and clinically significant. The concept of 'crossover' in trials is also explained, where groups switch treatments halfway through the study.