MPharm Master of Pharmacy Master of Pharmacy: Research and Biostatistics 5 — Questions and Answers
Question 1: In a pharmacogenomics study using genome-wide association (GWAS), what statistical threshold is typically applied to define significance?
- p < 5×10⁻⁸ (Correct answer)
- p < 0.05
- p < 0.01
- p < 1×10⁻³
Correct answer: p < 5×10⁻⁸
GWAS typically requires p < 5×10⁻⁸ to account for approximately one million independent tests performed across the genome (Bonferroni-corrected).
Question 2: In a dose-response study, the EC₅₀ represents:
- The drug concentration producing 50% of the maximal effect (Correct answer)
- The dose at which 50% of subjects show a response
- The elimination half-life at 50% of the maximum dose
- The concentration at which 50% of drug is protein-bound
Correct answer: The drug concentration producing 50% of the maximal effect
EC₅₀ is the effective concentration required to produce 50% of the maximal pharmacological response and is a key PD parameter.
Question 3: When a clinical trial's data safety monitoring board (DSMB) performs an interim analysis, they may use an O'Brien-Fleming boundary to:
- Preserve the overall Type I error while allowing early stopping for efficacy (Correct answer)
- Calculate the minimum sample size needed
- Adjust for covariate imbalance between arms
- Assess carry-over effects in crossover designs
Correct answer: Preserve the overall Type I error while allowing early stopping for efficacy
O'Brien-Fleming spending functions allocate alpha across interim looks, using very conservative early thresholds so the final alpha approximates the planned 0.05.
Question 4: A pharmacy researcher wants to measure inter-rater reliability for a subjective adverse event severity scale. The most appropriate statistic is:
- Cohen's kappa (Correct answer)
- Pearson's r
- Spearman's rho
- Intraclass correlation coefficient (ICC) for ratio-scale data
Correct answer: Cohen's kappa
Cohen's kappa measures agreement between two raters on categorical classifications beyond what is expected by chance.
Question 5: In adaptive clinical trial designs, 'response-adaptive randomization' modifies:
- Allocation probabilities based on accumulating outcome data (Correct answer)
- Eligibility criteria based on interim safety signals
- Sample size based on observed variance estimates
- Follow-up duration based on event rates
Correct answer: Allocation probabilities based on accumulating outcome data
Response-adaptive randomization shifts allocation toward the better-performing arm as outcome data accumulate, increasing ethical efficiency.
Question 6: Which pharmacokinetic parameter best reflects the total systemic drug exposure over time after a single dose?
- Area under the concentration-time curve (AUC₀–∞) (Correct answer)
- Peak plasma concentration (Cmax)
- Elimination half-life (t½)
- Volume of distribution (Vd)
Correct answer: Area under the concentration-time curve (AUC₀–∞)
AUC₀–∞ integrates the entire concentration-time profile from administration to infinity, representing total systemic drug exposure.
Question 7: In meta-analysis, heterogeneity among studies is quantified using the I² statistic. An I² of 75% indicates:
- High heterogeneity; a random-effects model is preferred over fixed-effects (Correct answer)
- Low heterogeneity; a fixed-effects model is appropriate
- Moderate heterogeneity; subgroup analysis is not warranted
- Publication bias is present in 75% of included studies
Correct answer: High heterogeneity; a random-effects model is preferred over fixed-effects
I² ≥ 75% indicates high statistical heterogeneity, suggesting true variability in effect sizes across studies and favoring a random-effects model.
MPharm Master of Pharmacy Master of Pharmacy: Research and Biostatistics 5
In a pharmacogenomics study using genome-wide association (GWAS), what statistical threshold is typically applied to define significance?