Using linear regression to assess dose-dependent bias on a Bland-Altman plot

Kwok M. Ho

Abstract

Assessing agreement and possible interchangeability between two methods of clinical measurement, when there is no gold standard to compare with, is common in medical research (1,2).