Study Designs Pt1: Parallel Group and Cross over EXAM 2 Flashcards
What are two common designs using two or more groups of patients?
-Parallel group (non-crossover)
-Cross-over
What is a Parallel group design?
Comparing relative effects of drugs or treatments on 2 or more groups
Most common: Drug vs. placebo
-could also be Drug vs. placebo vs. standard of care
or
-Drug 1 VS Drug 2 VS Drug 3 VS Placebo
Why is the Parallel group design considered robust?
Because it is tolerant to many types of problems such as missing data, missing visits, etc.)
What are the advantages of a Parallel group design?
-easy to conduct and it takes less time to achieve objectives bc the study in other groups run simultaneously, whereas in cross-over they have to switch
-Patients are enrolled for a shorter time and only exposed to one drug (in cross-over there is only one group of patients receiving the drug and placebo - enrolled longer and exposed to two treatments)
-A more diverse set of patients (patients in placebo or drug group (cross-over has only one group)
-suitable for phase 1, 2, and 3 trials
Disadvantages of a Parallel group design
-possibility that the experimental and placebo group is different - but this is minimized by random assignment
-requires more patients than in the cross-over design -> more expensive
What is a Run-in Period?
-Period of time before running a trial where the treatment or NO treatment is given
-incorporated to detect and exclude a subgroup of the study population (non-adherent patient - they are not taking their meds as prescribed, or look for outliers of placebo response when they get placebo in the run-in-period)
How can the data collected in the Run-in Period be helpful to optimize a study?
-Screen out non-compliant patients
-Screen for placebo response (placebo response is expected from any pt, they screen for a significantly higher placebo response)
-Ensure that the patients have the same baseline
-Exclude pt who are likely to have adverse effects
What is a Cross-over design?
-The same patients (e.g. n=500) receive both the investigational and control treatment during separate periods
-pt serves as their own control -> their reaction to each treatment is compared
-Less variability due to a smaller number of pt
-Data are highly affected by dropouts
(when dropping out early; X2 bc a pt is enrolled in drug and placebo group)
Timeline of Crossover
- Search for eligible patients
- Random Assignment of patients to control and placebo group
-2 groups: one group starts with the drug group and changes to the placebo group and vice versa
- Washout -> Pt cross-over from drug to placebo group and vice versa
What is the Washout period?
To avoid the Carryover Effect
If the drug works during a trial, you want the effect of the drug to stop before cross-over, to avoid carrying over the effects into the next phase
-> It might seem like the placebo effect is working but in reality, it is still the drug from the previous phase
What affects the length of the Washout period?
-The drug’s half-life (usually 5 half-lives)
-biological effect -> could be longer than the half-life (f.e. steroids have prolonged pharmacological effect days after elimination)
When should a Washout period be conducted?
-Before the Run-in-period ( creating Baseline levels)
-Between the swap of groups
When should the Baseline level of drug in patients be accomplished?
After the Washout period and right before random assignment into the groups
Is there a Run-in Period before the first phase of a Cross-over design??
What is the Period effect?
Conditions/diseases fluctuate in severity during the trial
-> It goes up during one phase of the study, but after cross-over, it doesn’t -> mask the actual patient’s response
What is a disease that is associated with the Period effect?
Osteoarthritis
What study design should be used to investigate disease with fluctuations?
Parallel Group study design preferred
What is the Sequence effect?
It happens when the sequence (order) may have an impact on the efficacy of a drug (f.e. Albuterol causes bronchodilation when the lungs are constricted -> if a steroid is tested and given previous to albuterol, it lowers inflammation and albuterol is able to penetrate the lung much better and cause much more dilation than it would be w/o the steroid -> the effect of albuterol is unintentionally improved
-To avoid this effect half of the patients are randomly assigned to the sequence (order) of the drug
-referred to as a 2-sequence study (during Cross-over)
Advantages of a Cross-over design
-minimize variability of drug response (bc the same patients for both groups)
-more efficient use of patients (same patients for both groups)
-smaller number of patients -> lower cost
Disadvantages of a Cross-over design
Patients are longer enrolled in a study (1 period in the drug group and 1 in the placebo group)
-the risk of dropout
-Dropout may be high -> f.e. in a study with 2 drugs -> twice as likely to experience adverse effects -> more likely to drop out
A single dropout can represent X2 bc one pt goes through 2 periods
More Disadvantages of Cross-over Design
-More exposure to drugs: In a study with active treatment (standard treatment), the patients are exposed to 2 drugs, composed to half of the patients being exposed to one drug each
-More exposure to placebo: reduces the likelihood of adverse effects, but patients are also untreated for a longer time (ethic?)
-Possibility of variation in patient’s disease (Period effect)
When are Cross-over designs appropriate?
-Condition is relatively stable (no fluctuations, no period effect)
-The treatment period is not too long (if treatment takes a year, we do not want to cross over and wait for another year)
-effects of the drug are reversible (drug effect is easy to wash away - to prevent the carry-over effect)
-when the carryover effect is not considered a problem
When are Cross-over designs inappropriate?
-Self-limited disease with a short duration bc the disease might go away before the cross-over
-> better to use a Parallel group design
When looking at a graph showing the sequence of a cross-over-design; What do Period and Sequence mean?
RECOGNIZE how different designs would look like
3 sequence 2 sequence, 1 period, 2 period
Period: Period where a drug is tested (could be the drug 1 in period 1 and the drug 2 in period 2)
Sequence: Describing the different types of treatment (drug 1, drug 2, drug 3, placebo)
??
Number of groups: f.e. concurrent treatment: Drug and Placebo -> then both swap
What would be a reason to use an Extra-Period Crossover?
An extra period is added to the second phase
-to investigate for side effects that may appear at a later phase of drug therapy
-the drug doesn’t work as well as expected in the short-term -> requires a longer time to manifest its beneficial effect
What are the problems with an Open-label Crossover without Randomization?
-open to Period effect
-open to sequence effect
-potential Bias
What does an “Invalid Initiation” in Cross-over-trial mean?
Starting at different Baseline points
Invalid Initiation in the Second part of the Cross-over-trial
There is no sufficient washout-time
so the placebo and medication have a different Baseline at the end of treatment II (makes Treatment III invalid)
What does the graph of the Valid Trial imply?
- sufficient washing out time -> both groups start Baseline
- Placebo and meds perform well in Period I, -> Washout and CROSS-OVER -> In Period II the meds perform well the placebo doesn’t
-> This is likely due to the period effect, fluctuation in period II which does not occur in period I (looks like the placebo doesn’t work in period II)
Shouldn’t the placebo not be working at all???
Ideal Collected Data from a cross-over-trial
- Both groups start at Baseline -> placebo doesn’t perform well, the meds do
- Sufficient washing time and cross-over
- Both start at the Baseline in period II placebo doesn’t perform well, and the meds do
