Depth of Anesthesia During Electroconvulsive Therapy

Background Electroconvulsive therapy (ECT) is an effective therapy for severe depression where an electric current is applied to the brain to induce an epileptic seizure (1). The success of ECT is dependent upon the degree and quality of epileptic activity (2). ECT is normally administered after administering a muscle relaxant to prevent the seizure activity from causing muculoskeletal injuries. To tolerate the muscle relaxant the entire procedure is performed under general anesthesia. Most hypnotics have anti-convulsant effects and therefore counteract the effects of ECT (3). This implies that deep anesthesia hinders an optimal treatment effect of ECT which is supported by observational data (4). This relation has been proven with lighter anesthesia measured through bispectral index (BIS) correlating to longer seizure duration (5). BIS is not optimized to estimate the depth of anesthesia under very dynamic anesthetic protocols as there is an inherent latency in the computation of the BIS value. Anesthesia for ECT generally consists only of induction and recovery with no stable anesthetic phase making BIS an inappropriate measure of anesthetic depth. Indices of anesthetic depth that are derived directly from the processed EEG-signal have a negligent latency and can therefore be used with more precision during anesthesia for ECT. Spectral edge frequency (SEF) is the frequency under which 95 % of the total amplitude of the EEG-signal is found and it decreases with increasing depths of anesthesia. The density spectral array (DSA) is a graphic representation of the frequency and amplitude of the EEG signal over time and it is possible to interpret the DSA in real time without extensive training. The SEF and DSA have not, to our knowledge, been previously tested during general anesthesia for ECT. In the postictal phase during recovery from ECT some patients become agitated and confused (6). This condition is not clearly defined in the literature but is commonly referred to as postictal agitation (PIA) and due to differing definitions the incidence is difficult to estimate. Incidences in the range of 10-20 % have been reported (7). In the post-operative period after cardiac surgery decreased regional cerebral oxygen saturation (SrcO2) has been linked to post-operative confusion (8). A possible link between decreased SrcO2 and PIA has not been studied before. Aim The aim of this research project is two-fold: 1. To study the relation between anesthetic depth measured with SEF and DSA and the quality of epileptic activity during ECT. 2. To study the relation between SrcO2 and PIA during ECT. Hypotheses 1. Lighter depths of anesthesia as measured by SEF and DSA have a positive correlation with epileptic seizure quality during ECT. 2. SrcO2 during ECT has an inverse correlation with the incidence of PIA. Method Setting and recruitment The study will be performed at the ECT-unit of the university hospital in Örebro. General anesthesia is provided by a nurse specially trained in anesthesia and an anesthesiologist. ECT is provided by specially trained nurses and staff at the ECT-unit. Patients recover from the postictal phase in an adjacent recovery room. Patients undergoing ECT through the outpatient clinic will be recruited to the study. Anesthetic protocol The protocol for anesthesia and ECT will be the same as is being used clinically at the ECT-unit. After preoxygenation anesthesia is induced using thiopental. After confirmation of adequate anesthetic depth (failure to respond to verbal command) suxamethonium is administered. Doses are chosen by the anesthesiologist. Patients are admitted to the recovery room when breathing spontaneously and can maintain a patent airway. Data collection SEF and DSA will be measured continuously from before pre-oxygenation to admittance to the recovery room using a SedLine® (Masimo, Irvine, CA, USA) monitor. The monitor will be blinded during the procedure so as not to affect the staff in any way. SrcO2 will be measured using the same monitor during the same period. Data regarding the seizure duration will be gathered from the electronic patient records. The staff caring for the postictal patients will estimate the degree of PIA according to previously used numerical scale (9).Data will be presented as mean ± SD or median (IQR) according to their distribution. Statistical analysis will be performed using a logistic regression model adjusted for potential confounders such as type of psychiatric medication, age, bilateral or unilateral stimulation.