A shorter treatment schedule can change what depression care feels like: fewer weeks organized around appointments, a different burden of travel and the possibility of earlier improvement. Accelerated TMS is scientifically interesting for the same reason. Compressing time, however, changes a protocol; it does not make every accelerated schedule the same treatment. [1]
Begin with three different meanings of “faster”
The THREE-D trial compared brief intermittent theta-burst stimulation with a conventional longer high-frequency rTMS session. It supported noninferiority for the depression outcome studied. That finding concerns the duration of a daily session within a treatment course. It should not be mistaken for evidence that an entire course can be delivered over a few days with identical results. [1]
An accelerated protocol delivers multiple sessions within a day. SNT adds a particular high-dose schedule and individualized functional-connectivity targeting. When a clinic or paper uses the word “accelerated,” ask which of these features is present: shorter sessions, more sessions each day, a particular interval between them, a different cumulative exposure, or imaging-based targeting. A shared name can hide substantial differences in the intervention. [2]
What the original controlled SNT trial showed
The original double-blind SNT trial treated 29 participants: 14 with active stimulation and 15 with sham. At four weeks, the mean reduction in MADRS scores was 52.5% with active treatment and 11.1% with sham. Those figures describe average percentage symptom reduction; they are not remission rates. The protocol used resting-state functional MRI to choose an individualized left prefrontal target. [2]
The trial gave a strong reason to pursue the method, while its small size limited precision. Dramatic percentages from early open-label work are especially vulnerable to being treated as stable expectations. A controlled comparison, uncertainty around the estimate and an independent or enlarged replication are more useful than a memorable headline number when deciding what a field has established.
The 2026 replication: encouraging, still bounded
A 2026 World Psychiatry report randomized 48 participants who met entry criteria, with 24 receiving active and 24 sham SNT. At one month, remission occurred in 50.0% of the active group and 20.8% of the sham group; response also favored active treatment. This provides a further controlled efficacy signal and a more grounded estimate than relying on an uncontrolled early series. [3]
The same report explored EEG findings, including left frontal beta-power changes and associations involving anterior cingulate beta activity. These are candidate markers of mechanism and outcome. A clinically useful prediction tool would still need reproducible measurement, prespecified thresholds and evidence that using it improves treatment decisions in new patients. A correlation within a promising trial is an important starting point, not the end of that process. [3]
Can acceleration work without expensive targeting?
A separate sham-controlled trial of 100 participants tested a pragmatic schedule using three iTBS sessions a day and a non-neuronavigated approach. At week five, mean HDRS-17 improvement favored active treatment. During the subsequent open-label phase, approximately three quarters received additional sessions. This is relevant evidence for service design, but it is not a direct comparison with the SNT protocol. [4]
The study also illustrates why a treatment advertised as a short course may still require further appointments. Distinguish the prespecified controlled endpoint from later open-label improvement and additional treatment. For patients and services, the complete pathway matters: assessment, the initial course, review, any additional sessions and the plan if symptoms return. [4]
Network findings are not yet a routine clinical compass
A randomized SNT imaging analysis investigated changes in connections involving the stimulation target and wider networks. It found a post-treatment difference involving the default mode network, but did not find significant group-by-time effects for every hypothesized connection. That mixed pattern is scientifically useful. It reminds readers that a plausible network story should be judged against all the planned comparisons, not only the positive figure. [5]
A brain image can make a mechanism feel more certain than the data justify. Group-level connectivity differences do not show that a clinician can inspect one scan and reliably select the best treatment for that person. Imaging can support discovery while clinical utility remains an open question. Replication, measurement reliability and outcomes following test-guided decisions are separate steps. [5]
Speed does not remove the need for protocol-specific safety
Expert TMS safety guidance addresses screening, devices and implants, seizure risk, hearing protection, operator competence and other practical safeguards. An accelerated schedule should be evaluated as the protocol actually delivered, including intensity, cumulative exposure and session spacing. Evidence from a familiar device name should not be assumed to cover every locally modified schedule. [6]
Assessment also needs to consider the depressive presentation, previous mood elevation, concurrent treatment and the ability to report adverse effects. For a person considering treatment, ask how the service documents symptoms, tolerability and follow-up, and what happens if benefit is incomplete. A faster schedule changes logistics; the clinical responsibility to evaluate the whole course remains. [6]
The next endpoint is a life beyond the initial course
One-month remission and sustained recovery are different outcomes. The controlled SNT reports establish a foundation for further work on durability; pragmatic studies highlight the possible need for additional sessions. The next phase should clarify retreatment, continuation strategies, patient-important functioning and the resources required to deliver the approach consistently. [4]
For now, the most useful description is exact: which protocol, in which population, compared with what, and measured when? That language makes room for progress without converting preliminary biomarker findings or selected short-term outcomes into a promise. Accelerated TMS deserves attention because the evidence is developing, and it deserves careful reading for the same reason. [5][6]
Sources & study types
- Blumberger et al. THREE-D: theta-burst vs high-frequency rTMS. The Lancet (2018).Randomized noninferiority trial · Accessed 27 October 2026
- Cole et al. Stanford neuromodulation therapy: double-blind randomized trial. American Journal of Psychiatry (2022).Sham-controlled randomized trial · Accessed 27 October 2026
- Kratter et al. SNT replication and candidate EEG biomarkers. World Psychiatry (2026).Sham-controlled randomized trial and EEG study · Accessed 27 October 2026
- Accelerated theta-burst stimulation for treatment-resistant depression. JAMA Psychiatry (2025).Pragmatic sham-controlled randomized trial · Accessed 27 October 2026
- Gajawelli et al. Target–default mode network connectivity following SNT. Translational Psychiatry (2024).Secondary randomized-trial imaging analysis · Accessed 27 October 2026
- Rossi et al. TMS safety and recommendations: expert guidelines. Clinical Neurophysiology (2021).Expert safety guidance · Accessed 27 October 2026
About this review
- Research checked
- 27 October 2026
- Clinical review
- Not independently reviewed
- Method
- Focused narrative review of selected primary studies and official guidance; not a systematic review.
International evidence is discussed in its source context. Indian authorization, availability and individual clinical decisions require separate assessment.