We are thankful for Dr. Morey’s excellent comments regarding possible inaccurate estimates in limits of agreement related to repeated measurements in the same subject with regards of our recent paper.1
This issue was raised during statistical revision of the paper, where we first presented our results with a Bland–Altman analysis using mean cardiac output values for both methods in all 43 subjects (resulting in only 43 x–y pairs for comparison). As recommended by the statistical reviewer, this was corrected to a Bland–Altman analysis of each repeated measurement (using all 215 x–y pairs), assuming a nonconstant change in cardiac output. This correction did not have any remarkable effect on the limits of agreement (0.41 to –0.24 for 43 x–y pairs and 0.40 to –0.24 for 215 x–y pairs). We then tried to apply the recommended modified Bland–Altman method for repeated measurements2 but for some reason were not able to reach assumed estimates. Instead we used the approach recommended by Zou to estimate limits of agreement3 and found that it was in agreement with our previous limits of agreement estimations (0.40 to –0.25 using Zou’s method for 215 x–y pairs). We had not seen the paper by Myles and Cui4 expanding on the modified Bland–Altman method.
Although it is important to be aware of possible changes in limits of agreement regarding Bland–Altman analysis and repeated measurements in the same subject, we did not notice any relevant change in limits of agreement when Zou’s method was applied to our data. Estimated bias, precision, and percentage error in our paper should not be affected by any changes in the limits of agreement due to possible fluctuations in cardiac output between measurements.
The author declares no competing interests.