- A cohort study followed 2,000 healthcare workers during a pandemic. Among the 800 workers who consistently wore N95 masks, 20 developed infection. Among the 1,200 workers who wore surgical masks, 60 developed infection. If mask-wearing behavior was associated with other protective behaviors, and this resulted in a 40% overestimate of the protective effect of N95 masks, what is the bias-adjusted relative risk?
Answer: Observed RR = 0.50, bias-adjusted RR = 0.70, indicating N95 masks remain protective but less than observed
First calculate observed RR: N95 group = 20/800 = 2.5%, Surgical mask group = 60/1,200 = 5%, RR = 2.5%/5% = 0.50. A 40% overestimate of protective effect means the observed protective effect (1 - 0.50 = 0.50 or 50% reduction) is 40% too large. True protective effect = 0.50/1.40 = 0.357 (35.7% reduction). Therefore, bias-adjusted RR = 1 - 0.357 = 0.643 ≈ 0.70. This indicates N95 masks remain protective but the effect is smaller than observed. Choice B incorrectly makes the effect stronger. Choice C gives too small a protective effect. Choice D incorrectly suggests harm.
- A pharmaceutical company reports that their new medication reduces heart attack risk by 50% compared to standard care (RR = 0.5). The baseline heart attack rate in the control group was 2% over 5 years. What is the absolute risk reduction (risk difference) provided by this medication?
Answer: 1 percentage point reduction
When you encounter questions about treatment effects, it's crucial to distinguish between relative measures (like relative risk) and absolute measures (like risk difference). Both tell important but different stories about a treatment's impact.
To find the absolute risk reduction, you need to calculate the actual difference in event rates between groups. Start with the control group's baseline risk of 2% over 5 years. With a relative risk of 0.5, the treatment group's risk becomes: 2%×0.5=1%. The absolute risk reduction is simply the difference: 2%−1%=1 percentage point.
Answer A (1 percentage point reduction) correctly captures this calculation. This means that for every 100 patients treated for 5 years, one additional heart attack would be prevented compared to standard care.
Answer B (50% risk reduction) confuses relative and absolute measures. While the relative risk reduction is indeed 50%, this isn't the absolute risk reduction the question asks for.
Answer C (1.5 percentage point reduction) appears to stem from incorrectly calculating the treatment group's risk, possibly as 0.5% instead of 1%.
Answer D (0.5 percentage point reduction) likely comes from misunderstanding how to apply the relative risk, perhaps subtracting 0.5 directly from the baseline risk.
Remember this key distinction: relative risk tells you the proportional change, while absolute risk reduction tells you the actual difference in event rates. Both are clinically important, but they answer different questions about treatment benefit. - A researcher finds that exposure to chemical Y increases cancer risk (RR = 2.5). The background cancer incidence in unexposed individuals is 8 per 1,000 over 10 years. What is the attributable risk among the exposed population?
Answer: 12 per 1,000 over 10 years
This question tests your understanding of attributable risk, a key epidemiological measure that quantifies the additional disease burden directly caused by an exposure. When you see relative risk (RR) and background incidence together, you're likely being asked to calculate either attributable risk or population attributable risk.
To find attributable risk among the exposed, you need to determine how much additional cancer risk is caused by chemical Y exposure. Start with the formula: Attributable Risk = Incidence in exposed - Incidence in unexposed.
First, calculate the incidence in exposed individuals. Since RR = 2.5, the exposed group has 2.5 times the cancer rate of the unexposed group: 2.5×8=20 per 1,000 over 10 years.
Now subtract the background rate: 20−8=12 per 1,000 over 10 years. This means chemical Y causes an additional 12 cancer cases per 1,000 exposed individuals.
Answer A (12 per 1,000) correctly represents this additional risk. Answer B (20 per 1,000) is the total incidence in exposed individuals, not the attributable risk. Answer C (8 per 1,000) is simply the background incidence rate. Answer D (2.5 per 1,000) confuses the relative risk value with an incidence rate.
Remember: attributable risk always equals (RR - 1) × background incidence. Here, that's (2.5−1)×8=12. This formula can save you time on exam questions involving attributable risk calculations. - Two studies examine the same exposure-disease relationship. Study A reports RR = 2.0 with a risk difference of 10%. Study B reports RR = 2.0 with a risk difference of 2%. What explains this apparent contradiction?
