How do you interpret a 95% confidence interval for a risk ratio?

• A. It contains the true RR in 95% of similar studies; if it includes 1, not significant at 0.05.
• B. It is the probability that the true RR lies within the interval for this study only.
• C. It indicates the range in which the true RR lies with 95% confidence, assuming no bias; if it includes 1, the observed association could be due to chance.
• D. It provides the p-value directly.

Answer: (C)

The main idea here is that a 95% confidence interval for a risk ratio shows the range of true values that are plausible given the data, and it reflects the precision of the estimate. If we could repeat the study many times and compute a confidence interval each time using the same design and analysis, about 95% of those intervals would contain the true risk ratio. This interpretation relies on the study being properly conducted and free of major bias.

Practically, the interval tells us where the true RR lies with 95% confidence for this study. If the interval does not include 1, the association is statistically significant at the 0.05 level; if it does include 1, the observed association could be due to chance. The width of the interval shows precision: a narrow interval means more precision, a wide one means more uncertainty. The interval is related to the p-value for testing RR = 1, with a two-sided test at alpha = 0.05.