ECT is usually described as a course of treatment with an end point. But depression has a habit of returning, which raises an obvious question: what happens afterwards?
Researchers looked at everyone in Denmark who started ECT over a twenty-year period — 19,944 people in total. Around 7.7% went on to have continuation or maintenance ECT, meaning further treatments spaced out over months rather than a single intensive course.
The people who received ongoing ECT were not a random sample. They were more likely to have diagnoses such as schizophrenia or schizoaffective disorder, which tells you something about who clinicians judge to need continued treatment.
Studies at this scale are unusual and valuable, because they show what actually happens in ordinary clinical practice rather than in the tightly controlled conditions of a trial. Real patients, real decisions, over two decades.
If ECT is part of your care or a family member's, the question of what comes next is a fair one to raise early rather than at the end of a course. Knowing the plan tends to make the whole process less frightening — and this research shows ongoing treatment is a recognised path, not an improvisation.
Electroconvulsive therapy (ECT) is typically given as an acute course (aECT) — a set of treatments delivered close together until symptoms improve. But some patients continue receiving spaced-out treatments afterwards, either to consolidate the response (continuation ECT, cECT, given within 180 days of finishing the acute course) or to prevent long-term relapse (maintenance ECT, mECT, given beyond that window). This study, by Jørgensen et al., asked how often this happens in real practice, who receives it, and whether it works.
How the study was designed
This was a cohort study — a design that follows a defined group of people over time and compares outcomes between subgroups, rather than randomly assigning treatments as a clinical trial would. The researchers used Denmark's National Patient Registry to identify everyone who started ECT between 2003 and 2022, then built an algorithm to classify treatment patterns as aECT-only, cECT, or mECT based on the timing and spacing of sessions.
To estimate c/mECT's effect on later hospitalisation and suicidal behaviour, they used several complementary statistical approaches: Cox proportional hazard regression (which estimates the relative risk, or 'hazard', of an event occurring over time between groups, expressed as a hazard ratio), propensity score matching (which pairs patients with similar characteristics who differ mainly in treatment received, reducing bias from the fact that treatment wasn't randomly assigned), and self-controlled case series analysis using Poisson regression (which compares each patient's own risk of an event during different time periods, effectively using patients as their own control). Using multiple methods that rely on different assumptions strengthens confidence when they agree.
What it found
Out of 19,944 people treated with ECT, 1,533 (7.7%) went on to receive c/mECT — 5.1% continuation only, 2.6% maintenance. This group was more likely to have schizophrenia (odds ratio 2.14) or schizoaffective disorder (odds ratio 2.42), and less likely to have unipolar depression (odds ratio 0.56), compared with those who received aECT alone. An odds ratio above 1 means the outcome was more common in the c/mECT group; below 1, less common.
Across all statistical models, c/mECT was linked to lower hospitalisation rates after the acute course ended — for example, a 6-month adjusted hazard ratio of 0.68 and an incidence rate ratio of 0.51, both indicating meaningfully reduced risk relative to aECT alone. There was no significant difference in suicidal behaviour between groups. The study also found that c/mECT was associated with substantially lower overall treatment costs compared with the period before the acute course ended.
Why it matters
Because this drew on two decades of nationwide, real-world data rather than a controlled trial, it captures how clinicians actually use c/mECT and what happens to patients afterwards — information that's hard to get any other way. The authors conclude that c/mECT, though rarely used, may deserve wider consideration for preventing relapse in patients who haven't responded sufficiently to other treatments.
Limits worth knowing
Because treatment wasn't randomly assigned, patients who received c/mECT may have differed from others in ways the statistics couldn't fully capture, even with matching and multiple models. The findings describe association, not proof of cause and effect.
If you or someone you know is considering ECT or navigating what comes after it, it's worth talking this through with a psychiatrist — and reaching out for support is always a reasonable step.