Answer: The studies examined populations with different baseline disease risks
When you encounter questions comparing relative risk (RR) and risk difference across studies, remember that these measures capture different aspects of association and are influenced differently by baseline disease risk.
The key insight is that relative risk and risk difference have a mathematical relationship that depends on baseline risk. If RR = 2.0, then the exposed group has twice the risk of the unexposed group. However, the absolute difference between these risks depends entirely on the baseline risk in the unexposed population.
For Study A: If baseline risk = 10%, then exposed risk = 20%, giving RR = 20%/10% = 2.0 and risk difference = 20% - 10% = 10%. For Study B: If baseline risk = 2%, then exposed risk = 4%, giving RR = 4%/2% = 2.0 and risk difference = 4% - 2% = 2%. Both studies show the same relative effect (doubling of risk), but different absolute effects due to different baseline risks.
Answer A correctly identifies that populations with different baseline disease risks explain this pattern. Answer B is wrong because RR and risk difference are not proportional - they have a multiplicative, not additive, relationship. Answer C is incorrect because different exposure definitions would likely change the RR estimates themselves, not just the risk differences. Answer D is wrong because sample size affects precision (confidence intervals), not the relationship between effect measures.
Remember: relative risk tells you about the strength of association, while risk difference tells you about public health impact. Both depend on baseline risk, but in different ways.
- A public health intervention reduces childhood obesity from 18% to 12% in the target population. Health officials want to expand this program to a neighboring region with a current childhood obesity rate of 24%. Assuming the same relative effectiveness, what would be the expected obesity rate after intervention in the new region?
Answer: 16% obesity rate in the new region
When evaluating intervention effectiveness across different populations, you need to distinguish between absolute and relative effectiveness. This question tests whether you understand how to apply relative effectiveness consistently.
The intervention reduced obesity from 18% to 12% in the original population. To find the relative effectiveness, calculate the relative risk reduction: the intervention retained 18%12%=32 of the original rate, meaning it reduced the rate by one-third relative to the baseline.
Applying this same relative effectiveness to the new region with 24% baseline obesity: 24%×32=16%. The answer is A) 16% obesity rate.
Now for the incorrect options: B) 12% assumes the intervention achieves the same absolute final rate regardless of starting point, which ignores biological and social factors that make interventions more challenging in higher-risk populations. C) 18% incorrectly applies the original baseline rate as the final outcome. D) 20% might result from incorrectly calculating the absolute reduction (18% - 12% = 6%) and subtracting it from 24%, giving 18%, but this isn't even listed correctly and misunderstands relative versus absolute effectiveness.
Remember this pattern: when interventions are described as having "the same relative effectiveness," multiply the new baseline by the same proportional change, not the same absolute change. Public health interventions typically show relative rather than absolute consistency across populations because underlying risk factors scale proportionally. - An epidemiologist calculates that smoking increases lung cancer risk 15-fold in a particular population. The lung cancer incidence among non-smokers is 2 per 10,000 person-years. For a city with 200,000 smokers, what is the annual excess burden of lung cancer attributable to smoking?
Answer: 560 excess lung cancer cases annually
When you encounter questions about excess disease burden, you're calculating how many additional cases occur due to a specific exposure compared to what would happen without that exposure.
To find the excess burden, you need three key pieces: the baseline risk in unexposed people, the relative risk from exposure, and the size of the exposed population. Here, non-smokers have 2 lung cancer cases per 10,000 person-years, smoking increases risk 15-fold, and there are 200,000 smokers.
First, calculate the absolute risk in smokers: 2×15=30 cases per 10,000 person-years. Next, find the excess risk (difference between smoker and non-smoker rates): 30−2=28 excess cases per 10,000 person-years. Finally, apply this to the population: 10,00028×200,000=560 excess cases annually.
Choice A (560 cases) correctly applies this three-step calculation. Choice B (600 cases) likely represents the mistake of multiplying the total smoker risk (30 per 10,000) by 200,000, forgetting to subtract the baseline risk that would occur anyway. Choice C (40 cases) appears to use the baseline non-smoker rate instead of the excess rate. Choice D (580 cases) is close to the correct answer but represents a calculation error, possibly in the arithmetic steps.
Remember: excess burden always requires subtracting the background rate from the exposed group's rate. You're calculating additional cases that wouldn't have occurred without the exposure, not the total cases in exposed individuals. - A cross-sectional study of 5,000 adults found that 200 of 1,000 smokers had chronic cough compared to 120 of 4,000 non-smokers. The researchers want to estimate what the relative risk would be if this were a prospective cohort study, but they're concerned about healthy worker bias since employed individuals were overrepresented among non-smokers. Which approach would best address this concern?
