Quiz #2 Flashcards
Define controlled clinical trial
A study testing a specific drug or other treatment involving 2 or more groups of patients with the same disease. One group receives treatment and the other receives an alternate treatment, a placebo treatment or no treatment. The groups are followed up to compare the differences in outcomes
Define Randomized Control trial
A controlled clinical trial where patients are randomly allocated to treatment and comparison groups
Define hypothesis
an idea/explanation tested through study and experimentation
Define Independent variable
The variable that is intentionally changed to observe its effect on the dependent variable
What is the difference between sample and population?
Sample: a representative unit of the population for study
Population: a set of individuals sharing specified properties
What is the difference between convenience sampling and probability sampling?
Convenience: the subjects selected for inclusion in the sample are easiest to access (like volunteer)
Probability: the subjects selected for inclusion in the study are done at random
Define cofounder
Unknown or unmeasured variables that may affect the relationship between exposure and outcome
Has a relationship with BOTH the exposure and the outcome
What is the difference between patient-level and cluster randomization?
They are both units of randomization
Patient-level: each subject is individually allocated to intervention or control
Cluster: subjects are allocated as part of a group (example: all patients of pharmacy A are intervention, all patients of pharmacy B are control)
What is the simple method of randomization?
Subject assignment without any consideration of previous assignments or affiliations
What is the stratified method of randomization?
Subjects grouped based on an important characteristic, then randomly sampled from that group into intervention or control
(example: female vs male, middle age vs elderly)
What is the block randomization method?
Enroll and randomize subjects within multiple smaller ‘blocks’ that eventually add up to the desired sample size
Done to make an even amount of people in one group and ensures blinding
What is the serial/factorial method of randomization?
Subjects are randomized into one group and then randomized into a second group
EX: randomized to intervention or control, then further randomize intervention group into different doses
What is quasi-randomization?
RED FLAG: not a true form of research where patients are not randomly allocated to a group and so confounding variables would be a problem
What is clinical Equipoise?
The point where there is no current clinical preference between treatments; equally likely that treatment A or treatment B will turn out to be superior
Ethics would not approve a trial unless this was proven by the experimenter
What is the difference between single, double and triple blinding?
Single = subject OR investigator Double = subject AND investigator Triple = subject AND investigator AND data analyst
What is an open label study?
No one is blinded
What is involved in a cross-over study design?
Subjects serve as their own control and each subject receives all treatments sequentially
Define Washout Period
Drug-free interval between treatments designed to eliminated carry-over effects from one treatment to the next during a cross-over study
What is an N-of-1 study?
A form of cross-over study but there is only one participant, also has a washout period.
Should not be performed for acute curable conditions and results can’t necessarily be generalized to other patients
Define variables a, b, c, and d from epidemiology calculations
a = exposure with outcome
b = exposure without outcome
c = no exposure with outcome
d = no exposure without outcome
What is the EER?
EER = Exposure event rate (proportion of exposed patients with the outcome)
EER = a/(a+b)
What is the CER?
CER = Control event rate (proportion of non-exposed patients with the outcome)
CER = c/(c+d)
What is the relative risk?
RR = EER/CER
What is the relative risk reduction?
RRR = (CER - EER)/CER
What is the absolute risk reduction?
ARR = CER - EER
What is the number needed to treat?
NNT = 1/ARR
How do you calculate the odds that a patient with the exposure has the outcome?
a/b
How do you calculate the odds that a patient without the exposure has the outcome?
c/d
How do you calculate the odds ratio?
(a +d)/(b+c)
Define descriptive statistics
Analyses that help describe, show or summarize our data
Define inferential statistics
Analyses that allow us to make conclusions about the relationship between independent and dependent variables, or generalize our sample’s findings to the population
What is the relationship between mean, median and mode in symmetrical vs positive skew vs negative skew data?
Symmetrical distribution: all the same
Positive skew: graph is shifted left, mode is the highest point, coming first, then median, then mean
Negative skew: graph is shifted to the right, mean is lowest , coming first, then median, then mode
What is a confidence interval?
The term used to describe the uncertainty around a point estimate; used because the estimate is generally taken from a sample and we cannot be 100% certain it is a ‘true’ value in the population
What is the null hypothesis?
Typically that the independent variable has no effect on the dependent variable
If there was a study wanting to see that two options are equivalent, the null would be that there is a difference
What is the alternative hypothesis?
The independent variable has a significant effect on the dependent variable -> what we want to see
What is the difference between type one and type two errors?
Type 1: we reject the null hypothesis and it turns out to be true
Type 2: we accept the null hypothesis and it turns out to be false
What is statistical power?
Power = 1 - B
Where B = probability of making a type 2 error
Power is therefore the opposite of a type 2 error
Define primary outcome
Outcome that will answer the primary or most important question
Sample size of the trial is calculated based on the change we expect to see in this outcome
Define secondary outcome
Other relevant outcomes about the same study but of less importance and are only purposeful in hypothesis generating
What is the difference in paired vs unpaired t-tests?
Paired = both measures come from the same subject
Unpaired = measures come from different subjects
What is ANOVA?
ANOVA is like a series of t-tests but accounts for there being >2 independent variables
What is the difference between efficacy and effectiveness?
Efficacy = a measure of the capacity of a treatment to produce the desired effect in a controlled environment (like an RCT)
Effectiveness = the actual effect of the treatment in practice
What is the difference between accuracy and precision?
Accuracy = how close the measurement is to the true value
Precision = measurements of how close measurements of the same item are
Define validity
Whether or not the variable reported actually measures the outcome of interest
What is the difference between sensitivity and specificity?
Sensitivity = ability of a test to correctly identify those patients with a disease
–> low rate of false negatives
Specificity = ability of a test to correctly identify those patients without the disease
–> low rate of false positives
What is the difference between statistical and clinical significance?
Statistical = a difference between two interventions resulting in p<0.05
Clinical = a difference between two interventions that is meaningful to the patient and to their health outcomes