Prediction interval analysis is underutilized and can be more helpful than just confidence interval analysis.

摘要

Prediction interval analysis (PIA) is a statistical technique that seems to be little known in the anesthesia community and appears infrequently in our literature. I performed a PubMed search and found only three articles in Anesthe-siology since 2005, none in 2007 nor 2008 to date, that used prediction intervals. Anecdotally, when queried, I find that it is very rare that a clinical anesthesiologist or researcher is aware of PIA. This letter is intended to bring PIA to wider attention, introduce concepts with a hypothetical example, demonstrate with a published example (in an agreement study) its potential benefits, and to encourage its wider use. Confidence intervals (CI) are well known and commonly used in our literature. Some of the differences between CI and prediction intervals (PI) will be contrasted.