Answer: Calculate the prevalence ratio as an approximation of relative risk and apply inverse probability weighting for employment status
In cross-sectional studies, prevalence ratio (PR) approximates relative risk when disease duration is similar between groups. Here: PR = (200/1,000)/(120/4,000) = 0.20/0.03 = 6.67. Healthy worker bias occurs because employed non-smokers may be healthier, artificially lowering their cough prevalence. Inverse probability weighting can adjust for employment status distribution differences. Choice B is incorrect because OR doesn't always approximate RR, especially with common outcomes. Choice C addresses age confounding but not the specific healthy worker bias. Choice D is complex and requires assumptions about chronic cough natural history that aren't provided.
- A cohort study of 5,000 healthcare workers finds that proper hand hygiene reduces healthcare-associated infection transmission to patients (RR = 0.4). The baseline transmission rate is 6 infections per 100 healthcare workers per month. What is the prevented fraction among the exposed (vaccine efficacy equivalent)?
Answer: 60% of infections are prevented by proper hand hygiene
When you encounter questions about prevented fraction (also called vaccine efficacy or attributable efficacy), you're measuring how much disease exposure to a protective factor prevents. This is a fundamental concept in evaluating preventive interventions.
The prevented fraction among the exposed uses the formula: Prevented Fraction=RRunexposedRRunexposed−RRexposed=11−RR=1−RR
Since the relative risk (RR) compares exposed to unexposed groups, and proper hand hygiene has RR = 0.4, this means healthcare workers with proper hygiene have 40% the infection risk of those without proper hygiene.
The prevented fraction = 1 - 0.4 = 0.6 = 60%. This tells us that proper hand hygiene prevents 60% of the infections that would otherwise occur.
Answer A correctly states that 60% of infections are prevented by proper hand hygiene. Answer B confuses the RR value (0.4 or 40%) with the prevented fraction—this is the residual risk, not what's prevented. Answers C and D incorrectly try to incorporate the baseline rate (6 infections per 100 workers), but prevented fraction is a relative measure independent of baseline incidence rates.
Remember: prevented fraction always equals (1 - RR) when you have the relative risk. Don't let baseline rates distract you—this measure focuses purely on the proportional reduction in risk among those with the protective exposure. - A case-control study cannot directly calculate relative risk, but can estimate it using the odds ratio when the disease is rare. If a cohort study of the same population found a relative risk of 3.2 for the association between exposure X and disease Y, and disease Y has a prevalence of 15% in the population, how well would the odds ratio approximate the relative risk?
Answer: The odds ratio would substantially overestimate the relative risk due to high disease prevalence
When you encounter questions about odds ratios approximating relative risk, you need to understand that this approximation only works well when disease prevalence is genuinely rare (typically <10%).
The relationship between odds ratio (OR) and relative risk (RR) depends critically on disease frequency. When disease prevalence is low, the odds ratio closely approximates the relative risk. However, as disease prevalence increases, the odds ratio increasingly overestimates the relative risk. At 15% prevalence, this overestimation becomes substantial.
Here's why: the odds ratio compares odds rather than risks. With higher disease prevalence, the denominator in odds calculations (1 - probability) becomes notably smaller, inflating the odds ratio compared to the relative risk, which directly compares probabilities.
Answer A correctly identifies that the odds ratio would substantially overestimate the relative risk due to the high (15%) disease prevalence. Answer B is incorrect because 15% prevalence is not considered rare in epidemiological terms - the "rare disease assumption" typically requires prevalence below 10%. Answer C incorrectly suggests underestimation when the odds ratio actually overestimates relative risk at higher prevalences. Answer D is completely wrong since the relationship between odds ratio and relative risk is never exact except in theoretical scenarios with zero disease prevalence.
Study tip: Remember the "10% rule" - odds ratios only approximate relative risk well when disease prevalence is below 10%. Above this threshold, odds ratios progressively overestimate relative risk, making case-control studies less reliable for estimating true relative risk.
- In a case-control study, researchers found that 60% of cases with myocardial infarction had a history of smoking compared to 30% of controls. If the study had been designed as a cohort study with the same underlying population and smoking prevalence of 35%, what additional information would be needed to calculate the relative risk?
