The efficacy of electroconvulsive therapy in a perinatal population : a comparative pilot study

Haxton, Chris and Kelly, Stephen and Young, David and Cantwell, Roch (2016) The efficacy of electroconvulsive therapy in a perinatal population : a comparative pilot study. Journal of Pain. pp. 113-115. ISSN 1095-0680 (

[thumbnail of Haxton-etal-JOP-2015-The-efficacy-of-electroconvulsive-therapy-in-a-perinatal-population]
Text. Filename: Haxton_etal_JOP_2015_The_efficacy_of_electroconvulsive_therapy_in_a_perinatal_population.pdf
Accepted Author Manuscript
License: Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 logo

Download (591kB)| Preview


This study aims to investigate the effectiveness of electroconvulsive therapy (ECT) as a treatment for postnatal depression compared with a matched nonpostnatal population. A secondary aim is to compare the number of ECT treatments needed to treat in both groups. As the numbers in this study are small, this would act as a pilot study, allowing a power calculation to determine the numbers needed for a more definitive study. Cases were identified from the local Scottish ECT Accreditation Network database. Twelve patients had received ECT and had complete data from all those admitted to the Glasgow Mother and Baby Unit since its opening. Each case was matched to 2 control subjects who had also received ECT and who were matched for age, sex, and severity of depressive symptoms. It was not possible to find 2 control subjects for each case, and 23 control subjects were allocated. Severity was matched using the Montgomery-Åsberg Depression Rating Scale (MADRS), which is completed for all patients receiving ECT in Scotland, at the beginning and end of their course. As all control subjects were matched for initial severity of symptoms using MADRS, the change in MADRS score between both groups was compared. When comparing the mean change in MADRS scores between both groups, it was seen that the perinatal patients scores dropped by a mean of 10.09 points more than those of the control subjects (95% confidence interval for difference, −0.54 to 20.73, P = 0.062). No difference was seen between the groups when comparing the number of treatments (7.8 vs 8). Further research is required.