Zou 2013 Limits Of Agreement

As a probability, CP offers an intuitive measure of compliance, which can be easily interpreted by users with almost all statistical refinements. It also requires a clinically acceptable difference (CAO) that must be predetermined before use, and the resulting interpretation is therefore directly related to the initial measurement scale. Carstensen B, Simpson J, Gurrin LC. Statistical models for assessing compliance in study comparison methods with replication measures. Int J Biostat. 2008;4(1):16. In this article, we showed how the limitations of the mixed effect of the tuning method were ideal for answering the question of which device was most consistent with the gold standard in measuring respiratory frequencies in COPD patients. The superiority of the boundaries of the tuning method over alternatives such as the calculation of correlation coefficients was discussed elsewhere. [2,4, 22, 23] For the sake of completeness, we presented the total limits of compliance with confidence intervals, but the real basis for decision-making was that of the participant`s interior and the entire standard deviations that allowed us to easily organize the devices to find out what quality was cut. This method is ideal for use in the situation where we are interested in comparing the advisor agreement or devices recording continuous measurements.

Repeated measures should be encouraged in some studies, as they allow us to quantify the compliance of the measures within the same theme and then compare them to the overall agreement. [2] The methodology used is relatively simple to use and should not be an obstacle to the analysis of measurement data repeated in practice. Lin L, Pan Y, Hedayat AS, Banhart HX, Haber M. A simulation study of the non-parametric global deviation index as a measure of compliance on the basis of quantimetric regression. J Biopharm Stat. 2016;26(5):937-50. The method of calculating compliance limits first involves calculating the average value and the standard deviation of type differences (e.g. B differences in respiratory frequency, measured simultaneously by the same participant with two different devices). The standard deviation is then multiplied by the 97.5% of a normal distribution (usually rounded to 2) and we add up or subtract that amount separately from the average calculated to obtain the upper or lower limit values. If m is the average value of the type differences and sd the standard deviation, the limits of the agreement are calculated as follows: m ± 2 SD. The limits of the agreement are expected to account for about 95% of the differences observed in the future; In reality, these are only estimates that are measured by uncertainty, and therefore 95% of confidence intervals are often calculated at the limits of the agreement itself.

The limits of the agreement must then be interpreted clinically to determine whether the agreement is acceptable or not. Ideally, the acceptable alangeder should be defined in advance in order to avoid any distortion in this decision. [4] Table 1 shows the numerical values of compliance limits calculated on the basis of i) of an ANOVA method with fixed effects, (ii) a mixed effects model using all possible data, and (iii) a mixed effects model after elimination of outliers.