Interpretation of Relative Risk 0.70 with 95% CI of 0.50 to 0.98
A relative risk (RR) of 0.70 with a 95% confidence interval of 0.50 to 0.98 indicates a statistically significant 30% reduction in risk with the intervention compared to the control group.
Understanding the Key Components
Relative Risk (RR) of 0.70
- Indicates that the risk of the outcome in the intervention group is 30% lower than in the control group
- Formula: RR = (Risk in intervention group) ÷ (Risk in control group)
- RR < 1.0 suggests a protective effect (reduced risk)
- RR = 0.70 means the intervention group has 70% of the risk compared to the control group
95% Confidence Interval (CI) of 0.50 to 0.98
- Provides the range within which the true RR is likely to fall with 95% confidence
- The CI does not cross 1.0, making this result statistically significant
- Lower bound (0.50): Best-case scenario - up to 50% risk reduction
- Upper bound (0.98): Worst-case scenario - minimal (2%) risk reduction
Clinical Significance Assessment
Statistical vs. Clinical Significance
- This result is statistically significant because the CI doesn't include 1.0 1
- Clinical significance depends on:
- The baseline risk of the outcome
- The severity of the outcome being prevented
- The costs, risks, and burdens of the intervention
Absolute Risk Reduction (ARR)
- Cannot be calculated directly from the RR without knowing baseline risk
- For example:
- With a 10% baseline risk: ARR = 10% × (1 - 0.70) = 3% (NNT = 33)
- With a 20% baseline risk: ARR = 20% × (1 - 0.70) = 6% (NNT = 17)
Strength of Evidence Assessment
Quality Considerations
- The American Gastroenterological Association guidelines use the GRADE approach to evaluate evidence quality 1:
- High: Very confident the true effect is close to the estimate
- Moderate: Moderately confident in the effect estimate
- Low: Limited confidence; true effect may differ substantially
- Very low: Very little confidence; true effect likely substantially different
Precision of the Estimate
- The width of the CI (0.50-0.98) suggests moderate precision
- The upper bound is close to 1.0, indicating the effect could be minimal
- The lower bound suggests the potential for a substantial 50% risk reduction
Common Pitfalls in Interpretation
Avoid These Mistakes
- Confusing RR with absolute risk reduction: RR is a relative measure and doesn't tell you how many people benefit
- Ignoring the confidence interval: The CI provides crucial information about precision and significance
- Assuming causality: Association doesn't prove causation; other factors may explain the relationship
- Generalizing beyond the study population: Results may not apply to different populations
Proper Context is Essential
- Consider the study design, population characteristics, and intervention details
- Evaluate for potential biases and confounding factors
- Compare with other studies examining similar interventions
Application to Clinical Decision-Making
When applying this evidence to clinical practice:
- Consider both statistical significance and clinical importance
- Weigh potential benefits against harms, costs, and patient preferences
- Recognize that even a modest 30% risk reduction may be valuable for serious outcomes
- Use the confidence interval to communicate the range of plausible effect sizes to patients
In summary, this RR of 0.70 (95% CI 0.50-0.98) represents a statistically significant protective effect of moderate magnitude, with reasonable confidence that the intervention reduces risk by between 2% and 50%.