B-type natriuretic peptide in high-risk major surgery patients.

criterion for assessment of observed variability. Given an inherent variability of +20% for each method under comparison, the combined variability (i.e. limits of agreement) should not exceed +30% of the mean SVV. Applying these strict criteria to the data, agreement of methods studied by the authors was unacceptable. Secondly, the authors compared SV derived from different methods at specified experimental conditions. Unfortunately, no information is provided as to the ability of these methods to reflect changes of SVV consistently. It has been suggested recently to analyse the change in a variable after a specific intervention also with a Bland–Altman plot, comparing the mean per cent change of both methods against the difference. This analysis would have revealed quickly, if these methods do track ensuing changes of SVV in a comparable fashion. This is a very important issue, since changes of SVV are thought to reflect a change in the fluid responsiveness of an individual patient. Thirdly, the authors compared methods during a single tidal volume (12 ml kg). This is quite surprising, as mechanical ventilation has an important impact on SVV, and consequently agreement between methods may be influenced by depth of tidal volume applied.