drug lit II exam II

In case-control studies, why can't we directly calculate risk (cumulative incidence)?

Option A

There is no control group in case-control studies

Option B

Exposure is always random in case-control studies

Option C

They measure incidence directly but not risk

Option D

Cases and controls are selected based on outcome status, not exposure in the population

Option D

Cases and controls are selected based on outcome status, not exposure in the population

Which randomization method is most appropriate when participants have different baseline risk profiles?

Option A

Simple randomization

Option B

Stratified randomization

Option C

Block randomization

Option D

Cluster randomization

Option B

Stratified randomization

Which design is best for establishing temporality between exposure and outcome?

Option A

Cross-sectional study

Option B

Descriptive study

Option C

Cohort study (prospective)

Option D

Case-control study

Option C

Cohort study (prospective)

In an intention-to-treat analysis, how are participants analyzed?

Option A

Analyses exclude any non-adherent participants

Option B

Only participants who completed the study are analyzed

Option C

According to the group they were randomized to, regardless of adherence

Option D

According to the treatment they actually received

Option C

According to the group they were randomized to, regardless of adherence

Which randomization method helps ensure equal numbers of participants across treatment groups throughout enrollment?

Option A

Cluster randomization

Option B

Stratified randomization

Option C

Block randomization

Option D

Simple randomization

Option C

Block randomization

What is the key difference between a cohort study and a case-control study?

Option A

Cohort studies start with exposure and follow outcomes; case-control studies start with outcome and look backward for exposure

Option B

Cohort studies use only one outcome; case-control studies use multiple outcomes

Option C

Cohort studies are retrospective only; case-control studies are prospective only

Option D

Case-control studies randomize participants; cohort studies do not

Option A

Cohort studies start with exposure and follow outcomes; case-control studies start with outcome and look backward for exposure

Which endpoint type directly reflects how a patient feels, functions, or survives?

Option A

Composite endpoint

Option B

Surrogate endpoint

Option C

Clinical endpoint

Option D

Endpoint without clinical relevance

Option C

Clinical endpoint

Which term describes a trial where the same participants receive both the treatment and control at different times?

Option A

Factorial design

Option B

Crossover design

Option C

Parallel design

Option D

Cluster design

Option B

Crossover design

Which study design is most efficient for studying rare outcomes or diseases within a defined cohort?

Option A

Cross-sectional study

Option B

Randomized controlled trial

Option C

Nested case-control study

Option D

Descriptive study

Option C

Nested case-control study

What is the main purpose of blinding in a clinical trial?

Option A

To increase sample size

Option B

To speed up data collection

Option C

To reduce measurement bias and placebo effects

Option D

To ensure participants follow instructions

Option C

To reduce measurement bias and placebo effects

What is power analysis?

Determining the sample size needed to detect a clinically meaningful effect at a specified alpha and power.

When should sample size estimation be performed?

A priori, before the study begins.

What determines which sample size calculation formula is used?

The type of primary outcome data and hypothesis test.

What four inputs are used in power analysis?

Variability, effect size, significance level (alpha), and desired power (1 − beta).

How is variability measured for continuous outcomes?

Population standard deviation.

How is variability assessed for dichotomous outcomes?

Event proportions in the treatment and control groups.

How does increasing variability affect required sample size?

Increases sample size.

How does decreasing variability affect required sample size?

Decreases sample size.

What is effect size?

The clinically meaningful treatment difference a study aims to detect.

How is effect size measured for continuous vs. dichotomous outcomes?

Difference between means vs. difference between proportions.

How does decreasing effect size affect required sample size?

Increases sample size.

How does increasing alpha affect required sample size?

Decreases sample size.

How does increasing desired power affect required sample size?

Increases sample size.

What alpha and power levels are typically used?

Alpha = 0.05; power = 80–90% (90% commonly used in clinical trials).

What is the target population?

The population to which researchers want to generalize their findings.

What is the source population?

The population from which study subjects are drawn.

What is the sampling frame?

The defined subset of the source population from which participants are selected.

What is the study sample?

The participants actually included in the study.

What is a representative sample?

A sample that accurately reflects characteristics of the target population.

How do sample size and heterogeneity affect sampling error?

Larger samples generally reduce error; greater heterogeneity increases it.

What is the main concern with nonrandom sampling?

Selection bias and limited generalizability to the target population.

What are the four probability sampling methods?

Simple random, systematic, stratified random, and cluster.

What are the three nonprobability sampling methods?

Convenience, quota, and consecutive.

How does simple random sampling select participants?

Randomly selects individuals from a complete sampling frame.

What is systematic sampling?

Randomly select a starting point, then choose every kth individual.

How is the systematic sampling interval calculated?

N/n = sampling frame size / desired sample size.

What is a major disadvantage of systematic sampling?

Periodic patterns can bias selection; selections may be predictable.

What is stratified random sampling?

Divide into homogeneous subgroups and randomly sample individuals from each.

What is cluster sampling?

Randomly select natural groups, then include all eligible individuals in selected groups (single-stage).

What is the key difference between stratified and cluster sampling?

Stratified: sample individuals from every subgroup. Cluster: select groups and include their members.

What is convenience sampling?

Select participants based on accessibility or availability.

What is quota sampling?

Set subgroup proportions, then recruit nonrandomly until quotas are filled.

What is consecutive sampling?

Recruit every eligible subject encountered sequentially during the study period.

What is a limitation of consecutive sampling?

Potential temporal or seasonal bias.

A researcher randomly selects clinics and enrolls all eligible patients within them. Which method?

