We appreciate the comments by Drs BA Cotton, TD Girard, and EW Ely on our study(1). Regarding their concerns, we agree with them that it could be very useful to examine the relation between S100B and delirium during intensive care unit (ICU) stay and long-term cognitive outcome for the survivors. It has been shown that blood S100B protein may be related to postoperative delirium after abdominal surgery(2). However, the relation between S100B and delirium has not been studied in ICU until today. As far as we know, delirium should be diagnosed and measured after removal of sedation and analgesia. For this reason, mechanically ventilated patients have been excluded from nearly all delirium studies(3). Even the modified Confusion Assessment Method for use in ICU is not applicable in sedated patients(3). Since only patients under mechanical ventilation and sedation were included in our study(1), in a critical period of their disease course, representative of the more severely ill ICU population, withdrawal of sedation to evaluate delirium presence was not feasible. Of course, the contribution of brain dysfunction to the increased S100B level in ICU patients is not disputed at all, as we discussed it in our study. We have also commented on the limitations of our study regarding the use of Glasgow Coma Scale, which is a rough assessment of the level of consciousness. On the other hand, because increased levels of S100B are toxic and induce apoptotic astrocytes death(4), brain dysfunction and long-term cognitive dysfunction could be the result and not the cause of this increase. Furthermore, it is important to notice that the correlation of S100B concentration with lactate, pH, Hb, CaO2, MAP, and sequetial organ failure assessment score was significant even on the first day of the study, upon ICU admission, probably too early for the development of delirium. Because there is increasing evidence of extracerebral potential sources or causes of S100B as in trauma(5), hemorrhage(6), even in the presence of normal computerized tomography or magnetic resonance imaging(7), I/R of gut and viscera(8), stress(9), and exercise(10), other tissues and/or other conditions, in addition to those regarding the brain, may be implicated and represent a mighty stimulus for S100B release. The findings of the above-mentioned studies cannot be ignored and suggest a more global/complex role of S100B in overall homeostasis. Finally, the term "without brain injury" used in our study refers to the admission diagnosis indicating that no obvious cerebral damage was assessed in any patient before intubation and mechanical ventilation. We agree that a more detailed and long-term study design is necessary to include all possible factors that correlate with S100B increase and define its impact on cognitive function, possibly on the outcome and may be on other parameters, in critically ill patients. Christina Routsi, MD Elizabeth Stamataki, MD Seraphim Nanas, MD Charis Roussos, MD, PhD Medical School of Athens University Evangelismos Hospital Athens, Greece