Ketamine for Depression — Body Handbook

Psychology · §464

Ketamine for Depression

The first class of antidepressant that works in hours instead of weeks. A single sub-anaesthetic dose of ketamine drops depression severity by roughly half within 24 hours for the patients who respond — about half of those who try it — and the effect can outlast the dose by days. Two routes are in clinical use: the FDA-approved nasal spray esketamine (Spravato), administered in a certified clinic with two hours of post-dose monitoring; and off-label IV racemic ketamine at sub-anaesthetic dose, the version with the larger effect size and no insurance coverage.

Decide · Course Evidence Moderate Chapter Psychology

For people whose depression has survived two or more antidepressants — roughly a third of people diagnosed with major depression — this is a treatment that can lift the mood in hours rather than weeks, on the level of an effective psychiatric intervention. Response rate is around half. The catch: it isn't a one-shot cure (the effect fades without continued dosing), it isn't cheap (a full course runs into the thousands), and it isn't a take-home medication — you sit in a monitored chair for two hours per session, twice a week for a month.

Conventional antidepressants raise serotonin or norepinephrine and take four to six weeks to do anything noticeable. Ketamine works on a different system entirely — glutamate, the brain's main excitatory signal — and the effect shows up within hours. At sub-anaesthetic doses, ketamine briefly blocks a specific subset of NMDA receptors on inhibitory neurons in the prefrontal cortex, which releases the brake on glutamate signalling. The resulting surge triggers BDNF release, activates the mTOR pathway, and drives new dendritic spine formation within 24 hours Li et al. 2010. The action isn't on serotonin; it's on synaptic plasticity itself. That's why it's fast and why a single dose can outlast its short half-life by a week — the synapses it grew are still there.

What the trials actually showed

The proof-of-concept was a seven-patient crossover trial in 2000: a single IV infusion of sub-anaesthetic ketamine produced rapid mood improvement that saline placebo didn't Berman 2000. The replication that put ketamine on the map came six years later from the NIMH.

The obvious criticism: ketamine's dissociative effects are unmistakable, so a saline-controlled trial isn't really blind. The most rigorous reply came from a 73-patient trial that used midazolam, a sedative with its own psychoactive footprint, as the active comparator — and ketamine still won, with 64% response at 24 hours versus 28% on midazolam Murrough 2013. For the intranasal version, Janssen's pivotal phase 3 trial showed a 22.5% remission rate at four weeks on esketamine plus an oral antidepressant, against 7.6% on placebo spray plus the same antidepressant Popova et al. 2019. Two of the three pivotal short-term initiation trials missed their primary endpoint, and the FDA accepted a withdrawal trial as the second positive — a regulatory call that drew published commentary in the same journal that published the positive result. The largest head-to-head trial then pitted IV ketamine against electroconvulsive therapy in 403 patients with treatment-resistant depression: ketamine came out non-inferior, with a substantially gentler cognitive side-effect profile than ECT Anand et al. 2023. The American Psychiatric Association's consensus statement summarised the field as having strong support for use in major depressive episodes, with long-term safety and durability still under-characterised Sanacora et al. 2017.

If you have treatment-resistant depression and don't try this

Roughly a third of people diagnosed with major depression don't recover on two or more antidepressants — the operational definition of treatment-resistant depression. For that group, the next decade carries two to four times the all-cause mortality of non-depressed peers, mostly driven by suicide and heart disease. The lived shape is harder to look at than the statistic. The friend who keeps saying they're tired stops accepting invitations. The work that used to feel meaningful turns into something they push through. The version of them that other people knew five years ago becomes the version their partner is grieving while they're still in the room. Conventional antidepressants take four to six weeks to do anything; for someone in an acute suicidal crisis, that timeline is itself the problem. Esketamine cut depressive symptoms within 24 hours in patients hospitalised with active suicidal intent — a speed no oral antidepressant can match Fu et al. 2020. The decision not to try ketamine isn't necessarily wrong, but it should be made with the time-to-response gap visible.

How a course actually runs

Two routes, similar shape. Both start with a dense induction phase and taper to maintenance. Neither is a take-home medication.

The two-hour observation isn't a formality. Blood pressure rises in roughly a third of sessions, sometimes substantially, and the dissociation — the room feels distant, time slows, the body feels far away — can be intense. Both resolve within about two hours; you wait until they do before going home. Plan the day around the session: no work, no errands, no childcare immediately after.