Answer: The incidence rate of myocardial infarction in the source population and follow-up time
To calculate relative risk, we need the incidence rates in exposed (smokers) and unexposed (non-smokers) groups. While the case-control study tells us about exposure distribution among cases and controls, it doesn't directly provide incidence rates. We need the disease incidence rate in the population and follow-up time to convert the exposure odds to actual risks in a hypothetical cohort. Choice A is incorrect because OR alone cannot be converted to RR without knowing disease incidence. Choice C is incorrect as it addresses measurement error, not the fundamental design difference. Choice D is incorrect because prevalence doesn't provide the temporal risk information needed for RR calculation.
- Based on the table above, what is the number needed to treat (NNT) for the diabetes prevention program?
Answer: 20 individuals need treatment to prevent one case of diabetes
From the table: Intervention group risk = 80/1000 = 0.08 (8%). Control group risk = 130/1000 = 0.13 (13%). Risk difference = 0.13 - 0.08 = 0.05 (5%). NNT = 1/risk difference = 1/0.05 = 20. Choice B uses incorrect risk calculation. Choice C represents 1/control group risk. Choice D represents 1/intervention group risk. Choice E represents half the correct calculation.
- A randomized trial comparing two surgical procedures found that Procedure A had a 5% complication rate while Procedure B had a 15% complication rate. The study included 300 patients per group. Six months later, a larger trial with 1,000 patients per group found complication rates of 8% for Procedure A and 12% for Procedure B. What can be concluded about the relative risk estimates?
Answer: The relative risks differ substantially (0.33 vs 0.67) suggesting different patient populations or procedural modifications
Study 1: RR = 5%/15% = 0.33. Study 2: RR = 8%/12% = 0.67. These represent a substantial difference, with the relative risk doubling from the first to second study. This magnitude of difference (0.33 vs 0.67) suggests meaningful heterogeneity that could be due to different patient populations, learning curves, surgeon experience, or procedural modifications. The larger sample size in Study 2 makes its estimate more precise, but the substantial difference in effect size indicates true heterogeneity rather than random variation.
- Two studies examined the association between vitamin D supplementation and respiratory infections. Study 1 (northern climate, winter months): 15% infection rate in supplemented group vs 25% in placebo group. Study 2 (tropical climate, year-round): 8% infection rate in supplemented group vs 10% in placebo group. A meta-analyst wants to determine whether the absolute or relative effect is more consistent across studies. What should they conclude?
Answer: Both measures show substantial heterogeneity (RR: 0.60 vs 0.80; RD: 10% vs 2%) suggesting different mechanisms or populations
Study 1: RR = 15%/25% = 0.60, RD = 25% - 15% = 10%. Study 2: RR = 8%/10% = 0.80, RD = 10% - 8% = 2%. Both relative risk (0.60 vs 0.80, a 33% difference) and risk difference (10% vs 2%, a 5-fold difference) show substantial heterogeneity. This suggests genuine differences between studies - possibly due to baseline vitamin D status, seasonal variation, or population characteristics. Choice A incorrectly suggests RR is more consistent. Choice B incorrectly suggests RD is consistent. Choice D incorrectly attributes RD differences only to baseline rates rather than recognizing true effect heterogeneity.
- In a study comparing two surgical procedures, Procedure A has a 3% complication rate and Procedure B has a 9% complication rate. If a hospital switches from Procedure B to Procedure A for all cases, what proportion of complications would be prevented?
Answer: 67% of complications would be prevented
When you encounter questions about reducing disease rates or complications, you're dealing with relative risk reduction - a key concept in epidemiology that measures the proportional decrease in adverse outcomes.
To find what proportion of complications would be prevented, you need to calculate how much the complication rate decreases relative to the original rate. Start with Procedure B's 9% rate and Procedure A's 3% rate. The absolute reduction is 9%−3%=6%. However, the question asks for the proportion of complications prevented, which requires dividing this reduction by the original rate: 9%6%=0.67=67%. This means that switching procedures would prevent 67% of the complications that would have occurred under Procedure B.
Option A correctly identifies this 67% relative risk reduction. Option B states "6 percentage points" - this describes the absolute risk reduction, not the proportion prevented. While mathematically correct, it doesn't answer what was asked. Option C gives 33%, which would be the remaining complication rate expressed as a fraction of the original rate (9%3%=33%), but this represents complications that still occur, not those prevented. Option D mentions "3 times fewer," which is imprecise language and doesn't properly quantify the proportional reduction.
