1. The two monitoring gaps in TMS care
Transcranial Magnetic Stimulation (TMS) produces meaningful clinical responses in treatment-resistant depression (TRD). Roughly 50–60% of patients who complete an acute course achieve clinical response; 30–35% achieve full remission.
The problem is that the word “complete” is doing a lot of work in that sentence, and so is the word “acute.” TMS programs have two distinct monitoring gaps, both documented in the peer-reviewed registry data.
Gap #1: Acute-course completion
- 24%of patients never finish their acute TMS course (Sackeim et al. 2020, NeuroStar Registry)
- 17.6%stop before reaching 30 sessions, even though the protocol calls for 36 (Sackeim et al. 2023)
- 82%higher remission rate for completers vs. dropouts; the gap comes from treatment exposure, not patient type (Sackeim 2023)
- 20%increase in completion when clinics add structured touchpoints mid-course (Neuronetics TrakStar data)
The most-cited reason for non-completion is not side-effect burden. 37% of VA dropouts cited “personal reasons” (Lapid-Brown et al. 2025). In practice: logistics, commute, motivation decay, unclear benefit. All of them are visible in continuous behavioral data.
Gap #2: Post-acute response durability
- 37.5%of acute-course responders lose response within 12 months (Dunner et al. 2014, 42 US sites, 257 patients)
- 36.2%of those losing response come back for a retreatment course; the rest don't (Dunner 2014)
- 84%of patients who are retreated regain response (Janicak et al. 2010)
- ~5 momedian time to relapse after a successful acute course (d'Andrea et al. 2023)
The maintenance gap is where the real leak is, and where the data is most actionable. 84% regain-of-response on retreatment means: if you can detect early decay and bring the patient back in, you almost always save them. The problem is detecting the decay at all.
In both gaps, the standard monitoring tool is a PHQ-9 sent by email, with 30–40% completion. Between PHQ-9s there is no continuous, objective signal. The first sign most clinicians see of a relapse is a missed appointment.
This isn’t a failure of clinical care; it’s a structural problem. Psychiatrists are trained to observe patients in-person. The tools for monitoring between visits haven’t kept pace with what technology can now deliver.
2. Why this costs your clinic
Every undetected relapse represents both a patient outcome failure and a revenue loss. When a patient drops out of your maintenance program following a relapse they managed alone, the typical lifetime value impact is $8,000–$19,000 per patient per year.
Across a 50-patient active TMS program:
Relapses per year (50% rate)
25 patients
Undetected (60%)
15 patients
Revenue lost
$135k–$285k
With Emobot: caught (70%)
18 relapses
Recoverable revenue
$94k–$200k
Net ROI on monitoring
4×
3. TMS monitoring options compared
Here’s an honest comparison of the monitoring options available to TMS clinics:
| Approach | Frequency | Patient burden | Completion | Accuracy |
|---|---|---|---|---|
| PHQ-9 by email | Every 6–8 weeks | Manual entry | 30–40% | Subjective |
| Phone check-in | Weekly/monthly | Active call | Variable | Clinician-dependent |
| Wearables (Fitbit etc.) | Continuous | Daily wear | Poor long-term | Single-modal |
| Voice apps (active) | When prompted | Active recording | Poor | Medium |
| Emobot (passive AI) | Continuous | Zero (passive) | 80%* | r=0.89 MADRS |
*When proposed by a clinician and installed with the patient during the visit
4. How passive AI monitoring works
Passive AI psychiatric monitoring uses smartphone sensors to capture behavioral and physiological signals, without requiring any active participation from the patient. The key word is passive: the system works in the background, during normal phone usage, without prompting the patient to do anything.
For TMS monitoring specifically, four signal streams are most clinically relevant:
Facial microexpressions
Captured during natural phone use (unlocking, scrolling)
Facial affect changes precede subjective mood reports by 24–48 hours
Vocal biomarkers
Prosody, rhythm, and energy during phone calls
Voice flattening is a strong early indicator of depressive relapse
Actigraphy
Accelerometer captures movement patterns, step count
Activity reduction is one of the earliest behavioral markers of relapse
Digital behavior
App usage patterns, screen time, communication frequency
Behavioral changes in phone use correlate with mood state shifts
5. Clinical evidence for passive monitoring
Emobot’s passive monitoring platform has been validated in 10+ peer-reviewed clinical studies across France, the United States, Canada, and Germany.
r=0.89
vs. MADRS (gold standard)
r=0.83
vs. PHQ-9 (self-report)
48h
Early relapse warning
The r=0.89 MADRS correlation is clinically significant. For context: the MADRS has a test-retest reliability of approximately r=0.80–0.90 when administered by different clinicians. Emobot’s AI achieves comparable agreement, continuously, automatically, at zero cost to the patient.