When this isn't appropriate

Concurrent benzodiazepines (alprazolam, clonazepam, diazepam) appear to blunt the antidepressant response, and clinics often ask patients to taper if it's clinically safe to. Heavy chronic recreational ketamine use causes severe bladder damage — interstitial cystitis, sometimes irreversible — in roughly a quarter of regular users. The same complication at therapeutic dosing is uncommon but has been reported with extended maintenance, so any new urinary symptoms (frequency, urgency, blood in the urine) get flagged early rather than worked through.

What it costs and how to get to a clinic

The cost gap between the two routes is enormous, and it tracks regulatory status rather than clinical evidence. Spravato is FDA-approved, on most commercial formularies, and covered by Medicare Part B. After a prior authorisation (typically requiring documented failure of two or more oral antidepressants), patient copays land at $10–$50 per session on commercial insurance, or roughly $140–$240 per session on Medicare's 20% coinsurance before any supplement. Cash price including monitoring runs $590–$1,600 per session. IV racemic ketamine for depression is off-label, almost never covered, and runs $400–$800 per session out of pocket in most US markets — a six-session induction at $2,400–$6,400 plus ongoing maintenance. HSA and FSA funds typically qualify when a clinician documents the indication. Both routes require travel to a clinic; Spravato's REMS restriction rules out home administration entirely, and reputable IV clinics also require in-person dosing for the monitoring. Plan for the day to be a write-off; no driving for 24 hours after any dose.

Who this is actually for

The evidence base is centred on adults aged 18–65 with major depressive disorder who have failed two or more adequate antidepressant trials of different classes. Roughly 30% of diagnosed major depression falls into that group. The 2020 expansion of the Spravato approval added a separate indication for major depressive disorder with active suicidal ideation; that pathway exists specifically because oral antidepressants take weeks to work, and for someone in immediate crisis weeks is too long Fu et al. 2020 Ionescu et al. 2021. For someone on a first-line antidepressant who is tolerating it and improving, this isn't the next move. For someone in their fifth medication trial with no response, it is the conversation to have with a psychiatrist. Evidence in over-65 patients is weaker — the one phase 3 trial in that age group missed its primary endpoint — and there is no approved indication in adolescents.

What gets oversold

"One infusion fixed me." The acute effect is real and dramatic. Durability without continued dosing is not. Across the trial evidence, a single dose drifts back toward baseline within one to two weeks.
"Esketamine is just better ketamine." Esketamine is the S-enantiomer of racemic ketamine and binds the NMDA receptor more tightly, but in head-to-head meta-analyses IV racemic ketamine produces a larger effect on depression than intranasal esketamine. Esketamine's advantage is regulatory and logistical — FDA-approved, insurance-coverable, a nasal spray — not pharmacological superiority.
"The trippy part is the point." Some clinics frame the dissociative experience as a psychedelic journey that drives the therapeutic effect. Controlled data don't support that — the intensity of dissociation during the dose doesn't predict the depression response that follows.
"Ketamine clinics are uniformly evidence-based." They aren't. Demand has outrun clinical guidelines; some clinics use unvalidated protocols (oral troches mailed for home use, doses outside the trial range), and the field's own consensus statement flagged the gap explicitly Sanacora et al. 2017.

Where this goes wrong in practice

The most common failure is stopping after the induction series. The six-infusion or four-week induction works, the patient feels noticeably better, they stop coming in, and the depression returns within weeks. Maintenance isn't optional for responders — it's the part that holds the response, and the randomised withdrawal data are unambiguous on the point Daly et al. 2019. The second failure is ordering oral or sublingual ketamine from a telehealth provider for unmonitored home use. The effect size in adequately blinded trials of oral routes is small, the monitoring isn't there, and the abuse-potential supervision is gone. The third is concurrent benzodiazepine use the patient didn't think to mention — Xanax or Klonopin taken for anxiety can blunt the antidepressant response. The fourth is choosing IV when cash isn't there for the long haul, then stopping early as the bills mount, instead of going through the prior-authorisation paperwork that would make the insurance-covered nasal-spray version affordable.

Other paths from the same starting point

For someone in the same evidentiary position — two or more antidepressants failed, looking at next steps — the non-ketamine options are: augmenting the existing antidepressant with an atypical antipsychotic (aripiprazole, quetiapine, brexpiprazole) or lithium; switching to an MAOI; electroconvulsive therapy (the longest-running rapid-acting treatment, with substantial cognitive side effects, and the only intervention that has gone head-to-head with ketamine at scale — non-inferior, not superior Anand et al. 2023); repetitive transcranial magnetic stimulation (FDA-approved, no cognitive side effects, daily sessions for four to six weeks, modest effect size); or dextromethorphan-bupropion (Auvelity), an oral NMDA-modulating drug approved in 2022. The choice is rarely between ketamine and nothing; it is between ketamine and one of those.

