The influence of frontal muscle electromyography on electroencephalography-based depth of anaesthesia monitoring

Abstract

Yleisanestesia koostuu kolmesta eri osatekijästä, jotka ovat tajuttomuus, reagoimattomuus kirurgian aikaansaamille ärsykkeille ja liikkumattomuus. Tajuttomuutta eli anestesian hypnoottista komponenttia voidaan mitata erilaisilla aivosähkökäyrän (EEG) tulkintaan pohjautuvilla laitteilla. Näistä laitteista onkin tullut hyvin suosittuja, sillä niiden on todettu vähentävän leikkauksenaikaisen hereillä olon riskiä ja nopeuttavan anestesiasta toipumista. Nämä laitteet rekisteröivät aivosähkökäyrää potilaan otsalta ja ohimolta. Rekisteröity aivosähkökäyrä muutetaan laitteen toimesta lukuarvomuotoon, ja kullekin laitteelle on annettu suositus siitä, mitä lukuarvoja anestesian aikana tulisi tavoitella. Otsalihasten lihassähkökäyrä (EMG) on yksi häiriötekijä, joka saattaa tehdä näiden laitteiden tulkinnasta vaikeaa tai jopa mahdotonta.

Tämän väitöskirjan tutkimuksissa selvitettiin 133 nukutetulla potilaalla, kuinka EMG:n ja EEG:n reaktiivisuus kivuliaalle ärsykkeelle vaikuttaa kahden anestesian syvyyden mittaamiseen tarkoitetun laitteen (BIS ja Entropy) toimintaan. Lisäksi tutkittiin, voiko anestesiologi nähdä EMG:n anestesiamonitorilta paljaalla silmällä, jolloin laitteen antama virheellisen korkea lukema ei johtaisi tarpeettomaan anestesia-aineiden lisäämiseen.

Väitöstutkimus osoitti, että 1) voimakas EMG-reaktio nostaa anestesian syvyyden mittaamiseen käytettyjen indeksien lukuarvoja, 2) EEG:n ja EMG:n taajuusalueet ovat osittain päällekkäisiä, joten niiden erottaminen Entropia- ja BIS-laitteiden algoritmeilla on käytännössä mahdotonta, 3) leikkauksen loputtua tapahtuva, lihasrelaksaation kumoamisen aiheuttama nousu Entropyn ja BIS:n lukuarvoissa on mitä todennäköisimmin EMG:n aiheuttamaa, 4) anestesian aikaiseen monitorointiin tarkoitettuja laitteita tulee edelleen kehittää, ja 5) anestesiologin tulisi tietää riittävästi anestesian aikaisista EEG:n ja EMG:n muutoksista, jotta vältettäisiin anestesia-aineiden ali- tai liika-annostelu.

During the last 15 years, monitoring the hypnotic component of general anaesthesia has become routine in clinical practise. Several commercially available methods have been introduced for this purpose. These monitors are based on either spontaneous electroencephalogram (EEG) or mid-latency auditory-evoked potentials (MLAEP). These monitors have been reported to decrease the incidence of intraoperative awareness, to reduce the amount of anaesthetic drugs given to the patient, and to reduce recovery time after anaesthesia. The most widely used of these monitors is the Bispectral Index Scale (BIS). Another EEG-based depth of anaesthesia monitor is Entropy. The EEG analysed by these depth of anaesthesia monitors is collected from the forehead of the patient, therefore, rendering the methods vulnerable to confounding factors. One obvious confounding factor is the electromyographic (EMG) arousal caused by the frontal muscles. Also, different types of EEG arousal can produce changes in the numerical values of the depth of anaesthesia monitors without a change in the actual depth of anaesthesia, making the interpretation of depth of anaesthesia monitors very challenging for the anaesthesiologist. The aim of this doctoral thesis was to study the effect of EMG and EEG arousal on the numerical values of BIS and Entropy, and to investigate if an anaesthesiologist is able to detect EMG arousal visually on the anaesthesia monitor.

In Study I, 31 patients undergoing gynaecological surgery under general anaesthesia were studied. The patients were anaesthetized with propofol-nitrous oxide (N2O) or propofol-N2O-remifentanil. The one-channel EEG collected by the Entropy strip was analysed off-line. Intubation of the trachea produced EMG arousal in 13/16 and 15/15 patients with and without remifentanil, respectively. EMG arousal caused a significant rise in Entropy values, even during propofol-induced burst suppression. The EMG activity started at frequencies below 20 Hz, contaminating the interpretation of the Entropy values.

In Study II, 38 patients were studied during skin incision, a typical noxious stimulus. Again, the one-channel EEG collected by the Entropy strip was analysed to discover the incidence of skin incision-associated EMG and EEG arousal under sevoflurane-N2O or sevoflurane-N2O-rocuronium anaesthesia. Skin incision caused EMG arousal in 0/19 and 13/19 patients with and without rocuronium, respectively. Skin incision produced EEG arousal in 17/19 and 15/19 patients with and without rocuronium. EEG arousal was classified as beta arousal in 30 patients, causing an increase in Entropy values. EEG delta arousal (two patients) caused a decrease in Entropy values. The power spectra of EEG and EMG overlapped significantly.

In Study III, during light propofol-remifentanil anaesthesia, the neuromuscular blockade (NMB) of 30 patients was antagonized with neostigmine or sugammadex, and the numerical values of BIS and Entropy were studied. The reversal of NMB caused a strong rise in the numerical values of Entropy and BIS in 5/15 patients with neostigmine, and in 5/15 patients with sugammadex. The EEG analysis suggested that the cause for the increasing values was the EMG contamination of EEG. In patients without EMG arousal, no EEG change was seen, and Entropy and BIS values remained unchanged.

In Study IV, the EEG information seen on the anaesthesia monitor of 34 patients under general anaesthesia was recorded with a high-definition video camera, and compared with the one-channel EEG collected by the Entropy monitor. The EMG arousal was only partly detectable on the anaesthesia monitor using 100 ?V scale, suggesting that technological improvements to the depth of anaesthesia monitoring are warranted. In conclusion, the numerical values of depth of anaesthesia monitors are affected by EMG and EEG arousals. EMG arousal and EEG beta arousal increase the index values, while EEG delta arousal decreases the index values. The rise in the numerical values of BIS and Entropy during the reversal of NMB is most likely caused by EMG; however, a small change in EEG may be difficult to see underneath a strong EMG activity. Finally, in order to make detection of EMG arousal more accurate, technological improvements are warranted. However, anaesthesiologists using depth of anaesthesia monitoring should also have the skills to interpret the raw biosignal on the monitor screen, in order to avoid over- and underdosing of anaesthetic agents.