Messung der Atem-Alkoholkonzentration mit einem neuen elektrochemischen Sensor Modelluntersuchung zur Querempfindlichkeit gegen�ber volatilen An�sthetika und klinische Anwendung

Abstract

Absorption of irrigating fluid in transurethral prostatic resection (TURP) and percutaneous nephrolitholapaxy (PNL) into veins or delayed absorption due to fluid extravasation may result in a TURP syndrome. The measurement of end-tidal breath alcohol concentration (et AC) as a monitor of absorption of irrigating fluid labelled with 2% ethanol is limited by the disturbance of infrared sensors by volatile anaesthetics and nitrous oxide (N2O) (Fig. 2). An electrochemical sensor is acceptable for this method. The aim of the present study was the evaluation of breath alcohol measurements using an electrochemical sensor device (Alcomed 3010, Envitec). The stability of the sensor in the presence of volatile anaesthetics was examined using a lung model. In a clinical investigation, the device was then applied to spontaneously breathing or mechanically ventilated patients inhaling volatile anaesthetics during endoscopic urological surgery. A two-chamber lung model filled with water for performing noninvasive measurements at the mouth of a patient has already been introduced by Brunner et al. (Fig. 1). With the addition of different amounts of ethanol to the temperature-controlled water, a constant ethanol concentration is achievable in the air above the water that is dependent on adjustments of the ventilator. Increasing concentrations of volatile anaesthetics (isoflurane, enflurane, halothane, and sevoflurane) were added to the fresh gas flow (2 l O2/3 l N2O) and etACs were measured using the manually triggered self-absorbent electrochemical sensor. First, regression equations were established between breath alcohol concentrations and increased volatile anaesthetic concentrations. Regression equations were then established between end-tidal anaesthetic gas concentrations and vaporizer adjustments in order to rule out an influence of ethanol on the anaesthetic gas monitor Ultima V (Datex). In the clinical investigation, 53 intubated and ventilated patients (33 undergoing PNL, 20 undergoing TURP) and 48 patients breathing spontaneously (32 with inhalation anaesthesia, 16 with spinal anaesthesia) were investigated. The etAC was measured with the Alcomed 3010 and compared with gas-chromatographically registered blood alcohol concentrations (BAC). The study had previously been approved by the Ethical Committee of the Medical University of Luebeck. Patients with liver disease and a history of toxic abuse were excluded. Only one value per patient (maximum BAC) was included in the statistics in order to avoid a cluster effect. The lung model experiments demonstrated that the measurement of etAC with an electrochemical sensor is free of interference by volatile anaesthetics (Table 1). The slope of the regression between the measured alcohol concentration and increased concentrations of anaesthetics did not differ significantly from baseline values. The measurement of end-tidal anaesthetic concentrations was not significantly different from vaporizer adjustments in the presence of increased alcohol concentrations (Table 2). During the clinical investigation, a regression between etAC and BAC was determined for both groups. For the group of patients breathing spontaneously, the correlation coefficient was 0.961 and the regression equation revealed etAC = 0.5677*BAC-0.1303 (Fig. 5). However, in the group of ventilated patients a biphasic course was shown that was dependent on BAC (Fig. 6). At BAC < 0.4%, a similar correlation (r = 0.856) to the spontaneously breathing group could be seen (regression equation: etAC = 0.617*BAC-0.020). Above 0.4% BAC there was no acceptable correlation (r = 0.444, regression equation: etAC = 0.202*BAC+0.104). The tested electrochemical sensor does not interfere with volatile anaesthetics and N2O as demonstrated by a lung model. There is a good correlation between etAC and BAC measurements in patients breathing spontaneously with special regard to the slope of the regression (s = 0.57).