Why outcomes measurement matters now
Interventional psychiatry is at an inflection point. TMS and esketamine have moved from experimental to mainstream, but mainstream adoption brings mainstream scrutiny. Payers are asking harder questions. Prior authorization requirements are tightening. Patients are better informed and asking for evidence of expected outcomes before committing to multi-week treatment courses.
Simultaneously, value-based care frameworks are expanding their reach into specialty mental health. The implicit assumption, that offering TMS or Spravato is sufficient evidence of quality care, is being replaced by a requirement to demonstrate measurable patient outcomes across the treatment episode and beyond.
Clinics that can demonstrate objective, continuous outcome measurement will have a compounding advantage: better payer relationships, stronger prior authorization approval rates, more compelling patient onboarding conversations, and the data infrastructure to participate in outcomes-based contracts as they emerge.
50%
TRD relapse rate within 6–12 months of response
60%
Of relapses undetected by the clinic
4×
Net ROI from passive monitoring in IP clinics
Current gaps in IP outcome tracking
Most interventional psychiatry clinics track outcomes at three points: baseline, end of acute treatment, and intermittently at follow-up visits. That structure misses the most clinically and financially consequential window: what happens between visits.
The between-visit blind spot
60% of relapses go undetectedStandard outcome tracking captures pre- and post-treatment snapshots. It does not capture the trajectory between visits, which is where relapse begins, where patient engagement deteriorates, and where the revenue leakage starts.
Dependence on patient self-report
30–40% PHQ-9 email completion ratePHQ-9, MADRS, and similar scales depend on patient recall, motivation, and self-awareness. A depressed patient who most needs tracking is least likely to accurately report their state, and most likely to underreport worsening symptoms to avoid confronting them.
Sparse data density
48h early warning window missedEven a well-run outcomes program produces 8–12 data points per patient per year. That resolution is insufficient to detect the early, subtle signals that precede a clinically significant relapse, which can unfold over days, not weeks.
No longitudinal response tracking
Lost protocol optimization opportunityMost clinics know whether a patient responded to acute treatment. Few have systematic data on response durability at 3, 6, and 12 months, making it impossible to optimize maintenance protocols or identify which patient profiles have the most durable responses.
What to measure, and when
Outcome measurement in IP clinics serves three distinct purposes, each requiring different data at different timepoints:
| Purpose | Timepoint | Measure | Tool |
|---|---|---|---|
| Treatment response | Baseline → end of acute | Depression severity change | PHQ-9 + passive index |
| Relapse detection | Continuous (daily) | Depression index trend | Passive monitoring |
| Response durability | 3, 6, 12 months | Sustained remission rate | Passive index + PHQ-9 |
| Protocol optimization | Across patient cohort | Response by profile/protocol | Aggregated passive data |
| Payer documentation | Pre-auth + ongoing | Objective severity + change | Passive index reports |
Objective vs. subjective measurement
The distinction between objective and subjective outcome measures is becoming clinically and commercially significant. Payers are increasingly skeptical of outcomes programs built entirely on self-reported data, with good reason. Self-report is subject to recall bias, social desirability effects, and the fundamental problem that the most symptomatic patients are the least reliable reporters of their own state.
Objective measurement, derived from physiological signals rather than patient self-report, provides a different quality of evidence. Emobot’s passive monitoring achieves r=0.89 correlation with MADRS (clinician-administered) and r=0.83 with PHQ-9 (self-report). The higher correlation with MADRS than PHQ-9 is telling: the objective signal is closer to what a trained clinician would assess than to what the patient reports about themselves.
Subjective measures
- PHQ-9 (patient self-report)
- GAD-7
- Patient-reported outcomes
- Symptom diaries
30–40% completion · Recall bias · Missing worst patients
Objective measures
- Passive depression index (r=0.89 MADRS)
- Facial expression biomarkers
- Vocal biomarkers
- Actigraphy + digital behavior
100% coverage · No patient effort · Continuous
The business case for better outcomes
Outcome measurement is a clinical imperative, but it is also a business one. The financial model of an IP clinic depends on patient retention through acute treatment and into maintenance. Every patient who relapses undetected and drops out represents lost revenue of up to $19,000. The 60% of relapses that go undetected in standard-of-care programs represent a recoverable revenue gap.
Clinics using Emobot’s passive monitoring have documented a 4× net ROI, primarily from three revenue mechanisms:
Early relapse detection → retained patients
Up to $19k per patient retained48-hour early warning allows proactive outreach before dropout. A patient who might have silently deteriorated and stopped coming is instead contacted, supported, and kept in the program.
Objective data → stronger prior authorization
Fewer denials, faster approvalsPayers are more likely to approve maintenance TMS and extended Spravato programs when supported by objective, continuous outcome data, rather than periodic PHQ-9 scores alone.
Outcomes documentation → payer contract leverage
Structural revenue advantageClinics with documented outcome data are better positioned for value-based contract negotiations and preferred network status, which affects patient volume over the long term.
A practical outcomes framework for IP clinics
Based on clinical experience across 150+ physicians in US, France, Germany, and Canada, here is the outcome measurement framework that works operationally for an IP clinic without creating additional staff burden:
Pre-treatment (baseline)
Once- PHQ-9 or MADRS at intake
- Enroll patient in passive monitoring (EmoDTx, 3 min setup)
- Record baseline passive index score
Acute treatment
Daily monitoring, pre-session review- Passive index updates daily, no staff action required
- Dashboard review before each session (2 min)
- Flag any patients with downward trend for clinical review
End of acute treatment
Once at treatment end- PHQ-9 or MADRS at treatment completion
- Compare to baseline + passive index trend
- Document response: remission / response / partial / non-response
Maintenance monitoring
Continuous + quarterly assessment- Continue passive monitoring throughout maintenance phase
- Alert-based outreach (no proactive staff review needed)
- Quarterly PHQ-9 at scheduled visits
Long-term follow-up
12 months post-treatment- 12-month passive monitoring for all treatment completers
- Flag response degradation for maintenance protocol review
- Use cohort data for protocol optimization
Built for IP clinics
See the outcomes dashboard for your practice
30-minute demo. We’ll walk through a full patient trajectory and show you exactly what the outcomes data looks like for TMS and Spravato patients.