Glossary
Confidence interval.
Direct answer
A confidence interval gives the range within which the true value plausibly lies according to the data collected. It replaces a single figure with a span and thereby shows how precisely a study could measure at all. A wide span means little information, a narrow span good precision.
In practice an interval is read from its edges: if the least favourable plausible value is still worthwhile, the conclusion holds. If the span runs from clear improvement to clear worsening, the study is simply too small. The interval says nothing, by contrast, about whether the study was fairly designed.
In small equine studies intervals are routinely wide, which makes an isolated average especially misleading. A result published without an interval is hard to interpret, whatever species was studied. Worth remembering: a wide span does not prove the absence of an effect, it signals a shortage of precision.
The evidence letter
Once a month: new syntheses, checked studies, and what has changed in the state of the evidence.