Closed-loop control of anaesthesia using the Bispectral Index

A.R. Absalom, G.N.C. Kenny. University Department of Anaesthesia, Glasgow Royal Infirmary, Glasgow, UK

 

We have built a computerised system for closed-loop control of anaesthesia, and present the results of a study of its performance in 10 patients undergoing total hip replacements under general anaesthesia supplemented by epidural anaesthesia.

 

Background

Our system consists of a Bispectral Index (BIS) monitor (A-1000, Aspect Medical Systems), a personal computer, and a target-controlled infusion (TCI) system. The BIS is downloaded every 5 seconds and passed to a PI (proportional-integral) control algorithm which compares the BIS value with the user-defined “target” value. The algorithm decides on adjustments to be made to the blood propofol concentration, and the TCI system then administers the new concentration, thus closing the loop.

 

Methods

After insertion and starting epidural anaesthesia, general anaesthesia was induced using the TCI system under the control of the anaesthetist. After the start of surgery, when the anaesthetist was confident the patient was adequately anaesthetised, automatic control of anaesthesia was initiated. The BIS at that moment was used as the target BIS. Physiological variables (heart rate, blood pressure, end-tidal carbon dioxide level, oxygen saturation and respiratory rate) and presence or absence of movement were recorded manually. Several electroencephalographic variables were automatically recorded onto the hard drive of the PC. Performance of the system was assessed using measures suggested by Varvel et al ¹. The day after surgery patients completed a satisfaction questionnaire.

 

Results

Table 1 summarises the BIS, and blood and effect site propofol concentrations during anaesthesia. During automatic control of anaesthesia the cardiovascular variables were stable (Figure 1). All patients breathed spontaneously, none had an oxygen saturation below 95% and none needed assisted ventilation. No patient moved during automatic control. Median Performance Error (MDPE) for the system was –2.06%, Median Absolute Performance Error was 8.26%, Wobble was 6.67% and the Median Offset was 1%. Figure 2 shows the absolute BIS error at different times during closed loop control of anaesthesia. All patients were completely satisfied with their anaesthetic, and there was no occurrence of awareness.

 

Discussion

Mortier has described a system using the BIS for automatic control of sedation². We have produced a system for automatic control of anaesthesia. It provided stable physiological conditions in a group of patients undergoing combined regional and general anaesthesia for hip replacement surgery. The accuracy of control of the BIS compared favourably with that reported by Mortier for propofol sedation, but was worse than that reported for closed loop control of neuromuscular blockade ³. In the first few cases we experienced problems with oscillation and found that control improved if a shorter bispectral smoothing interval was used. Later we also altered the control algorithm to take account of the burst suppression ratio, because in some cases there was a paradoxical increase in BIS during burst suppression. We are confident that with further adjustments, and the newer BIS software version, which has better algorithms for burst suppression, accuracy of control will be improved.

 

References

1.       Varvel JR, Donoho DL, Shafer SL. Measuring the predictive performance of computer-controlled infusion pumps. J Pharmacokin Biopharm 1992; 20: 63-94

2.       Mortier E, Struys M, De Smet T, Versichelen L, Rolly G. Closed-loop controlled administration of propofol using bispectral analysis. Anaesthesia 1998; 53: 749-754

3.       Kansanaho M, Olkkola KT. Performance assessment of an adaptive model-based feedback controller: Comparison between atracurium, mivacurium, rocuronium and vecuronium. Int J Clin Mon Comput 1997; 13: 217-224

 

Table 1: BIS, blood and effect site propofol concentrations at key events

 

 

 

 

 

 

Figure 1. Changes in heart rate and blood pressure. Symbols represent mean values and error bars the standard deviation.