A Randomised Controlled Trial of Standard Bilateral Electroconvulsive Therapy Versus High-dose Unilateral Electroconvulsive Therapy for Severe Depression

The study is a two-group parallel design randomised controlled non-inferiority trial and has been registered (ISRCTN23577151). Consented patients with major depressive disorder (DSM-IV) will be randomly allocated to a course of bitemporal (BT) ECT (1.5 x ST) or high-dose right unilateral (RUL) ECT (6.0 x ST). To facilitate generalizability of results, the trial takes place under "real world" conditions and so both groups continue usual care and medications during the treatment phase and thereafter. Patients are followed-up for 12 months after completing their allocated course of ECT. Completion of the primary outcome depression-rating measure (i.e. HDRS) and the secondary outcome of most interest (autobiographical memory, using the AMI-SF) will be prioritised in the data collection.

Patients, their treating clinicians and raters are blind to treatment; clinicians administering ECT are not involved in post randomisation assessments or formal data analysis. Success of blinding for patients and raters will be assessed after the second and final treatments. The trial statistician is also blinded to allocation status.

Sample size: In a large series (n = 253) of depressed patients, Petrides et al. (2001) found a mean (SD) reduction in 24-item HDRS of 25.6 (9.4) after treatment with bitemporalT ECT (1.5 x ST). We therefore estimate that 69 patients will be required per treatment group to have 80% power to demonstrate, using a one-sided equivalence t-test at 5% level, that mean reduction in 24-item HDRS achieved using high-dose RUL ECT is no more than 4 points (i.e. equivalent to 3 points on 17-item HDRS) less than that achieved using standard BT ECT, assuming a common within-group SD of change scores of 9.4 and equal expected group mean change scores.