What changes when it works

For the responder — roughly half of patients who try it — the change is fast and unmistakable. The day after the first infusion, the morning doesn't feel like a wall. The thought that has been on a loop for months, that nothing will ever feel different, loses its grip. Not gone, but no longer load-bearing. Within the first week the people who knew you before recognise the version of you they hadn't seen in years; the partner stops asking what's wrong because the answer to the question has changed. Sleep improves as the depression does, the morning tiredness gives ground, the capacity to read a page or follow a conversation comes back.

The shape over months depends entirely on maintenance. Without continued dosing, the effect of a single course fades over one to two weeks — this is not one-and-done. The randomised withdrawal trial of intranasal esketamine showed that responders kept on maintenance had a 51% lower relapse risk than those switched to placebo spray over the trial's follow-up; for the remitter subgroup, the gap was around 70% Daly et al. 2019. The people who hold the response are the people who keep dosing — every two to four weeks for IV, every one to two weeks tapering to biweekly for esketamine. Dosing holds the chemistry; it doesn't do the durable, non-drug part for you. The inner work of rebuilding the habits, relationships, and sense of meaning the depression hollowed out is something you pair with the maintenance, not something it replaces. For the half who don't respond, the first two or three infusions usually settle the question; if the depression hasn't moved by then, it's unlikely to move with more.

Psilocybin and MDMA-assisted therapies are a different mechanism, different regulatory status, and a different evidence base — ketamine sometimes gets grouped with them but doesn't share the structure. Ketamine is also used off-label for chronic pain syndromes, PTSD, OCD, and eating disorders, each with its own thinner literature. Depression itself is a category, not one thing: bipolar depression, postpartum depression, and seasonal patterns each have specific first-line interventions that aren't ketamine. Electroconvulsive therapy and transcranial magnetic stimulation are the device-based neighbours for treatment-resistant depression — each worth its own look.

Related in the handbook

Alternatives

— At the other end of the spectrum from exercise — fast and clinic-bound where exercise is slow and free.
— Inner work is the durable, non-drug side of treating low mood — often paired with, not replaced by, a fast intervention like ketamine.
— Ketamine works in hours where TMS takes weeks of daily visits — both are options once standard antidepressants fail.
— If the appeal is an altered state without a drug, holotropic breathwork is the breathing version — far thinner evidence and a long do-not-do list.
— Ketamine's effect fades without maintenance; pairing it with ongoing therapy is how people make the lift last.

— Ketamine's fast action makes it relevant when depression has turned dangerous and there's no time to wait on an SSRI.
— Unlike microdosing, ketamine actually beats placebo for depression — a useful contrast in what 'psychedelic medicine' has proven.

Substance + claimed effects

Ketamine is a non-competitive NMDA-receptor antagonist developed in 1962 as a dissociative anaesthetic and used routinely in emergency medicine and pediatric procedural sedation. At sub-anaesthetic doses — roughly one-tenth of an anaesthetic dose — racemic ketamine and its S-enantiomer esketamine produce a rapid, large-magnitude reduction in depressive symptoms within hours, a class of effect that conventional monoaminergic antidepressants (SSRIs, SNRIs, TCAs, MAOIs) cannot match on that timescale Berman 2000 Zarate 2006. Two delivery routes dominate clinical use: intranasal esketamine (Spravato, Janssen), FDA-approved for treatment-resistant depression in 2019 and for major depressive disorder with acute suicidal ideation in 2020, with monotherapy approval added in 2025 FDA 2019 FDA 2025; and off-label sub-anaesthetic IV racemic ketamine infused at 0.5 mg/kg over ~40 minutes, delivered in a network of free-standing ketamine clinics that have proliferated since the mid-2010s. Claims this entry covers holistically: rapid antidepressant effect in treatment-resistant depression, acute reduction of suicidal ideation, durability and the relapse problem, the dissociative experience as side effect (and possible mediator), hemodynamic and bladder safety profile, and the cost / access reality.

