Clinical Trials Flashcards
what is selection bias?
a distortion in a measure of association because the sample selection does not accurately reflect the target population
what is performance bias?
Performance bias occurs when the studiers are aware of the participant’s treatment and intentionally or unintentionally influences their responses to treatment.
what is detection bias?
Detection bias is any differences between groups changing the way the outcome is determined. For example, larger men have larger prostates and therefore may be easier misdiagnosed for prostate cancer via biopsy due to a physical difference
what is attrition bias?
Systematic differences between study groups in the number and way participants are lost from a study
what is reporting bias?
occurs when researchers do not report all of their data, which could lead to drawing false conclusions from the study outcome
how do we mitigate selection bias?
We use RANDOMIZATION to place the study groups… a computerized random number generator is preferred to truly have random allocation of subjects
how do we mitigate performance/detection bias?
We use BLINDING to keep subjects, investigators, and staff unaware of treatment allocation AFTER randomization. We use blinding of treatments to mitigate, such as making sure the two “drugs” are similar, both in appearance and maybe even known side effects
how do we mitigate attrition bias?
we use INTENTION TO TREAT PROTOCOL where all subjects in initial randomization are included in final analysis regardless if they receive treatment
how do we mitigate reporting bias?
Use studies that are performed using INDUSTRY SPONSORED RESEARCH
what are 3 methods of randomization to ensure allocation concealment?
simple randomization
block randomization
stratified randomization
what is simple randomization?
50/50 chance for each subject
what is block randomzation?
subjects randomized in “blocks” to make sure groups are equivalent sizes
what is stratified randomization?
Separate by a common factor (men separate from women)
what is allocation concealment and the benefits from doing so?
A successful implementation of a random allocation sequence…
Pros:shields all people involved from knowing upcoming treatment/control assignments
what is the difference b/w allocation concealment and blinding?
allocation concealment is randomization before the trial begins, blinding is a method to mask treatment after allocation occurs.
define a per-protocol analysis. when is it a preferred method?
collects data from only patients who adhere to the protocol and is preferred fr equivalence/non-inferiority trials
define an intention-to-treat analysis
collects data from all patients in initial randomization to help preserve the randomization
what is the purpose of a survival analysis?
It shows an estimate of median survival times over the duration of the study
what is a cox proportional hazards model?
creates survival curves with hazard functions that are proportional over time, and the relationship between each variable is linear
what is the purpose of a p-value? what value is considered statistically significant?
determine if we can reject the null hypothesis.
p-value should be 0.05 or lower
How can we use confidence intervals to show non-inferiority?
look at image 1
define relative risk
The risk of an event occurring in one group compared to the other group. A RR less than 1 means the outcome is decreased by the exposure
calculate relative risk
look at image 2
define relative risk reduction
how much risk is reduced relative to the risk reduction. shows how much the treatment reduces the risk of an outcome
calculate relative risk reduction
RRR=(1-RR)100= % RRR
example:RRR= (1-0.46)100 = 54%
meaning the treatment reduced the risk of the adverse outcome by 54%
define number needed to treat
Number needed to treat is based on the amount of patients within the specified period of time to prevent one adverse outcome
calculate number needed to treat
look at image 4
NNT= 1/risk difference
what is internal validity?
examines if the study design, conduct, and analysis answer the research question without bias. This encompasses all of our analyzing for potential bias
what is external validity? (what is PICOS)
examines if the study findings are applicable towards patients outside the trial study. For external validity we use “PICOS”, or Patients (exclusion criteria, characteristics), Intervention (differences b/w trial procedures and real world practice), Comparison (are placebo vs. active appropriate), Outcomes (does event matter?), and Setting
what does it mean if a study has subgroup analyses?
researchers create separate data based on differing characteristics from the overall number of participants in the study. For example, the researchers wanted to specifically see how African-Americans responded to the treatment and thus made additional data on that population group.
what are surrogate outcomes?
Surrogate outcomes depict data that is used to demonstrate the greater efficacy of the treatment drug by showing data on non-primary outcomes. For example, providing data on how the drug lowered A1C or blood pressure when the primary outcome was if the drug reduced mortality. They are a way to beef up the drug
what are composite outcomes?
Composite outcomes are when multiple outcomes are combined into one composite outcome to improve the power/efficiency of trials. For example, the researchers give one data point (major adverse cv events) for how the drug prevented nonfatal stroke, MI, and CV death. These are all separate outcomes, but combined can show a greater efficacy of how the drug prevented major CV events
define absolute risk reduction/difference
The proportion of patients who are spared the adverse outcome as a result of having received the experimental treatment rather than the control
calculate absolute risk reduction/difference
placebo risk: 0.76%
treatment risk: 0.35%
RD= 0.76% - 0.35% = 0.41%
the absolute risk of event is reduced by 0.41% compared to placebo
