Item-Level Symptom Trajectories Across the PHQ-9 During At-Home Ketamine-Assisted Therapy for Depression: A Real-World Retrospective Analysis of 36,847 Adults
Davis Carter, Ian Reardon, Jack Swain, Leonardo Vando
Preprints.org October 27, 2026 preprint DOI: 10.20944/preprints202610.2651.v1 (opens in new tab)
Study at a glance
AI-extracted from the abstract| Characteristics | Retrospective cohort analysis |
|---|---|
| Sample size | 36,847 |
| Population | Adults who initiated a new-client course of guided at-home sublingual or subcutaneous ketamine-assisted therapy for depression at Mindbloom, with a baseline PHQ-9 submitted between May 2021 and November 2025 |
| Duration | Six sessions; post-session assessments at Post-Session 2, 4, and 6 |
| Measures | PHQ-9 |
| Topics | Depression Esketamine Ketamine |
| Key findings | All nine PHQ-9 symptom dimensions improved during at-home ketamine-assisted therapy, with gains measurable by the first post-session assessment and strengthening through Post-Session 6. Post-Session 6 response rates ranged from 76.9% (sleep disturbance) to 94.9% (thoughts of being better off dead or of self-harm), and total PHQ-9 scores declined 52%, with 59.4% achieving a ≥50% reduction. The authors characterize the findings as exploratory and hypothesis-generating. |
Abstract
Background. The nine items of the Patient Health Questionnaire-9 (PHQ-9) index distinct symptom dimensions of major depressive disorder. Depression treatment is typically reported as change in the total score, so how individual symptom dimensions move during a rapid-acting treatment is seldom characterized. Ketamine is a rapid-acting glutamatergic antidepressant, but item-level trajectories during at-home, telehealth-supported ketamine-assisted therapy have not, to our knowledge, been described at this scale. Methods. We conducted a retrospective cohort analysis of 36,847 adults who initiated a new-client course of guided at-home sublingual or subcutaneous ketamine-assisted therapy for depression at Mindbloom, with a baseline PHQ-9 submitted between May 2021 and November 2025. For each PHQ-9 item, we analyzed the subset of patients elevated on that item at baseline (score of 2 or 3) and tracked change at three post-session timepoints (Post-Session 2, 4, and 6), computing within-subject change, percent reduction, paired Cohen's dz, the proportion improving by at least 1 point (response) and reaching 0 (resolution), and any worsening. The PHQ-9 total was analyzed among patients with at least moderate baseline depression (total ≥ 10), with response defined as a ≥ 50% reduction and remission as a total below 5. Results. Improvement was observed on every item, was measurable by the first post-session assessment (median 15 to 16 days), and strengthened through Post-Session 6. Post-Session 6 response rates (≥ 1-point improvement) ranged from 76.9% (sleep disturbance) to 94.9% (thoughts of being better off dead or of self-harm), with within-subject effect sizes from dz = 1.31 to 2.19, based on completer subsets of 175 to 1,769 patients per item. The largest percent reductions were on thoughts of being better off dead or of self-harm (73%), psychomotor change (71%), and negative self-evaluation (62%). Treatment-emergent ideation among patients with no or minimal baseline ideation was low, with emergence to a score of 2 or 3 in no more than 1.4% at any timepoint. The PHQ-9 total score declined 52% among Post-Session 6 completers (dz = 1.58), with 59.4% achieving a ≥ 50% reduction and 71.6% scoring below the moderate-depression threshold. Any patient-reported side effect was reported at 3.3% to 4.2% of post-session assessments and was not graded for severity or adjudicated for seriousness. Conclusions. In this large retrospective Mindbloom cohort, all nine PHQ-9 symptom dimensions improved during at-home ketamine-assisted therapy for depression, with early onset, progressive gains through six sessions, and large within-subject effect sizes. The findings are exploratory and hypothesis-generating, are consistent with a treatment effect, and would require controlled replication to confirm causality.
Comparable studies
Other observational and cohort studies on ketamine, most cited first.
| Study | Year | Design | Participants |
|---|---|---|---|
| Taming the Ketamine Tiger Volunteers from a prison population | 2010 | Observational study | n = 20 |
| Increased spontaneous MEG signal diversity for psychoactive doses of ketamine, LSD and psilocybin Humans | 2017 | Observational study | |
| Disruption of Frontal–Parietal Communication by Ketamine, Propofol, and Sevoflurane Surgical patients undergoing induction of anesthesia with intravenous ketamine | 2013 | Observational cohort | n = 30 |
| Do the dissociative side effects of ketamine mediate its antidepressant effects? Treatment-resistant inpatients with major depressive disorder or bipolar disorder | 2014 | Observational cohort | n = 108 |
| Concomitant BDNF and sleep slow wave changes indicate ketamine-induced plasticity in major depressive disorder Patients with treatment-resistant major depressive disorder | 2012 | Observational cohort | n = 30 |