Patient reported outcome measures: from the classics to AI.

Abstract

With its basis in the development of intelligence testing, classical test theory paved the way to develop patient-reported outcome measures - tools capable of quantifying otherwise immeasurable traits. In hand surgery, many of the popular outcome measures were developed using classical test theory. However, these techniques are increasingly being superseded by novel mathematical approaches. This review article will chart the progress from established scientific methods through to modern day item response theory and computerized adaptive testing, and explore the opportunities that artificial intelligence may afford.Level of evidence: V.