Evidence by addressing question

mechanism

Ketamine binds the open NMDA-receptor channel as an uncompetitive antagonist. The dominant explanatory model is the disinhibition hypothesis: preferential blockade of NMDA receptors on GABAergic interneurons reduces tonic inhibition of pyramidal neurons in the prefrontal cortex, producing a transient glutamate surge. The surge activates AMPA receptors, triggers BDNF release, and engages the mTORC1 signalling pathway, which drives synaptic protein synthesis, dendritic spine formation, and a measurable increase in synaptic density within ~24 hours of dosing Li et al. 2010. This is fundamentally different from the monoamine-reuptake mechanism of SSRIs — ketamine's effect is on glutamatergic synaptogenesis, not serotonin levels, which explains both the speed (hours, not weeks) and the duration (one infusion of ~0.5 mg/kg IV produces a peak effect at 24 h and a measurable benefit at 7 days in responders Zarate 2006). Esketamine has ~3–4× higher NMDA-receptor binding affinity than the R-enantiomer; whether this translates to clinically meaningfully different efficacy versus racemic ketamine remains debated. An alternate model attributes some antidepressant action to the hydroxynorketamine metabolite acting on AMPA receptors independently of NMDA blockade; the evidence base supports both NMDA-dependent and NMDA-independent contributions, and the two are not mutually exclusive.

evidence

The proof-of-concept came from a 7-patient placebo-controlled crossover trial showing rapid mood improvement after a single sub-anaesthetic IV infusion Berman 2000. The landmark replication was an 18-patient NIMH crossover RCT in treatment-resistant depression: 71% response rate at 24 hours after ketamine versus 0% on saline placebo, with an effect size of 1.46 at 24 hours and 0.68 at 7 days Zarate 2006. Methodological critique focused on inert (saline) placebos being functionally unblinded by ketamine's psychoactive effects. The largest active-controlled IV trial (n=73) used midazolam as a psychoactive comparator and still showed superiority for ketamine at 24 hours (response 64% vs 28%) Murrough 2013. For intranasal esketamine, the pivotal phase 3 short-term TRD trial (TRANSFORM-2, n=223) showed a statistically significant 4-point MADRS advantage over placebo nasal spray at day 28, with 22.5% remission at week 4 versus 7.6% on placebo Popova et al. 2019; two of three pivotal short-term initiation RCTs failed to hit their primary endpoint and the FDA, unusually, accepted a randomised withdrawal trial as a second positive pivotal trial — a regulatory path that drew commentary in the same issue questioning whether the evidence base was sufficient for approval. Recent meta-analyses converge on IV racemic ketamine being more efficacious than intranasal esketamine (IV Hedges' g ≈ 1.5; esketamine Hedges' g ≈ 0.3), with response rates of 45–65% in RCTs for IV ketamine versus ~22–30% for intranasal esketamine over placebo. Real-world response rates are more modest (~30%) for both. The APA-affiliated consensus statement reviewed the field and concluded that the strongest evidence supports use in non-psychotic major depressive episodes; safety, durability, and risk of misuse remained under-characterised Sanacora et al. 2017. In the largest head-to-head trial against electroconvulsive therapy (ELEKT-D, n=403), IV ketamine was non-inferior to ECT in non-psychotic TRD over 3 weeks of treatment, with response rates of 55% (ketamine) versus 41% (ECT) and a more favourable cognitive side-effect profile Anand et al. 2023.

protocol

Intravenous racemic ketamine: standard sub-anaesthetic dose of 0.5 mg/kg infused over 40 minutes, induction schedule typically 6 sessions over 2–3 weeks, then tapered to maintenance every 2–4 weeks based on response. Dose-response meta-analysis suggests efficacy at doses ≥0.2 mg/kg with plateau at 0.5 mg/kg; doses above 0.5 mg/kg do not produce additional treatment response. Intranasal esketamine: induction at 56–84 mg twice weekly for 4 weeks, then weekly for weeks 5–8, then every 1–2 weeks thereafter; suicidal-ideation indication mandates 84 mg from the first dose Popova et al. 2019 Daly et al. 2019. Both routes require monitored administration: minimum 2-hour post-dose observation with vital signs (Spravato is dispensed only at REMS-certified clinics; no take-home). IV ketamine clinics typically run sessions in a reclining chair with continuous BP/SpO₂ monitoring; no driving for 24 hours post-dose.

contraindications

Hard stops: uncontrolled hypertension, recent cardiovascular event (MI, stroke, ruptured aneurysm) within 6 weeks, current psychosis or active mania, history of intracerebral haemorrhage. Relative cautions: pregnancy (no controlled data; animal models show developmental neurotoxicity); active substance use disorder (especially with prior ketamine misuse history); history of severe dissociative experiences or trauma where re-experiencing is contraindicated. Concurrent benzodiazepines may blunt antidepressant response; opioid concomitance increases sedation. Transient blood-pressure elevation during infusion is universal — roughly 20–30% of patients exceed 180/100 mmHg systolic or +20% over baseline, with peak effect within an hour and resolution within ~2 hours Sanacora et al. 2017. Hypertensive emergency in clinic series is extremely rare. Bladder toxicity (ketamine-induced cystitis) is well-documented in heavy chronic recreational use (~25% of regular users); occurrence at therapeutic dosing is rare but case reports exist with extended maintenance — periodic urinary-symptom screening is recommended.

