CPI Clinical Research & Study Design 3 — Questions and Answers
Question 1: Which of the following best defines 'protocol deviation' versus 'protocol violation'?
- A deviation is major and affects data integrity; a violation is minor and does not
- A deviation is any departure from the protocol; a violation is a deviation that could affect subject safety or data integrity (Correct answer)
- Violations are reported to the IRB; deviations are not
- Both terms are interchangeable under FDA regulations
Correct answer: A deviation is any departure from the protocol; a violation is a deviation that could affect subject safety or data integrity
Protocol deviations encompass all departures from the protocol, while violations specifically threaten subject safety, rights, or data reliability.
Question 2: In an adaptive trial design, a pre-planned interim analysis may allow:
- Retroactive changes to the primary endpoint
- Sample size re-estimation or arm discontinuation based on accumulating data (Correct answer)
- Unblinding of the sponsor only
- Addition of new sites without IRB approval
Correct answer: Sample size re-estimation or arm discontinuation based on accumulating data
Pre-planned adaptations such as sample-size re-estimation or dropping ineffective arms are key features of adaptive designs, maintaining statistical validity.
Question 3: Which statistical concept describes the probability of detecting a true treatment effect when one exists?
- Type I error (alpha)
- Confidence interval width
- Statistical power (1 – beta) (Correct answer)
- P-value threshold
Correct answer: Statistical power (1 – beta)
Statistical power is 1 minus the Type II error rate (beta) and represents the probability of correctly rejecting a false null hypothesis.
Question 4: A Phase I dose-escalation study in oncology most commonly uses the 3+3 design to determine the:
- Maximum tolerated dose (MTD) (Correct answer)
- Minimum effective dose
- Recommended Phase III dose
- Pharmacokinetic half-life
Correct answer: Maximum tolerated dose (MTD)
The 3+3 design escalates doses in cohorts of three patients and defines the MTD based on dose-limiting toxicity occurrence.
Question 5: Which of the following is an example of information bias in clinical research?
- Differential loss to follow-up between treatment arms
- Systematic over-reporting of symptoms in one exposure group (Correct answer)
- Imbalanced baseline characteristics between groups
- Exclusion of non-English-speaking participants
Correct answer: Systematic over-reporting of symptoms in one exposure group
Information bias arises when data are collected or recorded differently across groups, such as when exposed participants are questioned more thoroughly about symptoms.
Question 6: The number needed to treat (NNT) is calculated as:
- Relative risk divided by absolute risk
- 1 divided by the absolute risk reduction (Correct answer)
- Odds ratio multiplied by baseline risk
- Relative risk reduction minus 1
Correct answer: 1 divided by the absolute risk reduction
NNT = 1 / ARR (absolute risk reduction); it tells clinicians how many patients must be treated to prevent one additional adverse outcome.
Question 7: When a study sponsor monitors a clinical trial site, which document is the primary source for verifying data entered into the CRF?
- Protocol synopsis
- Investigator Brochure
- Source document (medical record, lab report, etc.) (Correct answer)
- Case Report Form itself
Correct answer: Source document (medical record, lab report, etc.)
Source data verification (SDV) involves comparing CRF entries against original source documents to confirm accuracy and completeness.
CPI Clinical Research & Study Design 3
Which of the following best defines 'protocol deviation' versus 'protocol violation'?