Cluster sampling.

A researcher enrolls every eligible patient as they arrive at a clinic. Which method?

Consecutive sampling.

What distinguishes experimental from observational studies?

Experimental: researcher assigns intervention.

Observational: researcher observes exposure.

What distinguishes an RCT from a quasi-experimental study?

RCTs randomly assign interventions; quasi-experimental studies do not.

What distinguishes descriptive from analytical observational studies?

Descriptive: no comparison group.

Analytical: compares groups to assess associations.

What is the hierarchy of evidence from highest to lowest?

Systematic reviews/meta-analyses of RCTs → RCTs → cohort → case-control → cross-sectional → case reports.

What is a parallel RCT design?

Participants remain in their assigned treatment arm throughout the study.

What is a crossover RCT design?

Participants switch treatments and serve as their own controls.

Why is a washout period used in crossover trials?

To allow the previous treatment's effects to dissipate.

What is a carryover effect?

Residual effects of a previous treatment influence the next treatment period.

What is a period effect?

Outcomes change over time independently of treatment.

When is a crossover design most appropriate?

Stable disease with reversible, transient treatment effects.

What is a factorial RCT design?

Tests multiple interventions individually and in combination.

How many treatment groups are in a 2×2 factorial design?

Four: A only, B only, A+B, and neither.

What can factorial designs assess besides individual treatment effects?

Interactions or synergy between interventions.

How can strict eligibility criteria affect generalizability?

They improve study control but may limit applicability to other patients.

What is the most common RCT treatment allocation ratio?

1:1, although unequal ratios such as 1:2 or 1:3 are possible.

What is the purpose of a run-in period?

Screen nonadherent/ineligible patients, assess response, or wash out previous drugs.

How can run-in periods bias trial results?

Overestimate treatment benefits and underestimate adverse effects.

What is the primary purpose of randomization?

Balance patient characteristics and reduce confounding between treatment groups.

What is a limitation of simple randomization in small trials?

Unequal treatment-group sizes.

What is block randomization?

Random assignment within blocks to maintain balanced treatment-group sizes.

What is stratified randomization?

Randomization within baseline-characteristic subgroups to balance risk factors.

What is cluster randomization?

Random assignment of entire groups, such as clinics or hospitals.

When is a placebo control appropriate?

When no effective established treatment exists.

When should an active comparator be used?

When standard therapy exists and withholding treatment would be unethical.

What is the difference between clinical and surrogate endpoints?

Clinical: patient benefit/harm.

Surrogate: intermediate predictor of clinical outcomes.

Give examples of surrogate versus clinical endpoints.

Surrogate: LDL-C, HbA1c, BP.

Clinical: MI, stroke, survival.

What is a composite endpoint?

Two or more outcomes combined into one endpoint, such as cardiovascular death, MI, or stroke.

What makes a composite endpoint appropriate?

Components are related, similarly important, and expected to respond in the same direction.

What does a superiority trial aim to demonstrate?

A statistically significant difference between intervention and comparator.

What does a noninferiority trial aim to demonstrate?

The intervention is not worse than the comparator beyond a prespecified margin.

What is required to demonstrate noninferiority?

The confidence interval must exclude effects worse than the noninferiority margin.

What does an equivalence trial aim to demonstrate?

Treatment effects fall within prespecified upper and lower equivalence margins.

What is required to establish equivalence?

The entire confidence interval falls within both equivalence margins.

What is the difference between noninferiority and equivalence?

Noninferiority excludes unacceptable worsening; equivalence excludes meaningful differences in either direction.

Who knows treatment assignment in an open-label study?

Patients, investigators, and analysts.

Who is blinded in single-, double-, and triple-blind trials?

Single: patient; double: patient and investigator; triple: patient, investigator, and analyst (per lecture).

What are the main benefits of blinding?

Reduces measurement bias, placebo effects, and differences in adherence.

How are participants analyzed in intention-to-treat (ITT)?

By original randomization, regardless of adherence or treatment switching.

What is the main advantage of ITT analysis?

Preserves randomization and often gives a conservative treatment-effect estimate.

How are participants analyzed in per-protocol analysis?

Only sufficiently adherent participants, by their original assigned group.

How are participants analyzed in as-treated analysis?

According to treatment actually received, regardless of original assignment.

Which analysis methods do not preserve randomization?

Per-protocol and as-treated.

What distinguishes efficacy from effectiveness trials?

Efficacy: controlled conditions.

Effectiveness: real-world clinical practice.

What is clinical equipoise?

Genuine uncertainty about whether an intervention is beneficial or harmful.

What is the role of a DSMB?

Independently monitors trial safety/integrity and recommends early stopping when warranted.

What is a one-group pre-post study?

Measures outcomes before and after an intervention without random assignment.

What are four major threats to pre-post study validity?

History, maturation, testing effects, and regression toward the mean.

Which tests can analyze paired pre-post outcomes?

Paired t-test (normal continuous), Wilcoxon signed-rank (non-normal/ordinal), McNemar (binary).

What is an interrupted time-series design?

Multiple measurements before and after an intervention to assess changes in level or trend.

What analysis is used for interrupted time-series studies?

Segmented regression.

What is a major limitation of interrupted time-series studies?

Concurrent events (history effects) and cyclical trends can influence results.

What is the difference between a case report and case series?

Case report: one patient.

Case series: multiple patients.

What is the main limitation of descriptive ecological studies?

Group-level associations cannot establish individual-level exposure-outcome relationships.

What is a cohort study?

Groups defined by exposure are followed to compare subsequent outcome occurrence.