misconceptions

(1) One infusion is a cure. The single-dose effect is dramatic but transient — meta-analyses show return toward baseline within 1–2 weeks without repeat dosing or maintenance. (2) Esketamine = enhanced ketamine. The S-enantiomer is one ingredient of racemic ketamine; whether it is therapeutically superior is not established, and IV racemic ketamine outperforms intranasal esketamine on most head-to-head and meta-analytic measures of effect size. The commercial advantage of esketamine is regulatory (FDA-approved, insurance-coverable) and convenience (nasal spray), not pharmacological superiority. (3) Dissociation = therapeutic response. Some clinics market the dissociative or "psychedelic" experience as the active ingredient; controlled data show acute dissociative intensity is not consistently correlated with antidepressant response, and dissociation in the Spravato trials was a side effect, not a mediator. (4) Ketamine clinics are uniformly evidence-based. The proliferation of off-label IV and oral/troche ketamine providers has outpaced clinical guidelines; some clinics use unvalidated protocols (oral troches, high-dose at-home), and the APA consensus warned that lay demand and clinician supply have run ahead of long-term safety data Sanacora et al. 2017.

audience

The evidence base centres on adults aged 18–65 with major depressive disorder who have failed at least two adequate antidepressant trials of different classes — the operational definition of treatment-resistant depression. Estimated prevalence of TRD: roughly 30% of patients diagnosed with MDD. ASPIRE I/II additionally validated esketamine in MDD with active suicidal ideation with intent Fu et al. 2020 Ionescu et al. 2021. Limited data in older adults (one phase 3 trial, ESKETINTRD3005, in ≥65 failed primary endpoint; FDA approved across the adult age range nonetheless). No FDA approval for adolescents (off-label use exists; one midazolam-controlled adolescent trial is published). Bipolar depression: smaller evidence base, single infusion meta-analyses positive but not an FDA-approved indication. Postpartum depression: anecdotal use; brexanolone and zuranolone are the within-class GABAergic competitors, not ketamine.

stakes

Untreated treatment-resistant depression carries a roughly 2–4× increased all-cause mortality compared with non-depressed peers, dominated by suicide and cardiovascular causes. Functional disability is severe — TRD is among the leading causes of years lived with disability globally. For the subgroup with acute suicidal ideation with intent, conventional antidepressants take 4–6 weeks to reach therapeutic effect; the time-to-response gap is itself a clinical risk, and the rapid-acting indication addresses exactly that Fu et al. 2020. Ketamine is one of the few agents that meaningfully shifts that timeline.

payoff

For responders (45–65% of TRD patients in RCT settings; ~30% in real-world cohorts), the effect is rapid and substantial: a single IV infusion drives MADRS / HAM-D scores down by ~50% within 24 hours, with the response peaking at 24–72 hours and decaying over 1–2 weeks without maintenance Zarate 2006. Repeat dosing extends and consolidates the response. In SUSTAIN-1, maintenance esketamine plus oral antidepressant reduced relapse risk by ~51% in responders and ~70% in remitters over the trial period versus oral-antidepressant-plus-placebo-spray, with a number-needed-to-treat of 4 Daly et al. 2019. ELEKT-D demonstrated non-inferiority to ECT in non-psychotic TRD with substantially less cognitive impact — a meaningful payoff for patients who would otherwise face ECT's memory side effects Anand et al. 2023. The ASPIRE trials demonstrated rapid (within 24-hour) reduction in depression severity in patients hospitalised with active suicidal intent, though between-arm differences on the suicidality scale itself were not consistently significant — the major effect was on overall depression severity, with suicidality improving in both arms with intensive standard-of-care monitoring Fu et al. 2020 Ionescu et al. 2021.

