Evaluation of bioequivalence data MCQs With Answer

Introduction: This quiz collection on Evaluation of Bioequivalence Data provides focused practice for M.Pharm students preparing for exams and practical application in bioanalytical studies. The questions emphasize statistical principles, regulatory criteria, design choices, and data-handling strategies used in bioequivalence (BE) assessment — including log-transformation, two one-sided tests (TOST), confidence-interval interpretation, replicate and crossover designs, and methods for highly variable drugs. Each item targets common pitfalls and decision points encountered during BE analysis, helping students deepen conceptual understanding and interpret real study outputs correctly. Answers are provided to reinforce learning and to clarify why specific approaches are preferred in regulatory and analytical contexts.

Q1. Which statistical criterion is most commonly used by regulators to conclude average bioequivalence between test and reference products?

• Two-sided 95% confidence interval of arithmetic mean difference within ±20%
• Two one-sided tests (TOST) with 90% confidence interval for the geometric mean ratio within 80–125%
• One-sided t-test at alpha = 0.05 comparing means
• Nonparametric rank-sum test for median equivalence

Correct Answer: Two one-sided tests (TOST) with 90% confidence interval for the geometric mean ratio within 80–125%

Q2. Why are plasma AUC and Cmax generally log-transformed before statistical analysis in bioequivalence studies?

• To convert multiplicative variability into additive variability and approximate normality of residuals
• To make the data integers for ANOVA software
• To reduce the effect of outliers by truncation
• To obtain arithmetic mean ratios directly

Correct Answer: To convert multiplicative variability into additive variability and approximate normality of residuals

Q3. What is the purpose of the Two One-Sided Tests (TOST) procedure in bioequivalence testing?

• To test superiority of test vs reference
• To separately test for sequence and period effects
• To demonstrate equivalence by rejecting both null hypotheses that the ratio is ≤ lower limit or ≥ upper limit
• To estimate within-subject variance components

Correct Answer: To demonstrate equivalence by rejecting both null hypotheses that the ratio is ≤ lower limit or ≥ upper limit

Q4. Regulatory guidelines typically require reporting which confidence interval for bioequivalence decisions?

• 95% two-sided confidence interval
• 90% two-sided confidence interval for log-transformed parameters (equivalently two one-sided 5% tests)
• 99% confidence interval for safety endpoints only
• One-sided 97.5% confidence interval

Correct Answer: 90% two-sided confidence interval for log-transformed parameters (equivalently two one-sided 5% tests)

Q5. What is a primary advantage of a replicate crossover design over a classical 2×2 crossover design in BE studies?

• Eliminates the need for log-transformation
• Allows within-subject variability for each formulation to be estimated directly, enabling reference-scaling for highly variable drugs
• Reduces the number of periods required
• Makes non-parametric analysis mandatory

Correct Answer: Allows within-subject variability for each formulation to be estimated directly, enabling reference-scaling for highly variable drugs

Q6. When is scaled average bioequivalence (reference-scaling) typically applied?

• When between-subject variability is greater than 50%
• When intra-subject coefficient of variation (CV) for the reference product exceeds a regulatory threshold (commonly ~30%)
• When Tmax is highly variable
• When sample size is very small (<8 subjects)

Correct Answer: When intra-subject coefficient of variation (CV) for the reference product exceeds a regulatory threshold (commonly ~30%)

Q7. Which pharmacokinetic parameter is the primary measure of the extent of systemic exposure used in BE comparisons?

• Tmax
• Cmax
• AUC (area under the plasma concentration–time curve)
• Half-life (t1/2)

Correct Answer: AUC (area under the plasma concentration–time curve)

Q8. Which PK metric is most associated with the rate of absorption and is commonly evaluated alongside AUC in BE studies?

• AUC0–∞
• Cmax
• Apparent clearance
• Volume of distribution

Correct Answer: Cmax

Q9. In an ANOVA for a 2×2 crossover BE study, what does a significant sequence effect indicate?

• Carryover or systematic differences related to the order in which treatments were received
• That the test product is superior to reference
• That within-subject variability is zero
• No influence on BE interpretation and can be ignored

Correct Answer: Carryover or systematic differences related to the order in which treatments were received

Q10. Why does a 90% confidence interval correspond to two one-sided tests each at 5% significance in equivalence testing?

