experimental study design Flashcards
clinical trials
involve a human or more, not observational instead require a manipulating intervention
NIH clinical trial criteria
- involves human participants, 2. prospectively assigns participants an intervention, 3. evaluates an interventions effect, 4. measures health related biomedical or behavioral outcome
phases of clinical trials
preclinical: lab studies on cells and animals
phase I: small scale, a focus on safety and limiting dose toxicity
phase II: medium scale, tests efficacy and safety
phase III: large scale, randomized trials comparing standard treatment (necessary for FDA approval)
phase IV: post approval, monitoring for safety and side effects
randomization
eliminates selection bias and ensures comparability between groups
blinding
can be single or double, prevents bias in treatments expectations and assessments
SMART trials
modify interventions based on interim results
crossover designs
participants switch interventions after a washout period
adaptive randomization
adjusting treatment assignment based on early data
type I error
false positive - claims treatment worked when it did not
generalizability
ensuring results apply beyond study population
type II error
false negative - fails to detect a real effect
volunteer bias
participants may not represent the general population
complex logistics
recruitment, compliance and cost challenges