Observational Studies: Cohort vs Case Control Studies

templeim Clinical Reasoning and Biostatistics , , , , , ,

Thank you to Dr Axelrod and Dr Horwitz for two amazing lectures recently!

For Dr Axelrod’s lecture:

Here are the highlights from Dr Horwitz’s lecture:

Standard research study design with associated biases
Standard research study design with associated biases

 

All of the above biases are present in observational studies, but eliminated in RCT

Cohort vs Case Control Studies

Cohort

Risk ratio: Inc dz / everyone exposud vs (Inc of CA among /not exposed )

Case control

Case control easier when incidence is low, and when there is a long time bw exposure and development or dx of illness

Odds Ratio: Antecedent exposure / diseased pts vs antecedent exposure / non diseased (control)

Confounding variables

  • Exposure A is related to another exposure X, which may affect incidence of dz
  • Intermediate confounding: Bias introduced in the intermediate between exposure and outcome state
  • Loss to follow up

Types of Research

Change (Causal) Research

  • Harm: etiologic or risk factor (RR or OR)
  • Diagnostic accuracy
  • Prognosis
  • Therapeutic: Real world evidence vs RCT

Conformity / Consistency Research

  • Quality of Care
  • Inter- and Intra- Observer Variation