• Because regulators arbitrarily chose 90% for convenience
• Because two one-sided 5% tests produce an overall family-wise error of 10%
• Because constructing a 90% two-sided CI is algebraically equivalent to performing two one-sided tests at alpha = 0.05 each
• Because 95% CI would be too conservative for BE studies

Correct Answer: Because constructing a 90% two-sided CI is algebraically equivalent to performing two one-sided tests at alpha = 0.05 each

Q11. The geometric mean ratio reported in BE studies is best described as:

• The arithmetic mean of individual ratios
• The exponentiated difference of means on the log scale (test/reference)
• The median ratio of concentrations at each time point
• The ratio of pooled variances between test and reference

Correct Answer: The exponentiated difference of means on the log scale (test/reference)

Q12. Which statistical approach is most appropriate for comparing Tmax between formulations?

• Log-transformed ANOVA identical to AUC/Cmax tests
• Nonparametric methods (e.g., Wilcoxon signed-rank) because Tmax is often not normally distributed
• Scaled-average bioequivalence
• Chi-square test for proportions

Correct Answer: Nonparametric methods (e.g., Wilcoxon signed-rank) because Tmax is often not normally distributed

Q13. How are plasma concentrations below the Lower Limit of Quantification (BLQ) typically handled for AUC calculations in BE analyses?

• All BLQ values are always imputed by the mean of quantifiable values
• BLQ values prior to the first quantifiable concentration are set to zero; BLQ values between quantifiable points are handled according to prespecified rules to avoid bias
• BLQ values are deleted and ignored from the dataset
• BLQ values are replaced by LLOQ for all calculations

Correct Answer: BLQ values prior to the first quantifiable concentration are set to zero; BLQ values between quantifiable points are handled according to prespecified rules to avoid bias

Q14. What is the principal cause of carryover effect in crossover BE studies?

• Incorrect randomization
• Incomplete washout between periods leading to residual drug from the previous period
• Use of log-transformation on PK parameters
• Using geometric means instead of medians

Correct Answer: Incomplete washout between periods leading to residual drug from the previous period

Q15. Which factors primarily determine the sample size needed for a BE study?

• Intra-subject variability, expected test/reference ratio, alpha level, and desired power
• The number of blood sampling time points only
• Whether Tmax is tested parametrically or nonparametrically
• Choice of chromatography column in bioanalysis

Correct Answer: Intra-subject variability, expected test/reference ratio, alpha level, and desired power

Q16. In the standard ANOVA model for a 2×2 crossover, which terms are usually included?

• Formulation, period, sequence, and subject nested within sequence as random effect
• Only formulation and residual error
• Formulation, treatment-by-period interaction, and sampling time as covariates
• Sequence and subject as fixed effects only

Correct Answer: Formulation, period, sequence, and subject nested within sequence as random effect

Q17. When is a parallel-group study preferred for bioequivalence assessment?

• When the drug is highly variable and a replicate crossover is feasible
• For drugs with very long half-life or when washout is impractical so carryover cannot be avoided
• When within-subject variability is negligible
• Only when performing nonparametric BE tests

Correct Answer: For drugs with very long half-life or when washout is impractical so carryover cannot be avoided

Q18. How do regulators typically tighten acceptance limits for narrow therapeutic index (NTI) drugs?

• They always accept 80–125% limits regardless of drug class
• They may require narrower acceptance limits (for example, 90–111% for AUC) or stricter review depending on the agency and drug
• They switch to nonparametric median-based acceptance criteria
• Tmax is used as the sole criterion for NTI drugs

Correct Answer: They may require narrower acceptance limits (for example, 90–111% for AUC) or stricter review depending on the agency and drug

Q19. A statistically significant subject-by-formulation interaction suggests which of the following?

• That sequencing was done incorrectly
• That treatment effects are consistent and no further investigation is needed
• That the relative performance of formulations varies across subjects, indicating possible formulation-specific responses and complicating average BE interpretation
• That geometric mean ratio is exactly 1.00

Correct Answer: That the relative performance of formulations varies across subjects, indicating possible formulation-specific responses and complicating average BE interpretation

Q20. Why might a bootstrap approach be used when estimating confidence intervals in BE analyses?

• Because bootstrap always yields narrower intervals regardless of data
• To obtain distribution-free confidence intervals when standard parametric assumptions (normality of log-transformed residuals) may be violated or sample size is limited
• Because regulators require bootstrap in all BE submissions
• To avoid having to measure Cmax and AUC experimentally

Correct Answer: To obtain distribution-free confidence intervals when standard parametric assumptions (normality of log-transformed residuals) may be violated or sample size is limited

Download