Group sequential designs — MCQs

10 questions

An investigator is measuring the blood calcium level in a sample of female cross country runners and a control group of sedentary females. If she would like to compare the means of the two groups, which statistical test should she use?

You submit a paper to a prestigious journal about the effects of coffee consumption on mesothelioma risk. The first reviewer lauds your clinical and scientific acumen, but expresses concern that your study does not have adequate statistical power. Statistical power refers to which of the following?

You are currently employed as a clinical researcher working on clinical trials of a new drug to be used for the treatment of Parkinson's disease. Currently, you have already determined the safe clinical dose of the drug in a healthy patient. You are in the phase of drug development where the drug is studied in patients with the target disease to determine its efficacy. Which of the following phases is this new drug currently in?

In a randomized controlled trial studying a new treatment, the primary endpoint (mortality) occurred in 14.4% of the treatment group and 16.7% of the control group. Which of the following represents the number of patients needed to treat to save one life, based on the primary endpoint?

A pharmaceutical company conducts a randomized clinical trial in an attempt to show that their new anticoagulant drug prevents more thrombotic events following total knee arthroplasty than the current standard of care. However, a significant number of patients are lost to follow-up or fail to complete treatment according to the study arm to which they were assigned. Several patients in the novel drug arm are also switched at a later time to a novel anticoagulant or warfarin per their primary care physician. All patients enrolled in the study are subsequently analyzed based on the initial group they were assigned to and there is a significant improvement in outcome of the new drug. What analysis most appropriately describes this trial?

You are reading through a recent article that reports significant decreases in all-cause mortality for patients with malignant melanoma following treatment with a novel biological infusion. Which of the following choices refers to the probability that a study will find a statistically significant difference when one truly does exist?

A health system implements a new sepsis protocol across 20 hospitals. A researcher plans to evaluate effectiveness using a stepped-wedge cluster randomized design where hospitals sequentially adopt the protocol every 3 months. She calculates sample size based on individual patient outcomes (mortality) needing 2,000 patients total. The biostatistician identifies a critical error. Evaluate what modification is needed.

A 41-year-old research fellow designs a non-inferiority trial comparing oral to IV antibiotics for osteomyelitis. She sets the non-inferiority margin at 10% (cure rate difference), expects 85% cure in both groups, and calculates 300 patients per arm for 80% power with α=0.025 (one-sided). Her mentor suggests this underestimates required sample size. Evaluate the mentor's concern.

A pharmaceutical company tests a new antidepressant in 500 patients (250 per arm) and finds a 2-point improvement on a 52-point depression scale compared to placebo (p=0.04). The study was originally powered to detect a 4-point difference. The company seeks FDA approval citing statistical significance. Analyze the regulatory and scientific implications.

Q10

A meta-analysis of 5 previous trials testing a surgical technique shows a pooled effect size of 15% complication reduction (from 20% to 17%, p=0.30, I²=0%). An investigator wants to design a definitive trial. She calculates that 1,200 patients per arm would provide 80% power to detect this 3% absolute difference. Analyze whether this sample size is justified.

Group sequential designs US Medical PG Practice Questions and MCQs

Question 1: An investigator is measuring the blood calcium level in a sample of female cross country runners and a control group of sedentary females. If she would like to compare the means of the two groups, which statistical test should she use?

A. Chi-square test
B. Linear regression
C. t-test (Correct Answer)
D. ANOVA (Analysis of Variance)
E. F-test

Explanation: ***t-test*** - A **t-test** is appropriate for comparing the means of two independent groups, such as the blood calcium levels between runners and sedentary females. - It assesses whether the observed difference between the two sample means is statistically significant or occurred by chance. *Chi-square test* - The **chi-square test** is used to analyze categorical data to determine if there is a significant association between two variables. - It is not suitable for comparing continuous variables like blood calcium levels. *Linear regression* - **Linear regression** is used to model the relationship between a dependent variable (outcome) and one or more independent variables (predictors). - It aims to predict the value of a variable based on the value of another, rather than comparing means between groups. *ANOVA (Analysis of Variance)* - **ANOVA** is used to compare the means of **three or more independent groups**. - Since there are only two groups being compared in this scenario, a t-test is more specific and appropriate. *F-test* - The **F-test** is primarily used to compare the variances of two populations or to assess the overall significance of a regression model. - While it is the basis for ANOVA, it is not the direct test for comparing the means of two groups.

Question 2: You submit a paper to a prestigious journal about the effects of coffee consumption on mesothelioma risk. The first reviewer lauds your clinical and scientific acumen, but expresses concern that your study does not have adequate statistical power. Statistical power refers to which of the following?

