Biostatistics Flashcards
RRR
Relative Risk Reduction
1-RR
RR
Relative Risk
Risk of Tx/Risk of Control
A coffee creamer company recently learned that several lots ofproductwere contaminated with a human carcinogen. It is unknown what effect consuming this carcinogen in small quantities may have on humans. An immediate recall is announced. A massive media campaign is underway to identify people who used the lots of coffee creamer so they can receive medical attention immediately. The company plans to conduct a study to determine the effects of this exposure. They willfollowpeoplewho used the contaminated creamer over time and compare the rates of cancertopeople who used their coffee creamer, but who were not exposed to the contaminated lot. What study design does this trial represent?
A. Cross-sectional study
B. Case series
C. Case-control study
D. Cohort study
E. Case report
Cohort study
This is a prospective cohort study design.
Cohort studies start with an exposure/intervention and look for an outcome of interest (disease).
In this case the study starts with the exposure (contaminated coffee creamer) and patients will be followed prospectively to look for an outcome (disease such as cancer).
This design is useful when randomization to the exposure would be unethical.
Case-Control Study
Compares pts w/ a disease (cases) to those w/out the disease (control)
Retrospective
Start w/ disease looks for exposure
Cohort Study
Compares outcomes of a group of pts exposed & not exposed to a Tx
Prospective
Retrospective - less common
Start w/ exposure looks for disease
Cross-sectional Survey
Estimates the relationship between variables & outcomes (prevalence) at one particular time (cross-section) in a defined population
Case Report & Case Series
Describes an adverse rxn or a unique condition that appears in a single pt (case report) or a few pts (case series)
Outcome of the case is known
Case series is more reliable than a case report
Randomized Controlled Trial (RCT)
Compares experimental Tx to a control (placebo or existing Tx)
Pts are randomized (have equal chance of being assigned to the Tx or control)
Sometimes blinded (unaware if they are receiving Tx or control)
Randomized Controlled Trial (RCT)
Double-blind
Pt & investigator are unaware
Randomized Controlled Trial (RCT)
Single-blind
Pt is unaware
Randomized Controlled Trial (RCT)
Open label
or Unblinded
All parties know which Tx is being given to the pt
Randomized Controlled Trial (RCT)
Parallel
Subjects are randomized to the Tx or control arm for the entire study
Randomized Controlled Trial (RCT)
Crossover
Pts are randomized to 1 of 2 sequential Txs:
Group 1 - receive Tx A 1st, then crossover (change) to Tx B
Group 2 - receive Tx B 1st, then crossover (change) to Tx A
Factorial Design
Randomizes to more than the usual 2 groups to test a number of experimental conditions
Meta-analysis
Combines results from multiple studies in order to develop a conclusion that has greater statistical power than is possible from the individual smaller studies
Systematic Review Article
Summary of the clinical literature that focuses on a specific topic or questions
Types of Medical Studies
Least to Most Reliable
Expert Opinion
Case series & Case Report
Cast-Controlled Studies
Cohort Studies
Randomized Controlled Trials
Systematic Reviews & Meta-Analyses
OR Formula
AD/BC
Odds/Hazard Ratio
> 1
↑ in the likelihood
Odds/Hazard Ratio
< 1
↓ in the likelihood
Odds/Hazard Ratio
= 1
no correlation
Odds Ratio
How to express OR > 1 in %?
OR x 100 - 100
A pharmacist is considering which beta-blocker should be the preferred formulary agent at his institution. He has narrowed hisselection down to two agents. Each drug provides similar health benefits, has similar tolerability and is dosed once daily. The pharmacist will base his decision on the drug that can be purchased at the lower cost. This type of evaluation can be described as:
A. A cost-minimization analysis
B. A cost-effectiveness analysis
C. A cost-control analysis
D. A cost-benefit analysis
E. A cost-utility analysis
A cost-minimization analysis
If the benefit of two drugs or interventions is considered equivalent (e.g., there is no clinical benefit or harm of one compared to the other), a cost-minimization analysis can be used to determine which treatment is less expensive.
The rate of the primary outcome was 1.28% per year in the apixaban groupand 1.60% per year in the warfarin group (95% CI, 0.66 to 0.95; p < 0.001 for noninferiority; p = 0.01 for superiority).
