Making physical treatments for depression easier to reach
Test clinical benefit and whether people can complete treatment. Preserve urgent care and measure the time and costs borne by patients and providers.
First comparison to run
Start with adults with treatment-resistant major depression who are eligible for non-urgent outpatient iTBS. Compare standard specialist-center iTBS with the same stimulation course delivered at a regional site, with remote follow-up where appropriate. Assess symptoms, function and safety through 12 weeks before judging differences in visits and total burden. Test compressed schedules and image-guided targeting in separate comparisons to estimate their individual effects and costs.
Scope of the evidence
Most cited trials concern acute treatment in specific depression populations. Eligibility, comparators, scales, and follow-up differ, so the estimates cannot rank treatments or establish superiority across trials. Preserve current clinical standards, including ECT and urgent care pathways. This is a proposed access study for specialist care and ethics review. No new trial was conducted, and Korean approval or availability was not verified.
Published evidence
ECT in severe and urgent care
NICE NG222 considers ECT for severe depression after prior benefit and patient choice, when a rapid response is needed, or when other treatments have failed. Assess cognitive, anaesthetic, and delay risks. The US FDA device classification covers catatonia or a severe major depressive episode in major depressive or bipolar disorder at age 13 or older with treatment resistance or a need for rapid response due to severity.
A direct comparison of ECT and the rTMS used at the time
Eranti and colleagues randomised 46 people with severe depression referred for ECT to ECT (22) or repetitive transcranial magnetic stimulation, rTMS (24). At the end of acute treatment, 13 ECT participants (59.1%) and 4 rTMS participants (16.7%) met the Hamilton-scale remission criterion. These estimates apply to a small referral sample and the stimulation protocol reported in 2007.
THREE-D reduced stimulation time to three minutes
THREE-D randomised 414 people to standard 10 Hz rTMS or intermittent theta burst stimulation, iTBS. At treatment end, iTBS met the trial's non-inferiority criterion on the 17-item Hamilton Depression Rating Scale. Stimulation time dropped from about 37.5 to 3 minutes per session, and both groups attended on weekdays over 4–6 weeks. Scores followed similar trajectories through 12 weeks after treatment. The 12-week scale score was available for 264 participants, so missing follow-up matters when interpreting durability.
Provider costs modeled from the same trial
A 2019 Canadian analysis of THREE-D estimated direct health-system costs per acute course at USD 1,108 for iTBS and USD 1,844 for 10 Hz rTMS in 2018 dollars. It modelled equipment and staff costs using patient-level trial visits. Patient travel, time and post-course costs were outside its scope. Estimating total costs through 12 weeks requires separate local inputs.
Fewer treatment days without personalised imaging
Ramos and colleagues randomised 100 adults with treatment-resistant depression to accelerated theta burst stimulation using standard scalp targeting or to sham. Eighty-nine completed the randomised phase. At week five, response was 26/50 with active treatment and 11/50 with sham. The optional continuation phase cannot establish 12-week durability against standard iTBS. In 2026, US FDA clearance K260189 expanded the labelling of a specific device to cover accelerated schedules without relying on an individualised functional MRI targeting algorithm. A direct comparison with the standard schedule should measure treatment days, time on site per day, sustained response and safety in the same eligible population.
Two sham-controlled trials of accelerated SNT
Stanford neuromodulation therapy (SNT) combines functional MRI-guided personalised targeting and accelerated high-dose iTBS. In 2022, 29 people received their assigned treatment. Mean MADRS reductions four weeks later were 52.5% with active stimulation and 11.1% with sham. In the 2026 replication, one-month remission was 12/24 (50.0%) with active treatment and 5/24 (20.8%) with sham. Score reduction and remission are different outcomes. Both trials used sham controls. The contribution of targeting and the benefit against standard iTBS require separate comparisons.
The US FDA clearance for the SAINT device
In September 2022, the FDA granted 510(k) clearance K220177 to the Magnus Neuromodulation System with SAINT Technology. The US prescription indication covers adults with major depressive disorder who had inadequate improvement on prior antidepressant medication in the current episode.