failure-modes

(1) Stopping after the induction series. Without maintenance, the relapse rate within weeks is high (~50% by 1 month in responders without continuation dosing); ketamine's antidepressant effect is real but not durable as a one-shot. (2) Benzodiazepine concomitance. Several retrospective analyses suggest concurrent benzodiazepine use blunts the antidepressant response, hypothesised to be via GABAergic interference with the glutamate surge. (3) Choosing the wrong route for the patient. Intranasal esketamine has narrower effect-size range and modest separation from active placebo; IV racemic ketamine has larger effect but no insurance coverage. (4) At-home oral / sublingual ketamine without monitoring. The COVID-era telehealth expansion produced a wave of mail-order ketamine providers; effect size of oral/sublingual ketamine in controlled trials is smaller and the abuse-potential supervision is absent. (5) Underestimating the abuse risk. Ketamine is Schedule III; the long-term safety profile of repeated sub-anaesthetic dosing over years (as maintenance therapy implies) is not established, and signal exists for psychological dependence and bladder toxicity at chronic high exposure.

practicalities

Access is the dominant practical issue. Spravato is the only FDA-approved option for intranasal/inhaled ketamine and is insurance-coverable (Medicare Part B; most commercial plans after prior authorisation requiring documented failure of ≥2 oral antidepressants). Cash-pay session cost: $590–$1,600 (drug + monitoring); with commercial insurance copays typically $10–$50 per session; Medicare 20% coinsurance ≈ $140–$240. Full induction course ~$6,000 with standard insurance, $4,700–$10,500 cash. Spravato administration is restricted to REMS-certified facilities — no take-home — with mandatory 2-hour post-dose monitoring; patient cannot drive for the rest of the day. IV ketamine for depression remains off-label and almost entirely cash-pay: $400–$800 per session in most US markets, induction series $2,400–$6,400, plus consultation, integration, and maintenance fees. Insurance reimbursement for IV ketamine is uncommon; HSA/FSA funds usually qualify when prescribed. Geographic variability is large — ketamine clinics cluster in major metros and are sparse in rural areas; Spravato REMS clinics are more evenly distributed because reimbursement supports them.

history

Ketamine was synthesised by Calvin Stevens at Parke-Davis in 1962 and approved by the FDA as an anaesthetic in 1970. Recreational use emerged in the 1980s–1990s; DEA scheduled it as Schedule III in 1999. The antidepressant signal was first observed serendipitously in chronic-pain patients receiving sub-anaesthetic ketamine; Berman et al. published the first depression RCT in 2000 Berman 2000. The 2006 NIMH replication and subsequent free-standing IV ketamine clinics (starting around 2010) pre-dated regulatory approval by nearly a decade. Esketamine's approval in 2019 was the first novel-mechanism antidepressant approved in roughly 30 years FDA 2019. The 2025 monotherapy approval marked the transition from add-on-only to standalone indication FDA 2025.

alternatives

For TRD, the conventional next-step options after two failed oral antidepressants include: augmentation with atypical antipsychotics (aripiprazole, quetiapine, brexpiprazole — strong evidence, oral, insurance-covered, side-effect profile of metabolic syndrome and tardive risk); lithium augmentation (longstanding evidence, narrow therapeutic window, requires labs); MAOIs (effective in atypical depression, dietary restrictions); electroconvulsive therapy (the gold-standard rapid-acting treatment for severe TRD, with substantial cognitive side effects, non-inferior to IV ketamine in ELEKT-D for non-psychotic TRD Anand et al. 2023); repetitive transcranial magnetic stimulation (FDA-approved, no cognitive side effects, daily sessions over 4–6 weeks, modest effect size); vagus nerve stimulation (FDA-approved, surgical, slow onset). Within the rapid-acting / NMDA-glutamatergic class, the next entrant is dextromethorphan-bupropion (Auvelity), oral, FDA-approved 2022 for MDD; rapid-acting effects more modest than ketamine.

out-of-scope

Psychedelic-assisted psychotherapy with psilocybin or MDMA (different mechanism, different regulatory status, distinct evidence base). Ketamine for chronic pain syndromes (CRPS, neuropathic pain) — different evidence base. Ketamine for PTSD, OCD, eating disorders — emerging off-label use, evidence thin. Adolescent ketamine — limited approved indication.

Credibility range

Optimist case. Ketamine is the first genuinely rapid-acting antidepressant in 70 years of pharmacology. The replication base for the single-infusion effect in TRD is among the strongest in psychiatry: multiple independent RCTs across institutions, active-placebo controlled, with effect sizes (Hedges' g ≈ 1.5 for IV) that dwarf SSRI separation from placebo. The mechanism story is biologically coherent — glutamate / mTOR / synaptogenesis is a credible model of how depression's putative synaptic deficit gets reversed. Non-inferiority to ECT in ELEKT-D is a near-best-case clinical result, given ECT's 80-year track record as the most effective treatment for severe depression Anand et al. 2023. For acutely suicidal patients, ketamine offers a time-to-response that no oral antidepressant can match — clinically, the alternative is inpatient hospitalisation while waiting weeks for SSRIs to work Fu et al. 2020. The 2019 and 2025 FDA approvals reflect a regulator's judgement that the totality of evidence supports utility.

