Ketamine treatment for individuals with treatment-resistant depression: longitudinal qualitative interview study of patient experiences

Ketamine has attracted attention as a potential rapid-acting antidepressant, particularly for people with treatment-resistant depression (TRD). Earlier randomised and uncontrolled studies have shown that single intravenous infusions can reduce depressive symptoms and suicidal ideation, but effects are typically transient. Most published work has focused on single infusions; repeated or maintenance regimens are less well characterised, and there is little qualitative research exploring how patients experience ketamine treatment over time, including expectations, side-effects and the emotional impact when benefits fade. Lascelles and colleagues set out to explore longitudinal patient experiences of ketamine treatment delivered in a routine UK clinic. The study aimed to document patients' expectations before treatment, short- and longer-term effects on mood and suicidal thinking, side-effects, and overall perspectives on the value and limitations of ketamine therapy for TRD within a real-world, self-funded clinic context.

This was a longitudinal qualitative study conducted with self-funding patients at a single ketamine clinic in the UK. Inclusion criteria were age 18 or over, a diagnosis of TRD, assessed as appropriate for ketamine by clinic clinicians, no prior ketamine treatment for depression, capacity to consent and fluency in English. Clinic eligibility required current depression, prior trials of at least two antidepressant types (minimum 6 weeks each at adequate dose) and at least one psychological treatment. Standard clinic practice was an initial course of three intravenous ketamine infusions (0.5 mg/kg) given weekly, followed by a 3–4 week review; subsequent personalised care could include oral ketamine (taken at home, typically twice weekly), combined oral/intravenous schedules, or intermittent intravenous treatments. Patients commonly remained on other oral antidepressants. Recruitment took place during clinic assessment: clinicians provided information and referred interested individuals to the researcher for informed consent. Of 38 patients approached, 12 (31.6%) agreed to participate. Semi-structured one-to-one interviews were planned at three time points: before treatment, ~2 weeks after starting, and at ≧2 months or following treatment cessation within that period. Timing varied due to participant availability. Interviews were conducted face-to-face or by phone/video, lasted 20–60 minutes, and preceded by completion of the Beck Depression Inventory (BDI). Participants were also invited to keep diaries and later to provide retrospective reflections; however, the study relied principally on interview data because diary and email returns were limited. Transcripts were analysed using inductive thematic analysis following Braun and Clarke's approach, taking a semantic orientation (themes identified within and across participants using a 'following the thread' method). Two researchers (K.L. and F.B.) reached consensus on themes with supervision from L.M.; NVivo software supported the coding. Patient and public input was used to refine interview materials, and ethical approval was granted by the South Central Oxford Research Ethics Committee (reference 17/SC/0106).

Twelve participants (6 female, 6 male) were interviewed; median age was 57 years (range 21–70). Duration of depressive illness ranged from 10 to 50 years. Two participants had additional psychiatric diagnoses (bipolar I and obsessive–compulsive disorder). All had extensive prior treatment histories; three had electroconvulsive therapy. At baseline eight participants were taking antidepressant medication, four took two or more agents, two were on mood stabilisers and two on anxiolytics. Suicidal ideation had occurred at some point in all participants, and six had a history of self-harm. Engagement with the study interviews varied: eight participants completed all three interviews, three completed only the first two (treatment stopped early for side-effects or limited effect), and one completed interviews one and three only. Ten participants received at least three intravenous treatments; six progressed to oral ketamine and six remained in treatment at study end. Pre-treatment expectations were dominated by hope for symptom relief; most participants had learned about ketamine via internet searches (n = 5) or news coverage (n = 4). Half moderated expectations because of prior treatment failures. Main concerns were treatment not working (n = 5) and dissociative/psychotomimetic side-effects (n = 6); few expressed concerns about dependency. Mood outcomes were heterogeneous. Ten participants reported some improvement in mood at some point; two were classed as non-responders. Improvements ranged from transient (hours to days) to several weeks. Common symptomatic gains reported were reduced anxiety (n = 10), improved self-worth (n = 9) and increased energy (n = 8). Functional gains—better day-to-day functioning, reduced self-criticism and increased socialisation—were noted by nine participants each. Many noticed improvement after the first infusion (nine participants), and five described the second infusion as producing a notably greater effect, sometimes described as life-changing while it lasted. Eight participants reported reduced suicidal ideation; for four this cessation was stark, and for four the reduction was subtler or less intense. In several cases reduced suicidal thinking paralleled mood improvement, but for a minority suicidal ideation remained lower even as mood deteriorated. Longer-term patterns varied among those receiving three or more treatments (n = 10): two experienced sustained improvements, two remained non-responders, and six described waning benefit over time. Of the six whose benefit waned, three retained reduced suicidal ideation despite mood relapse, while three experienced return of strong suicidal thoughts, including some preparatory or ambivalent behaviours in two individuals. Six participants moved to oral ketamine; four found oral dosing less effective than infusions, two reported oral benefit exceeding infusions (one citing cumulative dosing, one reduced dissociative effects). By study end two participants reported substantial ongoing recovery, three reported limited residual benefit, five had ceased to benefit and two had never derived benefit. Financial cost was a recurrent problem, particularly where benefit was not sustained. Adverse effects were common. Dissociation occurred in 11 of 12 participants; six reported perceptual disturbances (seeing/hearing unusual things), five blurring of vision and three nausea. For most these effects were transient (under 1 to 3 hours). Post-treatment tiredness (n = 6) and headaches (n = 4) were reported; one severe 4-day headache following the first infusion led to treatment termination. A minority expressed some concern about potential dependency or the risk of obtaining non-clinical ketamine if costs became prohibitive. Participants suggested treatment improvements including individualised dosing, closer clinical monitoring, clearer management of expectations and provision of adjunctive psychological therapy.

Lascelles and colleagues interpret their findings as broadly complementary to existing trial and case-report literature: ketamine often confers rapid improvements in mood and reduces suicidal ideation, but these benefits are frequently transient. Considerable heterogeneity in intensity and duration of response was evident, with a minority sustaining benefit over the study period. The particularly notable response after the second infusion for some participants mirrors observations in other reports. Authors note a possible partial dissociation between effects on mood and suicidal ideation, since in some cases suicidal thinking remained reduced even as mood declined; this is consistent with recent reviews suggesting an independent impact of ketamine on suicidal ideation. Side-effects experienced during administration—most commonly dissociation and transient perceptual changes—aligned with wider literature. Although dependency as an adverse outcome has not been established clinically, some participants expressed concern about misuse or the temptation to procure street ketamine if therapeutic treatment became unaffordable. Media coverage appears to influence help-seeking, contributing to patient hope and, in some cases, disappointment when benefits do not persist. The emotional consequences of treatment waning included marked declines in hope and, for a few participants, increased suicidal ideation, a response the authors underline as important to disclose during consent. Strengths of the study include its real-world clinic setting and a longitudinal qualitative design that captures patient perspectives beyond standard clinical measures. Limitations acknowledged by the authors are the small sample size, short follow-up, variability in interview timing and treatment courses, reliance mainly on self-report interview data (although the BDI was used at key points), and limited generalisability because all participants were self-funding. The authors also note uncertainty about whether findings would generalise to licensed esketamine nasal-spray treatment delivered in different clinical contexts and doses. They conclude that informed consent should explicitly address the high likelihood of relapse and the potential emotional consequences, including increased hopelessness when treatment ceases to be effective, and they emphasise the need for research into adjunctive interventions—particularly psychological therapies—and closer monitoring to sustain early benefits and detect relapse promptly.