ECT efficacy and treatment course: a systematic review and meta-analysis of twice vs thrice weekly schedules.

Background: Electroconvulsive therapy (ECT) guidelines, across various regulatory bodies, lack consensus as to the optimal frequency of treatment for individual patients. Some authors postulate that twice weekly ECT may have a similar efficacy to thrice weekly, and may have a lower risk of adverse cognitive outcomes. We did a systematic review and a meta-analysis to assess the strength of associations between ECT frequency and depression scores, duration of treatment, number of ECTs, and remission rates.

Methods: We searched on Medline, EMBASE, CINAHL and the Cochrane Central Register of Controlled Trials (to December 2009), and searched reports to identify comparative studies of frequency of ECT. We did both random-effects (RE) and quality effect (QE) meta-analyses to determine the risk of various outcomes associated with lesser frequency as compared to the thrice weekly frequency.

Results: We analysed 8 datasets (7 articles), including 214 subjects. Twice-weekly frequency of ECT was associated with a similar change in depression score (QE model SMD -0.11 [-0.55-0.33] and RE model SMD -0.17 [-0.77-0.43]) as compared to thrice weekly ECT. The number of real ECT's trended towards fewer in the twice weekly group. There was a statistically significant longer duration of treatment with a twice weekly protocol (QE model 6.48 days [4.99-7.97] and RE model 4.78 days [0.74-8.82]). There was a statistically significant greater efficacy for thrice weekly ECT compared to once weekly ECT (QE model SMD 1.25 [-0.62-1.9] and RE model SMD 1.31 [0.6-2.02]).

Conclusions: Twice weekly ECT is associated with similar efficacy to thrice weekly ECT, may require fewer treatments and may be associated with longer treatment duration when compared to thrice weekly. These epidemiological observations support the routine use of twice weekly ECT in acute courses, though choice of frequency should take into account individual patient factors. These observations have implications for resource utilisation e.g. costs of duration of admission vs cost of provision of ECT, as well as issues of access to inpatient beds and anaesthetist time.