Skeptic case. Two of the three pivotal short-term esketamine initiation trials failed their primary endpoint; the FDA, unusually, allowed a withdrawal study (SUSTAIN-1) to count as a second positive pivotal trial — a regulatory path described as unprecedented and reviewed critically in the same journal that published the positive trial. Effect size of intranasal esketamine over active placebo is modest (4-point MADRS difference; Hedges' g ~0.3). The blinding problem is fundamental: ketamine's dissociative effects are unmistakable, and even midazolam-controlled trials are not fully blind. Long-term safety data beyond 1–2 years are limited; bladder toxicity at the doses used in chronic recreational users (~25% incidence) is a precedent that the field has not adequately reconciled with maintenance-dosing protocols. The free-standing IV ketamine clinic industry runs off-label with no consistent oversight, and at-home oral ketamine telehealth providers have proliferated without long-term data. Ketamine is Schedule III with documented abuse potential. The 51–70% relapse-prevention numbers from SUSTAIN-1 conflate efficacy with discontinuation effects — patients withdrawn from active medication may relapse partly because they were withdrawn. Janssen-funded trials dominate the esketamine evidence base; investigator-initiated trials are sparser.

Author's call. Ketamine is real, important, and over-marketed. The proof-of-concept replication in IV racemic ketamine is one of the most robust findings in psychiatric pharmacology, and for the subset of patients who have failed multiple lines of conventional treatment, the rapid response is genuinely transformative on a timescale nothing else offers. But it is not a one-shot cure: durability requires maintenance, and the maintenance evidence base is thinner than the acute-response evidence. The intranasal esketamine evidence is weaker than the IV racemic ketamine evidence, but esketamine's regulatory approval and insurance coverage make it the accessible option for most patients. The action verb is decide, not do — this is a prescription intervention requiring a psychiatrist, prior authorisation, monitored administration, and an informed conversation about long-term risks that the field doesn't yet fully understand. Evidence: 4. Controversy: 3 — active debate about regulatory standards, durability, and the gap between trial and real-world response, but no foundational disagreement on whether the acute effect is real.

Stakeholder + incentive map

Janssen / J&J. Patent-holder for esketamine; funded the pivotal trials. Commercial incentive to position Spravato as first-line in TRD; benefited from the 2025 monotherapy approval that allows use without concomitant oral antidepressant.
IV ketamine clinics. Cash-pay business model; ~$400–$800 per infusion. Incentive to recommend induction + maintenance protocols. Variable adherence to clinical-trial protocols; some use unvalidated routes (oral troches, at-home).
Academic psychiatry / APA. Generated the consensus statement urging caution; Sanacora et al. 2017 explicitly flagged that lay demand had outpaced safety data Sanacora et al. 2017. Investigator-initiated trials (ELEKT-D, Murrough) have produced the more conservative-but-rigorous evidence.
FDA. Approved Spravato through Breakthrough Designation with an unusually flexible pivotal-trial standard; published commentary in academic journals questioned whether the bar was lowered.
Patient advocacy / lay media. Strong demand from patients with TRD and their families; media coverage has been favourable, sometimes uncritical of effect sizes and durability.
Insurance payers. Cover Spravato (FDA-approved) with prior authorisation; do not cover off-label IV ketamine. This creates a coverage cliff that drives some patients toward esketamine despite weaker effect-size evidence.
Telehealth ketamine providers (Mindbloom, Joyous, others). Mail-order oral/sublingual ketamine; the regulatory gray zone exists because Schedule III allows certain telehealth prescribing flexibilities. Long-term data essentially absent.

Population variability

TRD vs first-line MDD. Evidence base is strongest in TRD (failed ≥2 oral antidepressants); efficacy in less-severe MDD is plausible but understudied.
Bipolar depression. Smaller but consistent positive signal; not FDA-approved indication. Concern about manic switching is real but appears low-frequency in monitored trials.
Age. Phase 3 trial in ≥65 (ESKETINTRD3005) failed primary endpoint; older adults may respond less robustly or require different dosing. Adolescent data sparse; not FDA-approved.
Suicidal ideation. Specifically approved (2020) for MDD with acute suicidal ideation; clinical use as bridge-to-treatment while conventional antidepressants ramp up.
Concurrent benzodiazepine use. Retrospective signal suggests blunted response.
Prior trauma / dissociative history. Increased risk of distressing dissociation; requires pre-screening and integration support.
Substance-use history. Higher abuse-potential risk; ketamine misuse history is a hard relative contraindication.

