Confidence Intervals

When a study reports that a drug lowers blood pressure by 8 mmHg, that single number, the point estimate, is only the best guess from one sample. Study a different sample of patients and you would get a slightly different number. The confidence interval (CI) captures this uncertainty by giving a range of plausible values for the true effect, not just one point. You will see it written as '8 mmHg (95% CI 5 to 11)'. What does '95%' actually mean? The technically correct interpretation is about the method: if you repeated the same study many times and computed a 95% CI each time, about 95% of those intervals would contain the true value. In everyday practice clinicians read it more loosely as 'we can be reasonably confident the true effect lies between 5 and 11 mmHg'. The precise definition matters less than the habit of always looking at the whole interval, not just the point estimate. Two features of a CI tell you almost everything you need. First, its width reflects precision. A narrow interval (8 mmHg, 95% CI 7 to 9) means a precise, trustworthy estimate; a wide one (8 mmHg, 95% CI minus 2 to 18) means the study could not pin the effect down, it might even be zero or harmful. Width depends mostly on sample size: bigger studies give narrower intervals. Second, whether the interval crosses a value of 'no effect' tells you about significance. For a difference, the no-effect value is 0; for a ratio such as an odds ratio or relative risk, it is 1. If the CI for a difference includes 0 (or a ratio's CI includes 1), the result is not statistically significant at that level. This is why confidence intervals are more informative than p-values alone. A p-value gives a yes/no verdict on significance; a CI shows the size of the effect AND its precision AND its significance in one expression. Two studies can both be 'significant', yet one with a CI of 2 to 4 and another with 0.1 to 12 tell very different stories about how much you actually know. For a clinician the practical move is simple: read the point estimate to see the likely effect, then read the interval to ask two questions, is the whole range clinically meaningful, and is it narrow enough to act on? An effect that is significant but whose interval stretches from trivial to huge is rarely enough to change practice on its own.