A. The probability of detecting an association when no association exists.
B. The probability of not detecting an association when an association does exist.
C. The probability of detecting an association when an association does exist. (Correct Answer)
D. The first derivative of work.
E. The square root of the variance.

Explanation: ***The probability of detecting an association when an association does exist.*** - **Statistical power** is defined as the probability that a study will correctly reject a false null hypothesis, meaning it will detect a true effect or association if one exists. - A study with **adequate statistical power** is less likely to miss a real effect. *The probability of detecting an association when no association exists.* - This describes a **Type I error** or **false positive**, often represented by **alpha (α)**. - It is the probability of incorrectly concluding an effect or association exists when, in reality, there is none. *The probability of not detecting an association when an association does exist.* - This refers to a **Type II error** or **false negative**, represented by **beta (β)**. - **Statistical power** is calculated as **1 - β**, so this option describes the complement of power. *The first derivative of work.* - The first derivative of work with respect to time represents **power** in physics, which is the rate at which work is done. - This option is a **distractor** from physics and is unrelated to statistical power in research. *The square root of the variance.* - The **square root of the variance** is the **standard deviation**, a measure of the dispersion or spread of data. - This is a statistical concept but is not the definition of statistical power.

Question 3: You are currently employed as a clinical researcher working on clinical trials of a new drug to be used for the treatment of Parkinson's disease. Currently, you have already determined the safe clinical dose of the drug in a healthy patient. You are in the phase of drug development where the drug is studied in patients with the target disease to determine its efficacy. Which of the following phases is this new drug currently in?

A. Phase 4
B. Phase 1
C. Phase 2 (Correct Answer)
D. Phase 0
E. Phase 3

Explanation: ***Phase 2*** - **Phase 2 trials** involve studying the drug in patients with the target disease to assess its **efficacy** and further evaluate safety, typically involving a few hundred patients. - The question describes a stage after safe dosing in healthy patients (Phase 1) and before large-scale efficacy confirmation (Phase 3), focusing on efficacy in the target population. *Phase 4* - **Phase 4 trials** occur **after a drug has been approved** and marketed, monitoring long-term effects, optimal use, and rare side effects in a diverse patient population. - This phase is conducted post-market approval, whereas the question describes a drug still in development prior to approval. *Phase 1* - **Phase 1 trials** primarily focus on determining the **safety and dosage** of a new drug in a **small group of healthy volunteers** (or sometimes patients with advanced disease if the drug is highly toxic). - The question states that the safe clinical dose in a healthy patient has already been determined, indicating that Phase 1 has been completed. *Phase 0* - **Phase 0 trials** are exploratory, very early-stage studies designed to confirm that the drug reaches the target and acts as intended, typically involving a very small number of doses and participants. - These trials are conducted much earlier in the development process, preceding the determination of safe clinical doses and large-scale efficacy studies. *Phase 3* - **Phase 3 trials** are large-scale studies involving hundreds to thousands of patients to confirm **efficacy**, monitor side effects, compare it to commonly used treatments, and collect information that will allow the drug to be used safely. - While Phase 3 does assess efficacy, it follows Phase 2 and is typically conducted on a much larger scale before submitting for regulatory approval.

Question 4: In a randomized controlled trial studying a new treatment, the primary endpoint (mortality) occurred in 14.4% of the treatment group and 16.7% of the control group. Which of the following represents the number of patients needed to treat to save one life, based on the primary endpoint?

A. 1/(0.144 - 0.167)
B. 1/(0.167 - 0.144) (Correct Answer)
C. 1/(0.300 - 0.267)
D. 1/(0.267 - 0.300)
E. 1/(0.136 - 0.118)

Explanation: ***1/(0.167 - 0.144)*** - The **Number Needed to Treat (NNT)** is calculated as **1 / Absolute Risk Reduction (ARR)**. - The **Absolute Risk Reduction (ARR)** is the difference between the event rate in the control group (16.7%) and the event rate in the treatment group (14.4%), which is **0.167 - 0.144**. *1/(0.144 - 0.167)* - This calculation represents 1 divided by the **Absolute Risk Increase**, which would be relevant if the treatment increased mortality. - The **NNT should always be a positive value**, indicating the number of patients to treat to prevent one adverse event. *1/(0.300 - 0.267)* - This option uses arbitrary numbers (0.300 and 0.267) that do not correspond to the given **mortality rates** in the problem. - It does not reflect the correct calculation for **absolute risk reduction** based on the provided data. *1/(0.267 - 0.300)* - This option also uses arbitrary numbers not derived from the problem's data, and it would result in a **negative value** for the denominator. - The difference between event rates of 0.267 and 0.300 is not present in the given information for this study. *1/(0.136 - 0.118)* - This calculation uses arbitrary numbers (0.136 and 0.118) that are not consistent with the reported **mortality rates** of 14.4% and 16.7%. - These values do not represent the **Absolute Risk Reduction** required for calculating NNT in this specific scenario.