Knowledge gaps

Long-term (5+ year) safety of maintenance dosing — cognitive, bladder, dependence — remains uncharacterised.
Optimal maintenance schedule for IV ketamine has no consensus; clinic practice varies widely (every 2 weeks vs every 6 weeks).
Whether oral/sublingual ketamine achieves clinically meaningful effects relative to placebo, in adequately blinded trials, remains unsettled.
Predictors of response: no validated clinical or biomarker model exists for who responds vs who doesn't.
Whether psychotherapy adjunct (ketamine-assisted psychotherapy) adds durable benefit beyond medication-only protocols — the integration premise is plausible but RCT evidence is thin.
The dissociation-as-mediator question — is the psychoactive experience load-bearing, or is it an irrelevant side effect of the active antidepressant mechanism?
Equivalence of esketamine to racemic ketamine: no large head-to-head RCT exists.

Scope vs. brief. The brief named depressive symptoms, suicidal ideation, dissociative side effects, durability of response, and clinic access. All five are addressed end-to-end — depressive-symptom efficacy in evidence and payoff, suicidal-ideation indication in audience and stakes, dissociation as side effect in protocol and as misconceptions ("the trippy part is the point"), durability in payoff and failure-modes, and clinic access in practicalities. No narrowing.

Action / cadence calls. Chose decide over do because this is a prescription intervention requiring a psychiatrist, prior authorisation, and an explicit risk-benefit conversation that conventional antidepressants don't demand. Chose course over weekly/daily because the canonical protocol is a bounded induction phase followed by indefinite maintenance — the induction is the editorially salient shape, even though maintenance dosing extends.

Contraindications. The closed vocabulary fits the major hard stops (pregnancy, breastfeeding, cardiac-condition, uncontrolled-hypertension) but does not have a token for "history of substance misuse" or "current psychosis / mania," which are equally important relative contraindications and are called out in the article body. Worth proposing a substance-use-history token for future-rev — would also apply to several other entries (alcohol, opioid analgesia).

Audience scoping. Trial evidence is centred in 18–65; the one phase 3 trial in ≥65 missed its primary endpoint and there is no approved adolescent indication. Deliberately did not populate audience.ages — the entry's reader is realistically any adult considering this for themselves or a family member, and over-scoping the ages field would shrink reach for an editorial entry that talks honestly about the age gradient in its body.

Evidence score = 4, not 5. The RCT base is large and the FDA approvals are settled, but two of three pivotal esketamine short-term initiation trials missed primary endpoint, the FDA's acceptance of a withdrawal trial as a positive pivotal was unprecedented, and long-term safety data beyond 1–2 years are limited. The skeptic case is real enough to keep this off 5.

Controversy = 3. The regulatory path for esketamine, the dissociation-as-mediator question, durability, and the off-label IV clinic ecosystem are all genuinely contested among reasonable experts. Not 4 because there is no foundational disagreement on whether the acute antidepressant effect is real — the disputes are around adequacy of evidence and long-term unknowns.

Future-link candidates.

Electroconvulsive therapy — the head-to-head non-inferior comparator and the rapid-acting alternative for severe TRD.
Repetitive transcranial magnetic stimulation — the device-based outpatient alternative.
Atypical antipsychotic augmentation — aripiprazole / quetiapine / brexpiprazole for TRD, the conventional oral next step.
Psilocybin-assisted therapy — the related but mechanistically distinct rapid-acting alternative on the regulatory horizon.
Treatment-resistant depression as a condition entry separate from this treatment entry.

Separate-entry candidates. Ketamine for chronic pain, ketamine for PTSD, and at-home oral / sublingual ketamine via telehealth providers each have their own evidence base and stakeholder map; flagged in out-of-scope and worth their own entries rather than expanding this one. The at-home telehealth entry in particular is a meaningful safety / regulatory story that this entry only gestures at.

Bladder toxicity decision. Included a warning-callout reference because it is the most under-discussed long-term risk for maintenance-dosing patients and the case-report signal at therapeutic doses is rising. Did not lean on the recreational-user prevalence numbers in reader-facing prose because the doses and frequencies don't transfer directly — kept that detail in the research dossier.