Enter Study Inputs
Example Data Table
| Group | Events | Total | Absolute Risk | Absolute Risk Difference | Relative Risk | NNT |
|---|---|---|---|---|---|---|
| Treatment Group | 18 | 240 | 7.50% | -5.00% | 0.60 | 20 |
| Control Group | 30 | 240 | 12.50% |
This example shows a lower observed event probability in the treatment group during the same follow-up period.
Formula Used
Absolute Risk (Exposed) = a / n1
Absolute Risk (Reference) = c / n0
Absolute Risk Difference = (a / n1) - (c / n0)
Relative Risk = (a / n1) / (c / n0)
Odds Ratio = (a / b) / (c / d), where b = n1 - a and d = n0 - c
NNT/NNH = 1 / |Absolute Risk Difference|
Confidence Intervals use normal approximation for risks and risk differences, plus log-scale intervals for relative risk and odds ratio.
How to Use This Calculator
- Enter labels for the compared cohorts.
- Provide event counts and total sample sizes for both groups.
- Choose a confidence level and preferred decimal precision.
- Click Calculate Absolute Risk to generate the results.
- Review absolute risks, confidence intervals, NNT or NNH, and cohort comparison metrics.
- Use the CSV button for spreadsheet work and the PDF button for reports.
Frequently Asked Questions
1. What does absolute risk mean?
Absolute risk is the observed probability of an event in one group. It equals events divided by total participants during the stated follow-up period.
2. What is the difference between absolute risk and relative risk?
Absolute risk reports event probability directly. Relative risk compares one group’s probability with another group’s probability by division.
3. When should I use NNT instead of risk difference?
Use NNT or NNH when you want an intuitive treatment impact measure. Risk difference remains the core absolute effect estimate behind that number.
4. Why can confidence intervals become wide?
Intervals widen when sample sizes are small, event counts are rare, or group risks vary heavily. Wider intervals mean greater statistical uncertainty.
5. Can I use this for case-control data?
Case-control studies usually do not estimate absolute risk directly because sampling begins with outcome status. They often emphasize odds ratios instead.
6. What if one group has zero events?
The calculator still reports absolute risks and risk difference. Relative risk and odds ratio intervals use a small continuity correction for stability.
7. Does this calculator replace a full regression model?
No. It summarizes grouped binary outcomes only. Adjusted analyses need regression methods when confounding, matching, or covariate control matters.
8. What units should the event counts use?
Use simple counts of participants, items, or observations with the event. Totals must cover the same population and same time horizon.