Question 5: A pharmaceutical company conducts a randomized clinical trial in an attempt to show that their new anticoagulant drug prevents more thrombotic events following total knee arthroplasty than the current standard of care. However, a significant number of patients are lost to follow-up or fail to complete treatment according to the study arm to which they were assigned. Several patients in the novel drug arm are also switched at a later time to a novel anticoagulant or warfarin per their primary care physician. All patients enrolled in the study are subsequently analyzed based on the initial group they were assigned to and there is a significant improvement in outcome of the new drug. What analysis most appropriately describes this trial?

A. Per protocol
B. As treated
C. Non-inferiority
D. Intention to treat (Correct Answer)
E. Modified intention to treat

Explanation: ***Intention to treat*** - **Intention-to-treat (ITT)** analysis includes all participants randomized to a treatment arm, regardless of whether they completed the intervention or switched treatments, reflecting a real-world scenario and preserving randomization benefits. - This approach minimizes bias from **loss to follow-up** or **treatment crossovers** and provides a more conservative estimate of treatment effect. *Per protocol* - **Per-protocol analysis** only includes participants who completed the study exactly as planned without any deviations. - This method is susceptible to **selection bias** because it excludes patients who may have experienced adverse events or treatment failures, potentially overestimating treatment efficacy. *As treated* - **As-treated analysis** analyzes patients based on the actual treatment received, rather than the treatment to which they were randomized. - This approach can introduce **confounding** and selection bias, as patients who switch treatments may do so for reasons related to their prognosis or treatment response. *Non-inferiority* - A **non-inferiority trial** design aims to show that a new treatment is not appreciably worse than an active control, rather than proving superiority. - This describes a **type of study design** or hypothesis, not an analysis method for handling patient data after randomization with non-adherence. *Modified intention to treat* - A **modified intention-to-treat (mITT)** analysis typically excludes a small, predefined group of patients from the ITT population, such as those who never received any study drug or were found to be ineligible after randomization. - While similar to ITT, it involves specific exclusions that are not described in this scenario, where all randomized patients were analyzed **based on initial assignment**.

Question 6: You are reading through a recent article that reports significant decreases in all-cause mortality for patients with malignant melanoma following treatment with a novel biological infusion. Which of the following choices refers to the probability that a study will find a statistically significant difference when one truly does exist?

A. Type II error
B. Type I error
C. Confidence interval
D. p-value
E. Power (Correct Answer)

Explanation: ***Power*** - **Power** is the probability that a study will correctly reject the null hypothesis when it is, in fact, false (i.e., will find a statistically significant difference when one truly exists). - A study with high power minimizes the risk of a **Type II error** (failing to detect a real effect). *Type II error* - A **Type II error** (or **beta error**) occurs when a study fails to reject a false null hypothesis, meaning it concludes there is no significant difference when one actually exists. - This is the **opposite** of what the question describes, which asks for the probability of *finding* a difference. *Type I error* - A **Type I error** (or **alpha error**) occurs when a study incorrectly rejects a true null hypothesis, concluding there is a significant difference when one does not actually exist. - This relates to the **p-value** and the level of statistical significance (e.g., p < 0.05). *Confidence interval* - A **confidence interval** provides a range of values within which the true population parameter is likely to lie with a certain degree of confidence (e.g., 95%). - It does not directly represent the probability of finding a statistically significant difference when one truly exists. *p-value* - The **p-value** is the probability of observing data as extreme as, or more extreme than, that obtained in the study, assuming the null hypothesis is true. - It is used to determine statistical significance, but it is not the probability of detecting a true effect.

Question 7: A health system implements a new sepsis protocol across 20 hospitals. A researcher plans to evaluate effectiveness using a stepped-wedge cluster randomized design where hospitals sequentially adopt the protocol every 3 months. She calculates sample size based on individual patient outcomes (mortality) needing 2,000 patients total. The biostatistician identifies a critical error. Evaluate what modification is needed.

