Acta Anaesthesiol Scand 1999; 43: 470–475 Copyright C Acta Anaesthesiol Scand 1999 Printed in Denmark. All rights reserved ACTA ANAESTHESIOLOGICA SCANDINAVICA ISSN 0001-5172 A comparison of the monitors INVOS 3100 and NIRO 500 in detecting changes in cerebral oxygenation G. GRUBHOFER, W. TONNINGER, P. KEZNICKL, P. SKYLLOURIOTIS 1 , M. EHRLICH 1 , M. HIESMAYR and A. LASSNIGG Department of Cardiothoracic and Vascular Anaesthesia & Intensive Care, 1 Department of Cardiac and Thoracic Surgery, University Clinic of Vienna, Austria Background: Measurements of cerebral haemoglobin oxygen- ation of 2 near-infrared spectroscopy devices (INVOS 3100 and NIRO 500) were compared during and after hypocapnia. Methods: Fifteen awake, healthy volunteers, who hyperventi- lated to obtain end-tidal CO 2 (EtCO 2 ) values of approximately 20 mmHg, were studied. During hyperventilation and 8 min thereafter, EtCO 2 , INVOS 3100 (RSO 2 Ωregional cerebral oxygen- ation) and NIRO 500 recordings (HbO 2 Ωoxyhaemoglobin, HbΩ deoxyhaemoglobin, Hb-diffΩHbO 2 –Hb, CtO 2 Ωoxidised cyto- chrome oxidase aa3) were analysed. Results: Hyperventilation induced a significant decline in EtCO 2 from 30.5 to 14.7 mmHg (P∞0.001) and RSO 2 from 67.1% to 62.7% (PΩ0.025). At hypocapnia, only Hb (π1.61∫0.48 mmol/L; P∞0.001) and Hb-diff (ª3.01∫2.0 mmol/L; P∞0.001) indicated a decline in cerebral haemoglobin oxygenation. Within 8 min after hyperventilation, both EtCO 2 and RSO 2 normalised to values insignificantly different from baseline. In contrast, Hb and Hb- diff remained significantly different (Hb: π2.52∫1.28 mmol/l; M ONITORS using near-infrared spectroscopy (NIRS) can determine cerebral oxygenation non-invasively and are suggested to have important clinical applications (1, 2). NIRS has been validated in clinical studies (3–7), as well as in experimental set- tings changing cerebral oxygenation by means of hy- pocapnia (8), increasing PaCO 2 (9, 10) or hypoxia (11– 14). However, only a few comparisons concerning the accuracy of various NIRS devices are available (15, 16). In this study, decreases in cerebral haemoglobin oxygenation state were induced by hypocapnia. Sys- temic hypocapnia is known to impair cerebral oxygen delivery by decreasing cerebral blood flow (CBF), which will be compensated for by an increased extrac- tion rate of oxygen followed by an increase in deoxy- genated haemoglobin in cerebral venous vessels (17, 18). Furthermore, severe systemic hypocapnia reduces the oxygen pressure in cerebral tissue (19–21) exten- sively enough to induce anaerobic glycolysis and lac- tate production in cerebral cells (20, 22, 23). 470 P∞0.001, Hb-diff: ª4.31∫4.0 mmol/L; P∞0.001). A correlation with EtCO 2 was found for RSO 2 (RΩ0.35; P∞0.001) and CtO 2 (RΩ0.42; P∞0.001). All volunteers were continuously awake and none presented clinical symptoms of cerebral hypoxia. Conclusion: Changes in cerebral haemoglobin oxygenation state were reflected more accurately by INVOS 3100 than NIRO 500. The cause may be the different technology of the monitors, since INVOS 3100 eliminates the contribution of extracranial oxygen- ation. Received 8 April, accepted for publication 10 November 1998 Key words: Brain metabolism; oxygenation, cerebral; near-in- frared spectroscopy; hemoglobins; oxyhemoglobins; carbon dioxide physiology; comparative study. c Acta Anaesthesiologica Scandinavica 43 (1999) During and after moderate hypocapnia, we com- pared the measurements of cerebral haemoglobin oxygenation of 2 commercially available devices (IN- VOS 3100 and NIRO 500). INVOS 3100 uses 2 receiv- ing optodes, which is supposed to eliminate extra- cranial contribution, whereas NIRO 500 is a conven- tional NIRS monitor. For comparison of the monitors, the following questions have been explored: 1) Are changes in cerebral haemoglobin oxygen- ation tracked equivalently by the monitors? 2) Do the measures of both monitors correlate with the degree of hypocapnia? 3) Does the trend of Cytochrome aa3, a unique measure of NIRO 500, follow the changes in oxidised haemoglobin? Patients and methods Fifteen right-handed healthy volunteers not receiving any medication, 8 male and 7 female (mean age 35 years, range 22–43 years), were included in the study.