info analytics ozaki final

What does subgroup analysis examine in clinical research?

Examining treatment effects in specific subsets of a study population.

Why must results from subgroup analyses be interpreted with caution?

Multiple comparisons increase the risk of type I error.

What does a p-value for interaction tell you in a subgroup analysis?

Whether treatment effects differ between subgroups.

Do p-values for interaction measures the effectiveness of a treatment compared with the control group?

No.  Effectiveness in each subgroup is measured by its own HR and CI; the interaction p-value asks if subgroup effects differ.

How do patients aged ≥75 years respond to LCZ696 compared with younger patients?

Show similar treatment effects based on hazard ratios and interaction p-values.

What do subgroup results show about differences in LCZ696 effects by race?

No differences; the interaction p-value is not significant.

Why are confidence intervals wider in the subgroup of Black patients taking LCZ696?

The sample size is small and causes imprecision.

What is the primary goal of conducting a noninferiority trial?

To show a new treatment is not unacceptably worse than a standard therapy.

What does a noninferiority trial NOT attempt to demonstrate?

Prove a drug is better.

What does the noninferiority margin represent within a trial design?

The maximum acceptable loss of efficacy.

How does selecting a wide noninferiority margin affect the trial?

Makes the trial easier to meet but less clinically meaningful.

Which confidence intervals (margin 1.20) support noninferiority in a trial?

HR 1.08 (1.02–1.18) and HR 0.96 (0.90–1.05).

What did the polypill trial demonstrate when considering its CI and p-value results?

Both noninferiority (CI < margin) and superiority (CI < 1.0, p=0.02).

What is the primary purpose of using composite outcomes in a clinical study?

To combine endpoints and increase study efficiency.

Why can composite outcomes be misleading when minor components dominate the results?

They may exaggerate the perceived treatment benefit.

Why does variability in clinical importance among composite components create interpretation problems?

The results become misleading and harder to interpret.

How can composite outcomes obscure clinically meaningful results?

They can hide opposite effects among individual components.

What is incorrect about stating that factorial trials randomize patients to only two interventions?

The incorrect part is that factorial trials involve more than two interventions or factors. In a typical 2×2 factorial design, patients are randomized to all combinations of two independent interventions, creating four groups:

1. Intervention A + Intervention B

2. Intervention A + No B

3. No A + Intervention B

4. No A + No B

What does a factorial design create when assigning patients to multiple interventions?

Separate randomized groups for each treatment combination.

Do factorial trials always require larger sample sizes than standard trials?

A factorial trial does not always require a larger sample size than a standard two-arm trial. It depends on the goals of the study and how the sample size is calculated.

Can factorial trial designs incorporate non-drug interventions?

Yes, Factorial trials can include non-drug interventions, such as behavioral programs, lifestyle changes, or devices, alongside or instead of drugs.

Why does increasing the sample size alone fail to eliminate confounding?

Increasing the sample size reduces random error but does not remove confounding, because confounding is a systematic bias.

How does matching help control confounding during trial design?

By controlling confounding at the design stage.

How does randomization reduce confounding in clinical trials?

By balancing known and unknown confounders.

Can residual confounding still occur after statistical adjustment procedures?

Yes. Residual confounding can remain if confounders are unmeasured, mismeasured, or modeled incorrectly.

How does propensity score matching help in observational studies?

By reducing confounding through covariate balancing.

Which aspects of a study can misclassification affect?

The exposure, the outcome, or both.

Why is nondifferential outcome misclassification still a threat to validity?

It biases results and is not harmless.

What type of error decreases accuracy in a study measurement?

Systematic error.

What type of error decreases precision in a study measurement?

Random error.

Do Kaplan–Meier curves require every patient to have identical follow-up time?

No. Kaplan–Meier curves do not require identical follow-up times. They handle censored data, so patients can enter and leave the study at different times or be lost to follow-up.

Do survival curves serve only descriptive purposes in clinical trials?

No. Survival curves are descriptive and can also be used to compare groups statistically.

What information do Kaplan–Meier curves primarily display?

Event proportions over time and patterns of event occurrence.

Prevalence measures what?

Prevalence measures both new and existing cases at a specific point in time.

What do incidence proportion and incidence rate both measure?

They both measure new cases but differ because incidence proportion uses persons at risk and incidence rate uses person-time.

Is prevalence useful for identifying risk factors?

Prevalence is less useful for identifying causal risk factors because it mixes new and existing cases.

What does attributable risk represent?

The difference in disease incidence between exposed and unexposed groups.

How is population attributable risk estimated?

Population attributable risk equals attributable risk multiplied by exposure prevalence.

What does a β coefficient represent?

It represents the strength and direction of association between a predictor and an outcome.

What does a hazard ratio indicate?

It indicates relative event rates over time accounting for censoring.

What does an odds ratio indicate?

It compares the odds of an outcome between exposed and unexposed groups.

What does a relative risk indicate?

It compares the probability of an outcome between exposed and unexposed groups.

When is HR used?

HR is used in time-to-event analyses.

When is OR used?

OR is used when risk cannot be directly measured, such as in case-control studies.