A. Adjust for multiple time periods using Bonferroni correction
B. Use hospital-level outcomes instead of patient-level outcomes as unit of analysis
C. Increase alpha to 0.10 to account for cluster randomization reducing power
D. Include random effects for both hospital and time period in power calculation
E. Account for intra-cluster correlation coefficient (ICC) requiring substantial sample size inflation (Correct Answer)

Explanation: ***Account for intra-cluster correlation coefficient (ICC) requiring substantial sample size inflation*** - In cluster-randomized designs, observations within the same cluster (hospital) are not independent; the **Intra-cluster Correlation Coefficient (ICC)** quantifies this correlation and must be used to calculate a **design effect**. - Neglecting the ICC leads to an **underpowered study** because the effective sample size is smaller than the total number of individual patients measured. *Adjust for multiple time periods using Bonferroni correction* - **Bonferroni correction** is used to control for **Type I error** when performing multiple independent hypothesis tests, not for determining sample size in nested longitudinal designs. - While the stepped-wedge design involves multiple time points, the primary analysis typically uses a **single model** (e.g., GEE or GLMM) that accounts for time as a fixed effect. *Use hospital-level outcomes instead of patient-level outcomes as unit of analysis* - While the hospital is the **unit of randomization**, using hospital-level means as the unit of analysis simplifies the data and causes a significant loss of **statistical information** and precision. - Modern biostatistical methods utilize **multilevel modeling** to maintain the richness of patient-level data while adjusting for the cluster-level randomization. *Include random effects for both hospital and time period in power calculation* - While random effects are important for the **analysis phase**, the "critical error" identified in the prompt refers to the initial failure to inflate the sample size based on **clustering (ICC)**. - Power calculations for stepped-wedge designs are complex and certainly involve time parameters, but **ICC-based inflation** is the most fundamental adjustment required when moving from individual to cluster randomization. *Increase alpha to 0.10 to account for cluster randomization reducing power* - Increasing the **alpha level** (significance threshold) is not a standard or scientifically acceptable method to compensate for the loss of power due to **clustering**. - Standard practice mandates maintaining an **alpha of 0.05** while appropriately increasing the **sample size** or number of clusters to reach the desired power (usually 80-90%).

Question 8: A 41-year-old research fellow designs a non-inferiority trial comparing oral to IV antibiotics for osteomyelitis. She sets the non-inferiority margin at 10% (cure rate difference), expects 85% cure in both groups, and calculates 300 patients per arm for 80% power with α=0.025 (one-sided). Her mentor suggests this underestimates required sample size. Evaluate the mentor's concern.

A. Correct; non-inferiority trials require larger samples than superiority trials for equivalent power (Correct Answer)
B. Incorrect; non-inferiority trials actually require smaller samples due to less stringent hypotheses
C. Correct; dropout rates in antibiotic trials necessitate 20% inflation of calculated sample size
D. Incorrect; the calculation appropriately uses one-sided alpha for non-inferiority testing
E. Correct; the margin should be set at 5% requiring doubling of sample size

Explanation: ***Correct; non-inferiority trials require larger samples than superiority trials for equivalent power*** - **Non-inferiority trials** are designed to exclude a difference greater than a pre-specified margin, which typically requires a **larger sample size** than superiority trials investigating the same outcome. - Because we are proving that the new treatment is "not much worse" (rather than "better"), the **statistical threshold** often necessitates higher enrollment to achieve adequate **power**. *Incorrect; the calculation appropriately uses one-sided alpha for non-inferiority testing* - While it is true that **non-inferiority testing** uses a **one-sided alpha (0.025)**, this does not negate the fact that such trials inherently require more participants. - The mentor's concern is about the **total N**, which remains insufficient despite using the correct one-sided alpha convention. *Correct; the margin should be set at 5% requiring doubling of sample size* - There is no universal rule that the **non-inferiority margin** must be 5%; it is determined by **clinical significance** and regulatory standards for the specific condition. - While a 5% margin would indeed increase the sample size, the 10% margin is often standard in **antibiotic trials** for osteomyelitis. *Incorrect; non-inferiority trials actually require smaller samples due to less stringent hypotheses* - This is a common misconception; non-inferiority trials are actually more demanding because the **null hypothesis** assumes the treatments are different (inferior). - Disproving **inferiority** within a tight **margin (delta)** is statistically more intensive than proving a treatment is superior to a placebo. *Correct; dropout rates in antibiotic trials necessitate 20% inflation of calculated sample size* - While **attrition bias** is a concern, there is no fixed rule that every trial needs a **20% inflation** factor. - The mentor's concern is specifically about the **base calculation** and the statistical nature of non-inferiority designs rather than just the **dropout rate**.

Question 9: A pharmaceutical company tests a new antidepressant in 500 patients (250 per arm) and finds a 2-point improvement on a 52-point depression scale compared to placebo (p=0.04). The study was originally powered to detect a 4-point difference. The company seeks FDA approval citing statistical significance. Analyze the regulatory and scientific implications.

