Patient Selection & Building a Program

Evidence status. Regulatory route depends on indication, device, and whether use is clinical or research. HDE, off-label clinical use, and significant-risk device research are not interchangeable; confirm current FDA and institutional requirements.

Refractory OCD

Surgical candidacy in OCD requires severe, chronic, and genuinely refractory illness with substantial functional impairment, typically a Yale-Brown Obsessive Compulsive Scale (YBOCS) score at the upper end of the severe band or into the extreme band. On the standard bands, 24 to 31 is severe and 32 to 40 is extreme; surgical programs commonly set the cutoff at ≥ 28 of 40, though thresholds vary. A documented illness duration is also required, conventionally at least five years, although some protocols accept three. Adequacy of prior treatment is the crux, and it must be real rather than nominal:

• Adequate trials of at least three selective serotonin reuptake inhibitors (SSRIs), with dose, duration, adherence, response, and intolerance documented. Many OCD protocols require at least 12 weeks total, including a sustained period at the maximum tolerated therapeutic dose; exact criteria depend on the protocol and patient. At least three is also the threshold written into the FDA humanitarian device exemption labeling, so a program operating under that exemption must document it.
• An adequate trial of clomipramine.
• Augmentation with an antipsychotic.
• An adequate course of cognitive behavioral therapy with exposure and response prevention (ERP), conventionally at least 20 sessions, or roughly 25 to 40 hours, delivered by a therapist with specific OCD expertise; some programs require two failed courses with different therapists. Session count, hours, therapist credentials and homework adherence should be documented rather than assumed, because this is the element most often nominal rather than real.

Pallanti and Quercioli emphasized that “treatment resistance” is frequently mislabeled when trials were underdosed, too brief, or lacked adequate ERP, and proposed a staging framework precisely to standardize the judgment. Confirming the adequacy and failure of ERP, in particular, separates a truly refractory candidate from an undertreated one.

Refractory depression

For treatment-resistant depression (TRD), DBS candidacy is determined by a research protocol rather than a universal clinical threshold. Protocols typically require a severe, sustained episode despite multiple adequate medication and augmentation trials, evidence-based psychotherapy, and ECT failure, intolerance, or a justified contraindication. Document the appropriateness and availability of rTMS, ketamine/esketamine, and other established treatments rather than claiming that every protocol mandates all of them. Confirm the diagnosis, including assessment for bipolarity, psychosis, substance use, and reversible medical contributors. The TRANSCEND trial, for example, requires inadequate response to at least four antidepressant treatments. Deep refractoriness alone does not establish that DBS will help.

Who decides, and how

The 2014 consensus requires that candidacy be determined by an independent, multidisciplinary committee rather than by the surgeon or any single clinician. A functioning committee typically includes psychiatry (ideally with disorder-specific expertise, e.g., an OCD specialist), functional neurosurgery, neurology, neuropsychology, and access to bioethics; some programs include an independent psychiatrist not otherwise involved in the patient's care. The committee verifies the diagnosis and refractoriness, weighs comorbidity and psychosocial supports, confirms the patient's capacity, and documents its reasoning. Crucially, the surgical team should not be the sole gatekeeper of who is operated on.

Capacity, consent, and vulnerability

Informed consent in this population demands particular care. The disorders themselves (severe depression with hopelessness, OCD with pathological doubt) can affect understanding, voluntariness, and risk appraisal, and desperation after years of suffering can blur the line between informed choice and coercion by circumstance. Best practice includes a formal capacity assessment, explicit discussion of the investigational or humanitarian-device status of the procedure, realistic framing of expected benefit and the real possibility of nonresponse, the burden of device management or the permanence of a lesion, and the involvement of the treating psychiatrist and, with the patient's permission, family or other chosen supports. Consent is a process, not a signature.

The instruments that define candidacy and response

Validated, disorder-specific rating scales are used both to establish baseline severity (candidacy) and to define response, and they should be administered prospectively, ideally by a rater independent of the surgical team. Response conventions are standardized: a ≥ 35% reduction in YBOCS defines OCD response; a ≥ 50% reduction in HAM-D or MADRS defines depression response, with remission defined by absolute thresholds (e.g., MADRS ≤ 10 or HAM-D-17 ≤ 7; exact cutoffs vary by protocol). Quality-of-life and functional measures, plus baseline and longitudinal neuropsychological assessment, complete the picture and help detect cognitive or personality change.