When is RR used?

RR is used in cohort studies where incidence can be measured.

If OR = 2.5 in smokers vs nonsmokers, what does that mean?

Smokers have 2.5 times higher odds of pneumonia than nonsmokers.

What percent increase in odds does OR = 2.5 represent?

It represents a 150 percent increase in odds.

If RR = 0.6 for depression in diabetes, what does it mean?

Diabetes is associated with a 40 percent lower risk of depression.

What range does the CI 0.4–0.8 suggest?

It suggests the reduction in risk is between 20 percent and 60 percent.

What does HR compare?

It compares the rate at which events occur between two groups over time.

Why is OR used in case-control studies?

Because actual risk cannot be measured since incidence is not known.

Can RR be estimated in a case-control study?

RR cannot be estimated because incidence is not measured.

Can RR and OR be estimated in cohort studies?

Both OR and RR can be estimated because incidence is measured.

Why is HR preferred over RR in survival analysis?

HR accounts for censoring and varying follow-up.

Can RR be estimated using logistic regression?

RR is not estimated from logistic regression; logistic regression estimates OR.

What does randomization do?

It assigns participants so each has equal chance of being in any group, reducing confounding.

What does restriction do?

It limits study participants to reduce confounders.

What does matching do?

It selects comparison subjects with similar characteristics to reduce confounding.

What does stratification do?

It compares outcomes within subgroups that share similar risk.

What does standardization do?

It mathematically adjusts crude rates to equalize group characteristics.

What does multivariable adjustment do?

It adjusts for many confounders simultaneously via modeling.

What is propensity score matching?

It matches individuals with similar baseline characteristics to balance covariates.

Is PSM used in experimental studies?

No. Propensity score matching (PSM) is primarily used in observational studies to balance baseline characteristics between treatment groups. In experimental studies, randomization already controls for confounding, so PSM is generally unnecessary.

Does PSM guarantee causal inference?

No, it does not guarantee validity of causal inference.

Does PSM remove unmeasured confounding?

No, it only addresses measured confounders.

What does misclassification bias from coding errors affect?

It reduces internal validity. Misclassification bias happens when people are put in the wrong exposure or outcome group due to coding errors. It can make results seem weaker, stronger, or just wrong, leading to incorrect conclusions and misleading recommendations.

Does confounding by indication affect internal or external validity?

It reduces internal validity because the treatment is given for a reason related to the outcome, making it harder to determine the true effect of the intervention.

Are results from VA or Medicare generalizable to all populations?

They may not be generalizable to the entire population.

Are self-reported adherence measures objective?

No, they are subjective and prone to bias.

What is a strength of pharmacy claims data?

They allow large-scale evaluation but may miss OTC or non-insurance purchases.

Do dispensing records confirm ingestion?

No, they confirm fills but not actual medication use.

What are spontaneous reporting systems used for?

They help detect unexpected adverse drug reactions early.

What is a limitation of spontaneous reporting systems?

They can contain missing or duplicate reports that distort signals.

Are spontaneous reporting systems designed to evaluate drug effectiveness?

No, they focus on safety, not effectiveness.

Are EHR-based safety studies free of bias?

No, they may contain measurement and recording errors.

What bias may affect EHR drug safety studies?

Confounding due to differences in patients or prescribing patterns.

What is a major limitation of EHR data?

Missing or incomplete data can cause misclassification bias.

What does interrupted time series evaluate?

It evaluates outcome trends before and after an intervention.

Can ITS distinguish immediate vs gradual changes?

Yes, it detects level changes and trend changes.

Is ITS useful for large-scale non-randomizable policies?

Yes, it is useful when randomization is not feasible.

Does ITS adjust for all simultaneous events?

No, it does not automatically adjust for other events happening at the same time.

What does a before-and-after study measure?

It measures outcomes before and after an intervention in the same group.

Does a before-and-after study capture long-term effects?

It is not ideal for long-term evaluation.

What external factors can influence before-and-after results?

Policy changes, seasonal effects, or concurrent events.

How does ITS improve on before-and-after?

ITS uses multiple time points to better separate real changes from random variation.

Is a composite outcome always a better indicator of treatment effect than individual outcomes?

No, it can sometimes obscure the effects of individual outcomes.

When a composite outcome is used, are the individual components typically not analyzed separately?

No, individual components are often analyzed separately to assess their clinical relevance.

When are composite outcomes particularly useful in clinical studies?

They are useful when the primary endpoints are expected to occur infrequently.

Should the heterogeneity of effects across individual components of a composite outcome be assessed?

Yes, assessing heterogeneity helps interpret the results accurately.

Is Edoxaban noninferior to warfarin for the efficacy outcome?

Edoxaban was noninferior to warfarin because the upper bound of the 95% CI (1.31) was below the noninferiority margin (1.38).

Is Edoxaban noninferior to warfarin for the efficacy outcome?

Edoxaban was not noninferior to warfarin because the upper bound of the 95% CI (1.91) exceeded the noninferiority margin (1.38). Edoxaban’s safety outcome would be considered inferior of the new drug compared to the standard because the entire CI > 1.0.