A. Approval warranted; the study achieved statistical significance with adequate power
B. Approval not warranted; observed effect is smaller than pre-specified clinically meaningful difference (Correct Answer)
C. Approval warranted; post-hoc power analysis shows adequate power for 2-point difference
D. Approval not warranted; the study was underpowered for the observed effect size
E. Approval warranted if sensitivity analyses confirm robustness of findings

Explanation: ***Approval not warranted; observed effect is smaller than pre-specified clinically meaningful difference*** - Although the result is **statistically significant** (p=0.04), the observed 2-point improvement is only half of the **pre-specified 4-point difference** deemed clinically relevant. - Regulatory bodies like the **FDA** prioritize **clinical significance** over mere p-values, ensuring that a drug provides a meaningful benefit to patients' lives. *Approval warranted; the study achieved statistical significance with adequate power* - Statistical significance does not automatically justify approval if the **effect size** is too small to provide a real therapeutic advantage. - Being **powered** for a 4-point difference means the study was designed to reliably detect a larger effect; a smaller effect may be a result of **over-testing** or limited clinical utility. *Approval not warranted; the study was underpowered for the observed effect size* - If a study finds a significant result (p < 0.05), it is by definition **sufficiently powered** to detect that specific effect size in that sample. - The issue here is not **power** or sample size, but rather the **magnitude of effect** failing to meet the pre-defined target for clinical relevance. *Approval warranted if sensitivity analyses confirm robustness of findings* - **Sensitivity analyses** help confirm that results are not driven by outliers, but they cannot transform a **clinically trivial** difference into a meaningful one. - Even a robust, consistent 2-point difference remains below the **Minimum Clinically Important Difference (MCID)** set at 4 points. *Approval warranted; post-hoc power analysis shows adequate power for 2-point difference* - **Post-hoc power analysis** is generally considered scientifically flawed and redundant once the **p-value** is already known. - Demonstrating power for a 2-point difference does not erase the fact that the drug failed to meet the **threshold of efficacy** defined by the researchers at the start.

Question 10: A meta-analysis of 5 previous trials testing a surgical technique shows a pooled effect size of 15% complication reduction (from 20% to 17%, p=0.30, I²=0%). An investigator wants to design a definitive trial. She calculates that 1,200 patients per arm would provide 80% power to detect this 3% absolute difference. Analyze whether this sample size is justified.

A. Not justified; a 3% absolute reduction lacks clinical significance for most surgical outcomes (Correct Answer)
B. Not justified; the high I² indicates substantial heterogeneity making pooled estimates unreliable
C. Justified; the non-significant p-value indicates need for larger, definitive trial
D. Justified only if cost-effectiveness analysis supports the intervention
E. Justified; the meta-analysis provides the best estimate of true effect size

Explanation: ***Not justified; a 3% absolute reduction lacks clinical significance for most surgical outcomes*** - A **3% absolute risk reduction** (from 20% to 17%) might be statistically detectable but is often considered too minor to justify the **cost, risks, or resources** of a new surgical technique. - Investigator's focus should be on whether the **Minimal Clinically Important Difference (MCID)** is met; designing a massive trial to find a tiny effect is often a waste of resources. *Justified; the meta-analysis provides the best estimate of true effect size* - While meta-analyses are high in the evidence hierarchy, a **p-value of 0.30** indicates the pooled effect is not statistically significant and may be due to **random chance**. - Using a non-significant, potentially **spurious effect size** to power a large trial leads to a high risk of a **futile study**. *Justified only if cost-effectiveness analysis supports the intervention* - Cost-effectiveness is a secondary consideration that follows the determination of **clinical efficacy and safety**. - Even if cost-effective, the trial remains unjustified if the **sample size calculation** is based on statistically unreliable (p=0.30) data. *Not justified; the high I² indicates substantial heterogeneity making pooled estimates unreliable* - This statement is factually incorrect as the prompt states **I²=0%**, which indicates **no observed statistical heterogeneity** among the trials. - **I²=0%** suggests that the results of the specific trials were consistent with each other, though all remained non-significant. *Justified; the non-significant p-value indicates need for larger, definitive trial* - A non-significant p-value in a meta-analysis does not automatically mandate a larger trial; it suggests the **null hypothesis** cannot be rejected. - Planning a trial based on a **3% difference** that failed to reach significance (p=0.30) ignores the likelihood that the **true effect size** might be zero.

Want unlimited practice?

Get full access to all questions, explanations, and performance tracking.

Start For Free