Study Designs Pt 2: Others EXAM 2 Flashcards
What is Latin Square Study Design?
-Cross-over study but more complex
-f.e. Drug A vs Drug B vs Drug C vs Drug D
Why can the Latin Square Study Design be problematic?
Sequence effect
-we want to have as many patients to receive Drug A, Drug B, C or D first
-randomize the order, such that every order is represented in a balanced fashion
-ARE THE SAME PATIENTS USED FOR EACH ROUND (A, B, C, D)
(B, C, A, D)
???
How is a Latin Square Study designed?
-each group receives each treatment, in random order
-each treatment is given first to one group, second to one group, third to one group,…
-The sequence of administration can be randomized
-washing period between each cross-over
What is a Factorial Design?
-Design whereby several factors are examined simultaneously
-Factors are the drugs or the control (placebo)
-Factors can have different levels (doses, duration)
(the same drug can be examined on different doses or durations)
What does an RTC have in order to be considered a Factorial Design?
2 or more interventions are evaluated separately + in combination and against a control
Example of Factorial Design
Hypertension
Examine the effect of Medication A and Medication B on Hypertension
Examine the influence of Diet on the efficacy of Med A and B
Examine the influence of Exercise on the efficacy of Med A and B
-with a 2x2 factorial design (table) it would be more efficient and cheaper to conduct
Characteristics of a Factorial Design
-2 factors (or more) are examined at the same time
-the factors are examined separately and against the control
f.e. how is factor A different from factor B, and how are they individually different from the control
-evaluate the efficacy of 2 therapies: multiple research questions answered in the same experiment
Why is the Factorail Design so efficient?
-Because multiple scenarios can be tested at the same time
-Therapy A vs Therapy B = Test for Synergy
-Therapy A vs Placebo B
-Therapy B vs Placebo A
-Placebo A vs Placebo B
Advantages of Factorial Design
-Evaluation of multiple factors and in combination with each other
-not limited to one aspect (f.e. simply Drug A vs Drug B)
-reduces the number of patients needed compared to parallel-group study
How can a Factorial design test for multiple variables?
Example:
Determine the effect of 3 doses on completing 2 different tasks, in young and old
-Age (2 variables), Dosage (3 variables), Type of Task (2 variables)
2x3x2 = 12 scenarios
What are Noninferiority Trials?
-Trying to show that a drug is NOT WORSE than an active control by more than the equivalence margin
(f.e. efficacy of 80-90%)
-if the treatment is lower than the equivalence margin, then the treatment is inferior (worse than the active control)
-if within the equivalence margin, the treatment is non-inferior
When are Noninferiority Trials used?
-Placebo-controlled superiority trials are not always ethical (must use standard care of treatment if available, f.e in cancer trials)
-Active-controlled superiority is not always possible due to the large number of patients
So if placebo-control cant be used and active-control is not possible then at least is tried to show that the new drug is non-inferior
Example of Noninferiority Trial
-New drug to treat otitis media
-Standard treatment is Augmentin
-f.e. Augmentin works by 90%
-The equivalent margin could be 80-90%
-if the new drug is below the margin it is considered INFERIOR, within the margin = NONINFERIOR