The framework around the operation

The consensus situates each case within a governance framework. In the United States, the Medtronic OCD HDE is device- and indication-specific; another manufacturer's system is not covered by that HDE. FDA requires use at a facility with IRB oversight, and clinical HUD use must be approved by an IRB or an appropriate local committee, except for defined emergencies. Approval for clinical use is distinct from approval of a clinical investigation. Research collecting safety or effectiveness data requires research review, consent, and applicable FDA protections; significant-risk use outside the approved indication generally requires an IDE. A registry is not automatically an IDE-requiring study, and institutional review should determine whether a given activity is research, quality improvement, exempt, or otherwise regulated. Other psychiatric DBS indications remain experimental; off-label clinical care and device research are distinct pathways that require explicit institutional decisions. Prospective outcome tracking and long-term follow-up remain essential.

Who is required, and why

Psychiatric neuromodulation is irreducibly multidisciplinary; no single specialty can run it. The core roster:

• Functional neurosurgery: stereotactic implantation/lesioning and surgical decision-making.
• Psychiatry with disorder-specific expertise: ideally an OCD/mood specialist who owns diagnosis, refractoriness adjudication, and longitudinal care; this is the clinical center of gravity, not a consultant role.
• Neurology: particularly for programming expertise borrowed from movement-disorder DBS.
• Neuropsychology: baseline and longitudinal cognitive/behavioral assessment.
• Bioethics: engaged in candidacy and consent, not only at crises.
• A program coordinator/nurse navigator and advanced-practice support: the operational glue managing the long, multi-step pathway.
• Where relevant, radiation oncology / Gamma Knife and focused-ultrasound expertise for incisionless lesion options.

These individuals constitute the multidisciplinary committee that determines candidacy; the surgeon is a member, not the gatekeeper.

Getting the right patients, worked up the right way

The pipeline determines program quality. Referrals come chiefly from psychiatry and specialized therapy programs (e.g., intensive ERP/CBT services), and the program's job is to add a rigorous filter rather than a fast track. Practical features of a functioning pipeline:

• A clear refractoriness gate: before surgical evaluation, verify that prior pharmacotherapy and, critically, psychotherapy were genuinely adequate, since inadequate ERP for OCD or under-tried ECT for depression is the commonest reason an apparent candidate is not one.
• Expectation-setting from first contact: honest framing of investigational or humanitarian-device status, the real possibility of nonresponse, the long latency to benefit, and the lifelong device commitment.
• Tolerance for a slow, low-volume pipeline: appropriate candidates are uncommon, and a program that values throughput over selection has misunderstood the field.

Matching the framework to the indication

Each indication's regulatory route follows its evidentiary status, and the program must operate the correct one for each:

• OCD DBS proceeds clinically under the FDA Humanitarian Device Exemption for bilateral AIC/ALIC stimulation, which still requires institutional review board or appropriate local committee oversight at the implanting center. It is the one FDA-recognized psychiatric DBS indication offerable outside a research trial, and it covers the specific HDE-approved device and indication only.
• All other psychiatric DBS indications discussed here (depression, Tourette, addiction, anorexia) are investigational or experimental. Significant-risk device research generally requires an Investigational Device Exemption and IRB-approved protocols; any off-label clinical pathway should be explicitly reviewed through institutional governance and tracked prospectively.
• Informed consent is a documented, capacity-based process appropriate to a vulnerable population, with the treating psychiatrist and patient-authorized supports engaged.
• Conflict-of-interest management for any device-industry relationships is explicit and disclosed.
• Access and reimbursement are a distinct hurdle from regulatory clearance: the OCD HDE permits the procedure but does not guarantee payer coverage, and real-world OCD DBS uptake has been low and even declined after approval, for reasons that likely include coverage, reimbursement, referral, and access barriers as well as residual uncertainty about benefit and burden. Build coverage navigation into the pathway.

This architecture operationalizes the 2014 multi-society consensus; a program that cannot articulate the regulatory basis for each indication it offers is not ready to offer it.

The part that lasts for years

Unlike a one-time operation, psychiatric neuromodulation creates a long-term care relationship. The program needs a dedicated programming pathway staffed by clinicians comfortable titrating against a slow mood-and-anxiety response over months, embedded psychiatric co-management that continues regardless of device response, and explicit crisis and suicidality protocols given the persistent risk in these populations. Continued evidence-based psychotherapy (ERP for OCD) after implantation is part of optimizing outcome, not an afterthought. The programming logic itself is developed on the core indications page.

Measuring, sharing, and growing

Because individual centers accrue small numbers, prospective data collection and registry participation (such as the International Tourette Syndrome DBS Registry, and institutional or multicenter OCD outcome databases) are how the field learns and how a program demonstrates quality. A research-integrated program also positions itself for the field's frontier: connectomic targeting, biomarker-driven and adaptive stimulation, and head-to-head target trials. A pragmatic staged build: