Myles & Cui. Use of the Bland Altman method to measure compliance with repeated measurements. BJA: British Journal of Anaesthesia, Volume 99, Issue 3, 1 September 2007, pages 309-311, doi.org/10.1093/bja/aem214. Retrieved from academic.oup.com/bja/article/99/3/309/355972 april 23, 2018 Errors resulting from the three types of confidence intervals show that the exact approach of the 96 cases presented in Tables 1, 2, 3 and 4 works very well. For the two approximate methods Chakraborti and Li  and Bland and Altman , the probabilities of coverage of their bilateral interval remain quite close to the nominal confidence levels. However, the corresponding unilateral interval procedures do not retain the same desired accuracy, unless the sample size is large. It is identical to a mean tukey difference graph, the name by which it is known in other fields, but was popularized in the medical statistics of J. Martin Bland and Douglas G. Altman.   The limitations of concordance include both bias and accidental errors and provide a useful measure for comparing likely differences between results, measured against two methods. If one of the methods is a reference method, compliance limits can be used as a measure of the total error of a measurement method (Krouwer, 2002). In particular, Bland and Altman [1, 2] proposed compliance limits of 95% for the assessment of differences between measurements by two methods. The breakpoints of the Bland Altman 95% compliance limits are the 2.5th percentile and the 97.5th percentile for the distribution of the difference between failed measurements.
In order to reflect uncertainty due to sampling errors, approximate interval formulas were presented to estimate the two individual percentiles. . . .