6. Implementation guide for TMS clinics
Most TMS clinics are fully operational with Emobot within one business day. Here’s the implementation pathway:
Clinic setup
Create your Emobot account at portal.emobothealth.com. Add clinical team members. No EHR integration needed.
Staff training
30-minute walkthrough with our team. Review dashboard, alert settings, patient enrollment flow. Training video available async.
Enroll first cohort
Start with your 5–10 highest-risk maintenance patients. Send enrollment links directly from the dashboard.
First dashboard review
Review patient timelines with your team. Adjust alert thresholds to match your clinical workflow.
Full rollout
Expand to all active TMS maintenance patients. Introduce in post-acute course conversations as standard of care.
7. ROI model for TMS monitoring
The ROI case for passive TMS monitoring is straightforward: the cost of monitoring is a fraction of the revenue recovered from catching relapses that would otherwise lead to patient dropout.
Clinics using Emobot report an average 4× net ROI, net of the subscription. For every $1 spent on monitoring, about $5 is recovered in retained patient revenue, leaving ≈$4 of net gain (the net multiple reported).
Run your own ROI calculation →TMS monitoring FAQ
What is TMS progress monitoring?
TMS progress monitoring means tracking a patient's response to transcranial magnetic stimulation between sessions — not just at intake and discharge. It pairs standardized scales (PHQ-9, MADRS) with objective, continuous signals so you can see whether a patient is responding, plateauing, or relapsing close to real time.
How do you monitor TMS therapy response between sessions?
Two ways: active and passive. Active monitoring relies on periodic questionnaires like the PHQ-9, which reach only 30–40% of patients and capture a single moment. Passive monitoring uses smartphone sensors — facial expression, vocal biomarkers, activity, and digital behavior — to measure response continuously with no patient effort, flagging deterioration roughly 48 hours before symptoms surface.
What makes TMS monitoring effective?
Effective TMS monitoring covers both the acute and maintenance phases, works regardless of patient engagement, produces objective data rather than recall, and provides enough lead time to act before a patient drops off. Passive, multimodal monitoring meets all four criteria; PHQ-9-by-email meets none of them reliably.
References
Every acute-course and maintenance-gap figure on this page is sourced from peer-reviewed registry data or published clinical trials. Citations are provided in full for clinicians, payers, and regulatory reviewers.
- Sackeim HA, et al. Clinical outcomes in a large registry of patients with major depressive disorder treated with Transcranial Magnetic Stimulation.J Affect Disord. 2020. NeuroStar Registry, N = 5,010.
- Sackeim HA, et al. The impact of pre-treatment and demographic factors on the outcome of transcranial magnetic stimulation for major depressive disorder. J Psychiatr Res. 2023.
- Dunner DL, et al. A multisite, naturalistic, observational study of transcranial magnetic stimulation for patients with pharmacoresistant major depressive disorder: durability of benefit over a 1-year follow-up period. J Clin Psychiatry. 2014; 75(12):1394–1401. (42 US sites, 257 patients, 37.5% lost response within 12 months, 36.2% retreated.)
- Janicak PG, et al. Durability of clinical benefit with transcranial magnetic stimulation (TMS) in the treatment of pharmacoresistant major depression. Brain Stimul. 2010; 3(4):187–199. (84% regain-of-response on retreatment.)
- d’Andrea G, et al. Long-term outcomes after Transcranial Magnetic Stimulation (TMS) for treatment-resistant depression. 2023. (Median time to relapse ≈5 months post-acute.)
- Senova S, et al. Durability of antidepressant response to repetitive transcranial magnetic stimulation: Systematic review and meta-analysis. Brain Stimul. 2019; 12(1):119–128. (52.9% lose response within 6 months.)
- Lapid-Brown SL, et al. VA transcranial magnetic stimulation program outcomes.2025. (37% of dropouts cited “personal reasons” rather than side-effects.)
- Neuronetics TrakStar® clinic data. (20% improvement in acute-course completion with structured mid-course touchpoints.)
Emobot’s own validation (MADRS r=0.89, PHQ-9 r=0.83, 10+ peer-reviewed studies) is available on the clinical evidence page.
8. Getting started
Two paths to get started with TMS patient monitoring:
Book a Demo
30 minutes. We’ll show you the dashboard live with a real patient case and answer every question.
Book a Demo →Activate your account
Self-serve setup in minutes. Invite your first patients the same day — no procurement, no waiting.
Activate account →