Carkeet A. Exact Parametric Confidence Intervals for Bland-Altman Compliance Limits. Optom Vis Sci. 2015;92:e71-80. The boundaries of agreement estimate the interval between some of the differences between the measures. Unfortunately, acceptable loas have not been fixed in advance. Clinically, after an average surface area of 24.1 cm2, it was necessary to assess whether the BA-LoA of USD 2.39 to $2.36 was close enough, with external confidence limits of 2.71 and 2.69 USD. Compliance limitations include both systematic errors (bias) and random errors (precision) and provide a useful measure for comparing likely differences between different results measured using two methods. If one method is a reference method, compliance limits can be used as a measure of the total error of a measurement method (Krouwer, 2002). Odeh RE, Owen DB. Tables of normal tolerance limits, sampling plans and sampling. New York: Marcel Dekker; 1980.
Ba LoA are random estimates of actual but unknown population boundaries. Bland and Altman  originally proposed approximate confidence intervals to assess the accuracy of these estimates. In recent decades, the importance of confidence intervals has been emphasized in general [6.7] and for BA LoA in particular [8,9,10]. Carkeet  proposed accurate confidence intervals for BA LoA using bilateral tolerance factors for normal distribution. Zou  and Olofsen et al.  proposed confidence intervals for BA LoA for several observations in each compartment. They extended their work to freely available online implementation, accompanied by a formal description of more advanced Bland-Altman comparison methods , and proposed a standard format for the communication of a BA agreement analysis. However, recent examples of Bland-Altman plots with samples of 31 , 6 and 12  or 14  are readily available in MEDLINE/PubMed (by searching for “Bland-Altman” and “Accord” and limiting the search to publication dates in August 2019). These studies did not quantify the uncertainty of BA LoA`s estimate on the basis of confidence intervals. Vock  pointed out that a tolerance interval or external confidence limits can provide the entire BA LoA with an area that will contain at least a certain percentage of future differences with known certainty. Kottner et al.
 indicated that compliance and reliability assessment is conducted either in specific centre-of-gravity studies or in larger diagnostic precision studies, clinical trials or epidemiological studies that indicate compliance and reliability as quality control. The latter is often carried out in subsamples, which leads to reduced to moderate sampling sizes . The delimitation shows the theoretical correlations between the different shooting sizes in order to obtain precise confidence intervals. In addition, seemingly accurate methods of reconciliation, which are also far removed from the main treasures, have unwanted confidence limits. It is found that the optimal sample size has a median or medium minimum and increases when the percentile approaches the extremes. Classically, a BA analysis involves the study of the comparison of methods, i.e. the comparison between two methods or two instruments. The agreement of more than two methods, instruments or, in particular, advisors, is also an area for further research [73,74,75].