Table 2.

Comparison of proposed reporting items for a Bland–Altman analysis across included studies.

Reporting Item	Flegal (2019) [30]	Abu-Arafeh (2017) [23]	Montenij (2016) [31]	Olofsen (2015) [13]	Chhapola (2015) [22]	Berthelsen (2006) [33]	Mantha (2000) [32]
(1) Pre-established acceptable limit of agreement	X	X	X		X	X	X
(2) Description of the data structure (e.g., no. of raters, replicates, block design)		X	X	X
(3) Estimation of repeatability of measurements if possible (mean of differences between replicates and respective standard deviations)		X		X	X	X	X
(4) Plot of the data, and visual inspection for normality, absence of trend, and constant variance across the measurement range (e.g., histogram, scatter plot)	X	X	X	X	X	X	X
(5) Transformation of the data (e.g., ratio, log) according to 4), if necessary		X				X
(6) Plotting and numerically reporting the mean of the differences (bias)	X	X	X	X
(7) Estimation of the precision, i.e., standard deviation of the differences or 95% confidence interval for the mean difference	X	X	X	X		X
(8) Plotting and numerically reporting the BA LoA	X	X	X	X	X
(9) Estimation of the precision of the BA LoA by means of 95% confidence intervals	X	X	X	X	X	X	X
(10) Indication of whether the measurement range is sufficiently wide (e.g., apply the Preiss-Fisher procedure [34])		X
(11) Between- and within-subject variance or stating that the confidence intervals of the BA LoA were derived by taking the data structure into account		X	X	X
(12) Software package or computing processes used		X		X
(13) Distributional assumptions made (e.g., normal distribution of the differences)	X	X	X
(14) Sample size considerations			X
(15) Correct representation of the x-axis					X		X
(16) Upfront declaration of conflicts of interest						X

BA LoA: Bland–Altman Limits of Agreement.