Remember: absolute risk reduction tells you the raw difference in rates, while relative risk reduction tells you the proportional improvement - always read carefully to determine which the question is asking for. - A study reports that the relative risk of developing diabetes for individuals with BMI ≥30 compared to BMI <25 is 3.2 (95% CI: 2.1-4.9). If the baseline incidence of diabetes in the BMI <25 group is 2 per 1,000 person-years, and you want to calculate the number needed to treat (NNT) for a weight loss intervention that reduces BMI from ≥30 to <25, what intermediate calculation is required?
Answer: Calculate the absolute risk difference between BMI ≥30 and BMI <25 groups using baseline incidence
To calculate NNT, we need the absolute risk difference (ARD). Given: RR = 3.2, baseline risk (BMI <25) = 2/1,000 person-years. Risk in BMI ≥30 group = 3.2 × 2 = 6.4/1,000 person-years. ARD = 6.4 - 2 = 4.4/1,000 person-years. NNT = 1/ARD = 1,000/4.4 ≈ 227. Choice A is incorrect because we don't need OR for NNT calculation. Choice C is incorrect because population attributable risk fraction is not needed for individual NNT. Choice D is incorrect because hazard ratio conversion is unnecessary and doesn't help with NNT calculation.
- A safety analysis reveals that a medical device has a 0.3% complication rate compared to 0.1% for the standard procedure. The relative risk is 3.0. A hospital performs 10,000 procedures annually. If they switch to the standard procedure, how many complications would be prevented per year?
Answer: 20 complications prevented annually
When you encounter questions about risk differences and prevention calculations, you're working with absolute risk reduction - the actual difference in event rates between two interventions.
To find how many complications would be prevented, you need to calculate the absolute difference in complication rates and apply it to the hospital's volume. The medical device has a 0.3% complication rate while the standard procedure has a 0.1% rate. The absolute risk reduction is 0.3% - 0.1% = 0.2%.
With 10,000 procedures annually, switching to the standard procedure would prevent: 0.002×10,000=20 complications per year.
Let's examine why the other answers are incorrect. Answer B (30 complications) might result from mistakenly using the medical device's complication rate (0.3% × 10,000 = 30) rather than the risk difference. Answer C (10 complications) could come from using only the standard procedure's rate (0.1% × 10,000 = 10) instead of the difference between rates. Answer D (300 complications) represents a decimal error - perhaps calculating 0.3% as 0.03 instead of 0.003, leading to 0.03 × 10,000 = 300.
The relative risk of 3.0 confirms our calculation is reasonable (0.3% ÷ 0.1% = 3.0) but isn't directly needed for this prevention calculation.
Study tip: Always distinguish between relative measures (like relative risk) and absolute measures (like risk differences). For prevention calculations, focus on the absolute difference in rates, not the relative comparison between groups. - A retrospective cohort study examines occupational exposure to asbestos and lung cancer risk over 30 years. Workers with high exposure have a lung cancer incidence of 45 per 1,000 workers, while those with low exposure have an incidence of 15 per 1,000 workers. What percentage of lung cancer cases among highly exposed workers is attributable to the excess exposure?
Answer: 67% of cases are attributable to excess exposure
When you encounter questions about exposure and disease risk, you're dealing with attributable risk concepts that help quantify how much disease can be blamed on a specific exposure.
The key here is calculating the attributable risk percent (AR%) among the exposed group. This tells you what percentage of disease cases in the exposed population is due to the excess exposure beyond baseline risk. The formula is: AR%=RiskexposedRiskexposed−Riskunexposed×100
Let's calculate: The highly exposed workers have 45 cases per 1,000, while low-exposure workers have 15 per 1,000. So: AR%=4545−15×100=4530×100=67%
This means 67% of lung cancer cases among highly exposed workers can be attributed to their excess asbestos exposure.
Looking at the wrong answers: B (33%) represents a common error where students calculate the unexposed risk as a percentage of exposed risk (15/45 = 33%). C (30%) is the absolute risk difference (45-15 = 30) but not converted to a percentage of the exposed group. D (75%) might result from incorrectly using 40 instead of 45 in the denominator.