Trials that compare targeting within a treatment course
Siddiqi and colleagues scanned all 40 adults with treatment-resistant depression before comparing connectivity- and scalp-based targeting within the same accelerated TMS course. At one month, median MADRS reductions were 24 and 18 points. Li and colleagues' single-centre trial randomised 123 people with major depressive disorder or bipolar II depression; 119 entered the modified intention-to-treat analysis. Structural and functional connectivity targeting outperformed the 5 cm target at six weeks, but the difference was not statistically significant at 12 weeks. In the multicentre BRIGhTMIND trial of 255 people, connectivity-guided iTBS and standard rTMS did not differ significantly on the primary outcome over 26 weeks. Both target selection and stimulation differed, so BRIGhTMIND cannot isolate the contribution of imaging. Durability and the added cost of imaging still need testing.
Two home-based direct current stimulation trials with differing results
In Woodham and colleagues' 174-person randomised trial, 10 weeks of remotely supported home transcranial direct current stimulation, tDCS, improved Hamilton depression scores by 2.27 points more than sham on average (95% CI 0.51–4.01). In Borrione and colleagues' 210-person, three-arm trial, six weeks of home tDCS, with or without a digital intervention, showed no statistically significant score difference versus sham. Duration, support, and comparators differed. The contribution of remote supervision would require a direct test under matched conditions.
The US prescription approval for Flow FL-100
On 8 December 2025, the US FDA granted premarket approval P230024 to Flow FL-100. It covers monotherapy or adjunctive treatment for the current moderate-to-severe episode of major depressive disorder in adults aged 18 or older who are not considered refractory to medication. It is a prescription device used under clinical care.
A trial of bright light in nonseasonal depression
Lam and colleagues assigned 122 adults with nonseasonal major depressive disorder to bright light, fluoxetine, the combination, or placebo for eight weeks. Light alone and the combination produced larger reductions in the primary outcome, MADRS score, than placebo. The study covered acute nonseasonal depression. Longer-term durability, mood elevation risk, and use in bipolar depression require separate evaluation.
Exploratory signals from low-intensity focused ultrasound
Oh and colleagues' 2024 sham-controlled trial randomised 26 people, of whom 23 completed the study. The active arm showed a signal of improvement in depression symptoms. Riis and colleagues randomised 22 people in a crossover trial using a different brain target. The 19-person per-protocol analysis found a difference in depression scores at 24 hours. The difference was not statistically significant in the full randomised sample. Two participants experienced worsening depression and suicidal ideation 24–72 hours after active stimulation. Whether stimulation caused these events remains uncertain. These small, early trials with distinct targets leave reproducibility, durability and harms to be tested.
Open questions
Where do people struggle to start or finish treatment?
Record referral delays, cancellations, missed visits, completion, and reasons. Count stimulation, time on site, travel, visits, and total course days separately. Measure patient expenses and time away from work or caregiving, plus provider staffing, facility and device costs, and capacity.
Do symptoms and daily functioning hold up when visits become easier?
Compare the same iTBS course delivered through a specialist centre or a regional site. Measure symptoms, daily functioning, deterioration, and rescue care at treatment end and 12 weeks later. Assess access gains after meeting prespecified clinical and safety criteria.
Whose burden falls with local delivery and remote oversight?
Prespecify comparisons by travel distance, mobility constraints, and work or caregiving schedules. Track whether local delivery or remote assessment cuts travel and whether technical failures, on-site support, staffing, or emergency response add burden elsewhere. Record who handles missed treatment and safety alerts and how quickly they respond.
Does personalised targeting justify the imaging and preparation time?
When comparing personalised with standard targeting, hold the stimulation course and delivery setting constant and randomise target selection. Compare imaging and analysis time, cost, and the share of patients who can access them alongside symptoms, functioning, and durability.
Proposed study
Randomise adults eligible for non-urgent outpatient iTBS to a specialist-centre pathway or a regional-site pathway. Deliver the same stimulation course in both groups. The regional pathway uses remote follow-up where appropriate. This first trial estimates the effect of the delivery pathway as a whole; it does not isolate local care from remote assessment. Route urgent ECT or crisis cases through established care. Before enrolment, register sample size, outcome scale, timing, a population-specific non-inferiority margin, and rules for missing data, treatment changes, and rescue care. Independent blinded raters assess treatment end and 12 weeks later. Report intention-to-treat and per-protocol confidence intervals. Monitor functioning, memory and cognition, mania or hypomania, suicide risk, worsening, and rescue care. Test accelerated schedules and personalised targeting separately. A clinical team oversees remote assessment. Minimise logs for scheduling, oversight, technical failures, and escalation. Restrict and audit access. Evaluate access and costs after meeting clinical and safety criteria.