Evaluation of two oxygen analyzers by computerized data acquisition and processing.

Monitoring of inspired oxygen concentration during anesthesia with nitrous oxide is becoming accepted as essential. This type of monitoring demands accurate monitors that respond rapidly. We evaluated two such devices for their response patterns to rapid changes in oxygen concentration, a galvanic or "fuel cell" unit and a polarographic device. Data were stored after analog-to-digital conversion. The response patterns to stepwise changes in nitrous oxide and oxygen mixtures were recorded at flow rates ranging from 2 to 10 L/min. Both units responded accurately to all changes in the absolute oxygen concentration; the polarographic unit was, on average, twice as fast. Responsiveness to nitrous oxide was low (less than 0.4% at 100% nitrous oxide), and the stability of the signals was good. The 90% response time (T90) was consistent for any stepwise increase or decrease in oxygen concentration between 0, 21, 33, 50, and 100%. After a step change from 0 to 100% oxygen at a gas flow rate of 10 L/min, the T90 was 5.8 seconds in the polarographic device and 11.4 seconds in the galvanic device (p less than 0.01). After a decrease from 100 to 0% oxygen, the T90 was 0.6 second longer in both monitors. Comparing flow rates of 2 L/min with 10 L/min, the T90 was delayed by 1.1 and 2.3 seconds for an increase, and by 1.4 and 2.9 seconds for a decrease in oxygen concentration. Experimental data suggest that both sensors respond adequately during routine clinical use. The faster response of the polarographic device is probably of limited clinical relevance, but it may aid in calibration.