In this trial, the primary outcome is reported as a hazard rate (rate per year). Calculate the hazard ratio of the primary outcome in patients receiving apixaban compared to warfarin.(Express the answer as a decimal; no units or commas; round the final answer to the nearest TENTH.)
0.8
Hazard ratio (HR) = 1.28%/1.6% = 0.8. It is calculated in a similar way as risk ratio.
When HR < 1, this is favorable for the study drug as it means there were fewerunfavorable events in the treatment group.
HR formula
Hazard rate Tx/Hazard rate Control
Continuous Data
Data is provided by some type of measurement
Ratio & Interval
Ratio Data
0 = None
Interval Data
0 does not equal none
Discrete Data
Categorical
Nominal & Ordinal
Nominal Data
arbitrary order
Ordinal Data
ranked in a logical order
Student’s T-test
Continuous Data
2 independent groups
Paired T-test
Continuous Data
2 paired groups
ANOVA
Continuous Data
≥ 3 independent/paired groups
Chi-Square
Nominal Data
≥ 2 independent groups
McNemar
Nominal Data
2 paired groups
Cochran’s Q
Nominal Data
≥ 3 paired groups
Wilcoxon Rank Sum
Ordinal Data
2 independent groups
Wilcoxon Signed Rank
Ordinal Data
2 paired groups
Kruskal-Wallis
Ordinal Data
≥ 3 independent groups
Friedman
Ordinal Data
≥ 3 paired groups
Incremental Cost Rations Formula
(C2 - C1)/(E2 - E1)
Cost-Minimization Analysis (CMA)
Compares cost of interventions w/ demonstrated equivalence
Cost-Benefit Analysis (CBA)
Compares benefits & costs in monetary units
Cost-Effectiveness Analysis (CEA)
Most common
Costs are monetary but outcomes are in clinical units (easty to quantify)
Cost-Utility Analysis (CUA)
Outcomes based on quality-of-life assessments
Intent to Treat
Data from ALL pts even the ones who did not complete the study
Per Protocol
Data from ONLY the pts who completed the study
The definition of sensitivity is best represented as:
A. The percentage of patients without a disease who have a negative test result.
B. The probability that an event will or will not occur.
C. The effect of an independent variable on a dependent variable.
D. The percentage of patients with a disease who have a positive test result.
E. The percentage of patients that will benefit from an intervention when a test result is positive.
The percentage of patients with a disease who have a positive test result.
Sensitivity is the percentage of “true-positive” results (a test is positive and the patient has the disease).
Sensitivity
“True Positive”
Test is positive in patients w/ the condition
Specificity
“True Negative”
Test is negative in patients w/out the condition
A clinical trial is conducted on a new drug, LoSod. LoSod is found to reduce mean serum sodium to 131 mEq/L (95% confidence interval:128.7 mEq/L - 133.3 mEq/L). What is the correct interpretation of this study result?
A. There is a 95% chance that the confidence interval range contains the true population value for reduction in serum sodium with LoSod.
B. When using this drug in a larger population, 95% of patients will have a serum sodium between 128.7and 133.3 mEq/L.
C. When using this drug in a larger population, there is 95% confidence that the serum sodium will be reduced to 131 mEq/L.
D. There is a 5% chance that the true population mean is within the stated range.
E. There is more than a 5% chance that the results of the study are rejected in error.
There is a 95% chance that the confidence interval range contains the true population value for reduction in serum sodium with LoSod.
A confidence intervalis a range of values which is likely to includethe “true” population value.
It is more helpful than the p-value in estimating the effect of the outcome on the general population.
Asmall, randomized, double-blind trial was conducted to evaluate a new treatment in patients with acute myocardial infarction. The primary endpoint, reduction in all-cause mortality, was not statistically significant.Which of the following is an accurate conclusion?
A. The inclusion and exclusion criteria were notdefined appropriately to reach statistical significance.
B. Further studies of the treatment should be discontinued immediately.
C. If there was any reduction in mortality, the treatment should be recommended for use.
D. A type I error may have occurred, preventing the ability to detect statistical significance.
E. The clinical trial may not have adequate statistical power.
The clinical trial may not have adequate statistical power.
Performing a larger trial will increase statistical power, which increases the probability that a type II error (not finding a difference when there is a difference) will be avoided.
