NAPLEX Flashcards: Study Design And Bias

What this deck covers

This deck focuses on Study Design And Bias, giving you a quick way to review the definitions, rules, and examples that matter most for NAPLEX.

How to use these flashcards

Work through these flashcards in short sessions. Try to answer each prompt before flipping the card, then revisit any cards you miss until the explanation feels automatic.

Flashcard 1: What bias occurs when baseline differences between groups distort the treatment effect estimate?

Answer: Selection bias. Non-random group differences at baseline can confound the observed effect, leading to inaccurate conclusions about treatment impact.

Flashcard 2: Identify the correct formula for relative risk using risks in exposed and unexposed groups.

Answer: $RR=\frac{Risk_{exposed}}{Risk_{unexposed}}$. This formula quantifies the multiplicative increase in risk attributable to exposure by comparing group incidences.

Flashcard 3: Which design is most efficient for studying rare exposures (for example, uncommon drug exposure)?

Answer: Cohort study. Selecting exposed individuals first enables following them to observe outcomes, ideal for uncommon exposures in populations.

Flashcard 4: Identify the correct formula for odds ratio using odds in exposed and unexposed groups.

Answer: $OR=\frac{Odds_{exposed}}{Odds_{unexposed}}$. It expresses the ratio of exposure probabilities given the outcome, useful for approximating risk in certain designs.

Flashcard 5: What is blinding (masking) in a clinical trial?

Answer: Keeping group assignment unknown to reduce bias. By masking participants, investigators, or both, it minimizes performance and detection biases that could skew results.

Flashcard 6: What study design randomly assigns participants to intervention versus control groups?

Answer: Randomized controlled trial (RCT). This design minimizes selection bias by randomly allocating participants, ensuring comparable groups for evaluating intervention effects.

Flashcard 7: What bias occurs when loss to follow-up differs between groups and is related to outcome?

Answer: Attrition bias. When dropout rates vary by group and relate to the outcome, it can introduce systematic error in effect estimates.

Flashcard 8: What measure of association is typically estimated directly in a case-control study?

Answer: Odds ratio. Since sampling is by outcome, it calculates odds of exposure in cases versus controls, approximating relative risk for rare events.

Flashcard 9: What bias occurs when outcome assessment differs systematically between study groups?

Answer: Detection (ascertainment) bias. Systematic variations in how outcomes are measured or diagnosed between groups can lead to over- or underestimation of effects.

Flashcard 10: What bias occurs when co-interventions or adherence differ between groups after assignment?

Answer: Performance bias. Unequal provision of additional treatments or differences in compliance post-assignment can confound the true intervention effect.

Flashcard 11: Which bias occurs when selecting participants makes the sample unrepresentative of the target population?

Answer: Sampling (selection) bias. Non-random or biased participant selection can lead to results that do not generalize to the broader intended population.

Flashcard 12: What measure of association is typically estimated directly in a cohort study?

Answer: Relative risk (risk ratio). In cohorts, incidence in exposed and unexposed groups directly yields this measure of how exposure modifies outcome risk.

Flashcard 13: What is a placebo effect in clinical trials?

Answer: Outcome change due to expectations, not active treatment. Psychological expectations of benefit can produce perceived or actual improvements, necessitating placebo controls for isolation.

Flashcard 14: What is the primary purpose of randomization in an RCT?

Answer: Balance confounders between groups on average. Randomization ensures that known and unknown confounders are distributed evenly, allowing causal inference about the intervention.

Flashcard 15: What bias is introduced when only statistically significant studies are likely to be published?

Answer: Publication bias. Selective reporting of positive findings skews the literature, overestimating effects in systematic reviews and meta-analyses.

Flashcard 16: What study type synthesizes results of multiple studies using quantitative pooling?

Answer: Meta-analysis. This method combines data from several studies statistically, increasing precision and power to detect effects across heterogeneous evidence.

Flashcard 17: What is confounding in clinical research?

Answer: A third factor distorts the exposure–outcome association. An extraneous variable linked to both exposure and outcome can create a spurious or masked relationship in the analysis.

Flashcard 18: What is a confounder's required relationship to exposure and outcome (not on causal pathway)?

Answer: Associated with exposure and independently affects outcome. To confound, the factor must correlate with exposure, influence outcome independently, and not mediate the causal path.

Flashcard 19: Which analysis principle keeps participants in their randomized groups regardless of adherence?

Answer: Intention-to-treat analysis. This approach preserves randomization benefits, providing a pragmatic estimate of effectiveness in real-world adherence scenarios.

Flashcard 20: Which design is most efficient for studying rare diseases or outcomes?

Answer: Case-control study. Starting with cases allows efficient recruitment for infrequent outcomes, making it suitable for rare disease investigations.

Flashcard 21: What is allocation concealment in a randomized trial?

Answer: Hiding the upcoming assignment from enrollers. It prevents investigators from influencing enrollment based on anticipated group assignment, thus preserving randomization integrity.

Flashcard 22: What is recall bias in observational studies?

Answer: Differential accuracy of reported past exposure. Participants with the outcome may remember or report exposures differently than controls, leading to biased association measures.