To the Editor: A recent article by Natalini et al. (1) suggested that ventilation-induced photoplethysmographic pulse variations (PPVpleth) were similar to those measured in arterial pulse pressure (PPVart). Agreement between both indices was in fact not very close (limits of agreement: −12% and 12%), but PPVpleth correctly identified patients likely to respond to fluid administration (as estimated from PPVart and ΔDownart measurements). In an accompanying editorial, Feldman suggested that reproducibility of their results may vary with oximeter models (2). Using a monitor from a different manufacturer, we recently found that PPVpleth predicted stroke volume (SV) response to volume loading with a threshold value similar to that in Natalini’s report (9.5% and 9%, respectively) (3). Agreement between PPVpleth and PPVart was also consistent between studies (−14% and 15% in our patients, who were more often hypotensive). However, PPVpleth predicted fluid responsiveness, and correlated with the fluid-induced change in SV, less accurately than PPVart or noninvasive PPV obtained using a Finapres device (3). PPVpleth is thus not equivalent to PPVart and should still be considered with caution. Natalini et al. also reported that photoplethysmographic systolic pulse variation (SPVpleth), but not its ΔDownpleth component, correlated with its arterial corresponding variable. This suggests that, in contrast with its arterial counterpart (4), ΔDownpleth may not predict fluid responsiveness. In an unpublished study (5), we measured these parameters in 19 patients in conditions apparently similar to those of Natalini’s study, and found similar correlation between SPVart and SPVpleth (r = 0.78; P = 0.001), and between ΔDownart and ΔDownpleth (r = 0.77; P = 0.001). In Natalini’s study, almost all patients received PEEP and many had negative ΔUp and ΔUppleth. This suggests that inaccurate measurement of ΔDown might have occurred. To be valid in the presence of PEEP, the “plateau” in systolic arterial pressure must be obtained at the level of PEEP. This necessitates an end-expiratory occlusion (6), a maneuver not possible with most respirators in the operating room. Reference systolic pressure measurement at airway pressure less than PEEP (typically when disconnecting the endotracheal tube) results in negative ΔUp and falsely increased ΔDown (depending essentially on the level of PEEP, lung compliance, and preload conditions). We thus believe that, similarly to arterial tracing analysis (4), correct quantification of the expiratory component of SPVpleth or, probably better, of PPVpleth (7), could be an informative complement to PPVpleth and deserves further evaluation. Benoît Tavernier, MD, PhD Marie-Sophie Destandau, MD Benoît Vallet, MD, PhD Federation of Anesthesia and Intensive Care Medicine University Hospital of Lille Lille Cedex, France [email protected]