Confidence intervals are one of the most useful tools in medical research, yet many students and clinicians find them confusing. This article explains what confidence intervals are, why they matter, how to calculate them for means and proportions, and how to interpret them correctly. By the end, you will have a practical framework for reading and reporting confidence intervals in real medical studies.
A confidence interval is a range of values that likely contains the true population parameter, such as a mean or proportion, based on sample data. Instead of reporting only a single point estimate, a confidence interval provides a range that reflects the uncertainty caused by sampling variability.
The most common confidence level is 95%, but you will also encounter 90% and 99% intervals. The confidence level describes how often the method works in repeated sampling, not how likely any single interval is to contain the true value.
Medical decisions are rarely based on certainty. Researchers study samples and must use statistics to make inferences about entire patient populations. Confidence intervals help judge whether a result is precise enough to guide diagnosis, treatment, or public health policy.
For example, a study may report that a new drug lowers blood pressure by 5 mmHg on average. Is this a meaningful improvement? If the 95% confidence interval is 2 to 8 mmHg, the range suggests a consistent, clinically important benefit. If the interval is -2 to 12 mmHg, the result is too uncertain to support a strong recommendation.
When your outcome is a continuous variable like blood pressure, weight, or cholesterol, you can construct a confidence interval around the mean. The general formula is:
Confidence interval = mean ± (critical value) × (standard deviation / square root of sample size)
The critical value depends on your confidence level and whether you use a z-score or a t-score. Use z-scores when the sample size is large (usually above 30) and the population standard deviation is known. Use t-scores for smaller samples, with degrees of freedom equal to n minus 1.
Suppose you have a sample of 25 patients with an average systolic blood pressure of 140 mmHg and a standard deviation of 15 mmHg. Because the sample size is small, use a t-score with 24 degrees of freedom. For a 95% confidence interval, the t-score is approximately 2.064.
First, calculate the standard error: 15 / √25 = 15 / 5 = 3. Then multiply by the t-score: 2.064 × 3 = 6.192. The margin of error is about 6.19 mmHg. The confidence interval is 140 ± 6.19, which gives a range from 133.81 to 146.19 mmHg.
| Step | Value | Calculation |
|---|---|---|
| Sample size | n = 25 | — |
| Mean | 140 mmHg | — |
| Standard deviation | 15 mmHg | — |
| Standard error | 3 | 15 / √25 |
| t-score for 95% CI, df = 24 | 2.064 | From t-table |
| Margin of error | 6.19 | 2.064 × 3 |
| Lower limit | 133.81 | 140 − 6.19 |
| Upper limit | 146.19 | 140 + 6.19 |
For categorical outcomes, such as the percentage of patients who experience a side effect, you calculate a confidence interval for a proportion. The formula is:
Confidence interval = p ± z × √(p(1 − p) / n)
Here, p is the sample proportion, z is the critical value from the standard normal distribution, and n is the sample size. This method works well when both n × p and n × (1 − p) are at least 10. For rare events, other methods like the Wilson score interval may be more accurate.
Example: A study follows 200 patients who receive a new medication, and 30 develop mild side effects. The sample proportion is 30/200 = 0.15. The standard error is √(0.15 × 0.85 / 200) = √(0.1275 / 200) = √0.0006375 = 0.02525. The margin of error is 1.96 × 0.02525 = 0.0495. The 95% confidence interval is 0.15 ± 0.0495, or 10.05% to 19.95%.
Many students misinterpret confidence intervals. A 95% confidence interval does not mean there is a 95% chance that the true value lies within the particular interval you calculated. Instead, it means that if you repeated the study many times and calculated an interval each time, about 95% of those intervals would contain the true value.
A confidence interval is a range of plausible values for the true effect based on the data you observed. It is not a probability statement about the parameter after the interval has been calculated.
The correct wording is to say you are 95% confident that the true value lies between the lower and upper limits. This subtle distinction matters for honest reporting and interpretation.
Even experienced researchers can misuse confidence intervals. Knowing the most common mistakes helps you avoid them in your own work and when you read the literature.
P-values have been a staple of medical statistics for decades, but they only tell you whether a result is statistically unlikely under the null hypothesis. Confidence intervals offer far more information because they estimate the magnitude of an effect and show the range of values consistent with the data.
A p-value can tell you whether a result is statistically significant, but a confidence interval tells you whether that result matters and how certain you are about it.
For example, a large study may produce a p-value of 0.04 and a small effect, while another study with the same p-value may show a much larger effect. The confidence interval reveals this difference. Also, if the null value (such as zero for a difference or one for a ratio) lies outside the 95% confidence interval, you can infer that the result is statistically significant at the 5% level.
Whether you are writing a lab report, a dissertation, or a journal manuscript, reporting confidence intervals correctly is essential. A well-reported interval allows readers to understand the precision and practical relevance of your findings.
Understanding confidence intervals is not just a statistical exercise; it is a core skill for evaluating medical evidence. Once you learn how to calculate and interpret them properly, you will be able to make more nuanced judgments about treatment effects and patient outcomes. Use the practice examples above, and always ask: What does this interval really tell me about the uncertainty in this result?
A 95% confidence interval means that if you repeated the same study many times under identical conditions and calculated an interval each time, approximately 95% of those intervals would contain the true population value. It does not mean there is a 95% probability that any single calculated interval contains the true value.