Study tip: Always remember that attributable risk percent asks "of all cases in the exposed group, what fraction is due to the exposure?" The denominator is always the risk in the exposed group, and you're finding what proportion of that risk is excess beyond baseline. - A cohort study follows 1,200 smokers and 800 non-smokers for 10 years to assess lung cancer development. Among smokers, 48 develop lung cancer. Among non-smokers, 8 develop lung cancer. What is the relative risk of lung cancer for smokers compared to non-smokers?
Answer: 4.0
When you encounter a cohort study asking about relative risk, you're calculating how many times more likely an exposed group is to develop an outcome compared to an unexposed group. The formula is: Relative Risk=Risk in unexposed groupRisk in exposed group
First, calculate the risk (incidence rate) in each group. For smokers: 48 lung cancers ÷ 1,200 smokers = 0.04 or 4%. For non-smokers: 8 lung cancers ÷ 800 non-smokers = 0.01 or 1%.
Now apply the relative risk formula: RR=0.010.04=4.0
This means smokers are 4 times more likely to develop lung cancer than non-smokers, making answer A correct.
Let's examine why the other options are wrong. Answer B (6.0) might result from incorrectly dividing 48 by 8 (the raw number of cases), but this ignores the different population sizes. Answer C (2.4) could come from miscalculating the denominator, perhaps using the wrong population figures. Answer D (5.0) doesn't correspond to any logical calculation error but serves as a close distractor.
Remember this pattern: relative risk questions always require you to first calculate the risk (cases ÷ population) in each group, then divide the exposed group's risk by the unexposed group's risk. Never work directly with raw case numbers—always convert to rates first, accounting for different group sizes. - A clinical trial tests a new antiviral drug for treating influenza. The drug reduces symptom duration from an average of 7 days to 4.2 days, representing a 40% reduction. However, when examining the proportion of patients who recover within 5 days, the drug increases this from 30% to 65%. What is the relative risk of 5-day recovery with the new drug?
Answer: 2.17
When you encounter questions about treatment effects and recovery rates, you're dealing with relative risk calculations. Relative risk compares the probability of an outcome occurring in the treatment group versus the control group.
To calculate relative risk, you divide the risk (probability) in the treatment group by the risk in the control group. Here, the "risk" we're measuring is actually the positive outcome of recovering within 5 days. With the new drug, 65% of patients recover within 5 days, compared to 30% with standard treatment.
Relative risk = Risk in control groupRisk in treatment group=0.300.65=2.17
This means patients are 2.17 times more likely to recover within 5 days when taking the new drug.
Looking at the wrong answers: B) 1.40 represents the 40% reduction in symptom duration mentioned in the question stem, but that's not what we're calculating here. C) 0.60 would be the relative risk if you incorrectly flipped the fraction (30%/50%), suggesting the drug was harmful rather than beneficial. D) 35 percentage points improvement is the absolute risk difference (65% - 30%), not the relative risk.
Remember that relative risk greater than 1 indicates increased likelihood of the outcome with treatment, while values less than 1 suggest decreased likelihood. Always pay attention to which specific outcome the question is asking about - here it's 5-day recovery, not overall symptom duration. - In a randomized controlled trial, 150 patients receive a new drug and 150 receive placebo. The adverse event rate is 20% in the treatment group and 8% in the placebo group. What is the risk difference (attributable risk)?
Answer: 0.12 or 12 percentage points
When you encounter questions about adverse events in clinical trials, you're typically being asked to calculate measures of risk association. Risk difference (also called attributable risk) is one of the most fundamental measures - it simply tells you how much additional risk is attributable to the treatment itself.
Risk difference is calculated as the absolute difference between the risk in the exposed group and the risk in the unexposed group. Here, the treatment group has a 20% adverse event rate and the placebo group has an 8% rate. So: Risk difference = 20% - 8% = 12% (or 0.12). This means the new drug is associated with 12 additional adverse events per 100 patients treated.
Let's examine why the other options are incorrect. Option B (2.5 times higher risk) represents the relative risk ratio (20%/8% = 2.5), which tells you the treatment group has 2.5 times the risk of the placebo group - this is a different measure entirely. Option C (8 percentage points) gives you only the placebo group's baseline risk, ignoring the treatment effect. Option D (20 percentage points) is just the treatment group's risk without accounting for what would have happened anyway (the baseline risk).
The correct answer is A: 0.12 or 12 percentage points.
Study tip: Remember that risk difference answers "How many extra cases per 100 people?" while relative risk answers "How many times higher is the risk?" Always subtract the unexposed group rate from the exposed group rate for risk difference calculations.
