The Relative Risk Calculator computes relative risk (RR) by comparing the probability of an outcome in an exposed group versus an unexposed group. It uses your event counts to produce RR and a confidence interval so you can judge how strong the difference is.
This article shows the exact inputs the calculator needs, how RR is interpreted, and what to watch for when data are small or when the outcome is rare.
What Is Relative Risk (RR)?
Relative risk compares the chance of an event happening in two groups. One group is exposed (or receives a factor), and the other is unexposed. RR tells you how many times more (or less) likely the outcome is in the exposed group compared with the unexposed group.
- RR = 1: no difference in risk.
- RR > 1: higher risk in the exposed group.
- RR < 1: lower risk in the exposed group.
Relative Risk Calculator Inputs (What You Need)
To compute RR, you must know how many people experienced the outcome and how many were in each group. The calculator uses four numbers:
- Exposed events: number of participants in the exposed group who had the outcome.
- Exposed total: total number of participants in the exposed group.
- Unexposed events: number of participants in the unexposed group who had the outcome.
- Unexposed total: total number of participants in the unexposed group.
All totals must be positive, and each event count must be between 0 and its corresponding total.
The Core Formula for Relative Risk
Relative risk is based on two risks (probabilities):
- Risk in exposed = a / n1
- Risk in unexposed = c / n0
Where:
- a = exposed events
- n1 = exposed total
- c = unexposed events
- n0 = unexposed total
The RR formula is:
RR = (a / n1) / (c / n0)
Confidence Interval (How Precise Is RR?)
RR alone does not tell you whether the difference is likely due to chance. A 95% confidence interval (CI) gives a range of plausible RR values. If the CI includes 1.0, the observed difference may not be statistically distinguishable from no difference.
The calculator computes a CI using the log method:
- Compute ln(RR).
- Compute the standard error (SE) for ln(RR).
- Build the CI on the log scale, then exponentiate back.
Continuity correction: If one group has zero events, the standard log-based CI can fail because it requires division by event counts. The calculator applies a small continuity correction to handle zeros safely.
How to Interpret Results
When you read the output, use both the RR value and its confidence interval.
| Result pattern | Meaning |
|---|---|
| RR = 1.0 | The outcome risk is the same in both groups. |
| RR > 1.0 (CI entirely above 1.0) | The exposed group has a higher risk, and the increase is consistent with the data. |
| RR < 1.0 (CI entirely below 1.0) | The exposed group has a lower risk, and the decrease is consistent with the data. |
| CI includes 1.0 | There may be no clear evidence of a difference, even if RR is above or below 1. |
Common Pitfalls (Avoid These Mistakes)
- Mixing denominators: totals must match the event counts for each group.
- Using RR for case-control studies: RR is most appropriate for cohort-style data where you can compute risks.
- Ignoring small sample sizes: with few events, CIs widen and estimates become unstable.
- Confusing RR with odds ratio: odds ratio (OR) is different, especially when outcomes are common.
Practical Example 1: Medication Effect on a Side Effect
Suppose 200 people take a medication (exposed) and 180 people do not (unexposed). In the exposed group, 24 people report a side effect. In the unexposed group, 10 report it.
- Exposed: a = 24, n1 = 200 → risk = 0.12
- Unexposed: c = 10, n0 = 180 → risk ≈ 0.0556
RR ≈ 0.12 / 0.0556 ≈ 2.16. This means the side effect risk is about 2.16× higher with the medication. The confidence interval will tell you whether this increase is likely to be real or could be due to chance.
Practical Example 2: Workplace Exposure and Injury Rates
Imagine a workplace safety change is introduced. Before the change, 12 injuries occur out of 500 workers (unexposed). After the change, 20 injuries occur out of 520 workers (exposed).
- Exposed: a = 20, n1 = 520
- Unexposed: c = 12, n0 = 500
The RR will indicate whether injuries increased or decreased relative to the earlier period. If RR is below 1 and the CI stays below 1, the data support a reduction in risk after the change.
Frequently Asked Questions
What does a Relative Risk Calculator measure?
A Relative Risk Calculator measures relative risk (RR), the ratio of an outcome probability in an exposed group to the probability in an unexposed group. It uses event counts and group totals to compute RR and typically reports a confidence interval to show precision.
How do I interpret an RR of 0.75?
An RR of 0.75 means the exposed group’s outcome risk is 25% lower than the unexposed group’s risk. Interpretation depends on the confidence interval: if it stays below 1, the reduction is consistent with the data; if it includes 1, evidence is weaker.
What if one group has zero events?
When a group has zero events, RR can be undefined because the risk in that group is zero. The calculator applies a small continuity correction to produce a stable RR and confidence interval. Results still require careful interpretation, especially with small totals.
Is relative risk the same as odds ratio (OR)?
No. Relative risk compares probabilities, while odds ratio compares odds. When outcomes are common, OR can differ substantially from RR. For rare outcomes, OR and RR often get closer, but they are not interchangeable.
Which study types are best for relative risk?
Relative risk is best for studies where you can compute risks in both groups, such as cohort studies or randomized trials. In case-control designs, you typically do not measure risk directly, so RR is not usually the primary measure.