Acute psychoactive effects of intravenous ketamine during treatment of mood disorders: analysis of the Clinician Administered Dissociative State Scale

Since 2000, randomized controlled trials have shown that subanesthetic doses of ketamine produce rapid antidepressant effects in unipolar and bipolar depression. At commonly used clinical doses (0.5 mg/kg IV over 40 minutes), ketamine also produces short-lived psychoactive effects—changes in perception and alertness—that typically resolve within an hour. These acute effects are often labelled "dissociative," and the Clinician Administered Dissociative State Scale (CADSS) has been the most widely used instrument to quantify them in ketamine research. However, the CADSS was developed and validated in other clinical contexts (notably PTSD) and, based on the authors' clinical and trial experience, may not fully capture the phenomenology elicited by subanesthetic ketamine in mood-disordered patients. Van Schalkwyk and colleagues set out to evaluate whether the CADSS adequately measures the acute psychoactive effects of intravenous ketamine and to identify aspects of the ketamine experience that the scale may miss. To address these aims they performed an exploratory factor analysis (EFA) of CADSS ratings obtained at the 40-minute infusion point from a pooled sample of 110 patients who received ketamine, and complemented the quantitative work with semi-structured qualitative interviews in a subsample to characterise first-person accounts of the acute experience. The mixed-methods approach was intended to show both where the CADSS aligns with patient experience and where it may require revision for ketamine research.

The investigators pooled existing CADSS data from three sources: a randomised controlled trial of ketamine versus saline with an open-label extension (N = 67), an open-label ketamine trial (N = 16), and clinical treatment records from an interventional psychiatry service (N = 27), yielding CADSS scores at the 40-minute infusion point for 110 individuals. The extraction notes some inconsistencies in the number of CADSS items reported across sections; the version used in ketamine trials was described as clinician-administered items scored 0–4. Inclusion/exclusion criteria for the original trials were referenced as published elsewhere; the clinical sample was subject only to routine clinical appropriateness for ketamine. Raters had a minimum education level and trial-specific rater training. Institutional review board approval was obtained for the qualitative component and the clinical data were deemed exempt. Quantitative analyses were conducted in SPSS v22. Data were checked for missingness (minimal, <2% per item) and distributional assumptions (skewness and kurtosis reported within acceptable ranges). Sampling adequacy was tested (Kaiser-Meyer-Olkin = 0.849) and Bartlett's test was significant (p < 0.001). An exploratory factor analysis was performed using principal axis factoring with oblique (Oblimin) rotation, anticipating correlated factors. Factor number was chosen by eigenvalues and inspection of a scree plot. Internal consistency was assessed with Cronbach's alpha; each item was evaluated for its correlation with the total score and its effect on alpha when removed. For qualitative data, semi-structured interviews were conducted with a convenience subsample of 10 participants (mean age 52.6 years, SD 13.6; 7 female; 9 with unipolar major depressive disorder, 1 with bipolar disorder). Interviews were performed by a co-author with no prior therapeutic relationship to participants and were audio-recorded and transcribed. Inductive thematic analysis was carried out in QSR Nvivo: iterative coding led to a code list and then to higher-order themes grounded in participants' descriptions rather than a pre-existing framework. The qualitative analysis was reviewed by a co-author and participants were engaged to check the salience of emerging themes. Finally, the researchers performed a comparative analysis to map overlap and divergence between quantitative factor structure and qualitative themes.

The pooled sample comprised 110 individuals assessed at 40 minutes into a 0.5 mg/kg IV ketamine infusion. The mean total CADSS score reported in the Results section was low (6.2, SD = 9.1), and 27 participants (about 25%) had a score of zero. Missing data were minimal and distributional checks did not indicate gross departures from normality. EFA produced a single-factor solution that the authors labelled a dissociation factor. Principal axis factoring with oblimin rotation identified a primary factor with an eigenvalue of 6.9; seven additional factors had eigenvalues between 1 and 2, but a scree plot showed a steep drop after the first factor. The single-factor solution explained approximately 30% of the variance and initially included 18 items, of which 16 had loadings > 0.35 and were retained as indicators. Five items (reported as items 13, 17, 19, 21 and 23) showed low correlations with the total score (ranging roughly 0.14–0.34) and increased internal consistency when removed. Cronbach's alpha for the single-factor set was 0.71. Two items (7 and 8) had the lowest factor loadings (0.35 and 0.32) and did not cluster clearly with the main themes. The highest-loading items (five of the six top loadings) related to perceptual disturbances, specifically altered sensory experience and time perception; other retained items reflected depersonalisation and memory changes. Qualitative interviews (n = 10) generated four prominent themes: altered time and sensory perception, unusual bodily sensations, a sense of peace, and disinhibition. Altered perception was the most frequently reported and corresponded closely to several high-loading CADSS items: six participants described experiences aligned with items that had high correlations (> 0.6) with the CADSS total (items 1, 2, 9 and 15). Participants described feeling "halfway under" or "spacey and out of it," often with a slowing of subjective time rather than acceleration. Unusual bodily sensations were reported by four participants; these descriptions did not always match the specific bodily anchors used in certain CADSS items (for example, reports of feeling "woozy" or light-headed rather than clearly "large" or "small"). A sense of peace—encompassing mood improvement, calmness, and altered cognitive appraisal—was explicitly reported by four participants, some of whom described enduring benefit from mentally revisiting the infusion. Disinhibition emerged in accounts from three participants who recalled being more talkative or less guarded; one participant reported concern about sounding intoxicated. The authors included brief participant quotes to illustrate these themes. When comparing quantitative and qualitative findings, the researchers noted overlap in altered perception items, but also identified phenomena (sense of peace, disinhibition) that were either absent from the CADSS or only weakly represented by items that were infrequently endorsed.

Van Schalkwyk and colleagues interpret their findings as evidence that the CADSS only partially captures the acute psychoactive effects of intravenous ketamine. Average CADSS scores in this mood-disorder sample were substantially lower than scores reported in the CADSS's initial validation cohort (notably PTSD), and a quarter of participants endorsed no CADSS items, despite qualitative reports of prominent subjective effects. The authors propose two non-exclusive explanations: the CADSS may not be well matched to ketamine-induced phenomenology, or the dissociative phenomena seen in PTSD are more intense than those provoked by the ketamine doses used here. The single-factor EFA solution identified a coherent cluster of items related to altered sensory and temporal perception, depersonalisation and memory change; these items are supported by participant narratives and therefore appear valid for capturing a core component of the ketamine experience. At the same time, qualitative analysis revealed domains not well captured by the CADSS—chiefly a sense of peace and disinhibition—and indicated that some CADSS items that ostensibly target similar phenomena were seldom endorsed. The investigators therefore argue that certain CADSS items have low yield in this population and that other relevant phenomenology may need to be added or reframed to better reflect ketamine's acute effects. The discussion situates these experiential categories within neurobiological models: altered sensory processing and time perception align with known ketamine effects on auditory and visual processing, while disinhibition and depersonalisation may relate to changes in default mode network connectivity; the "sense of peace" might reflect decreased self-referential processing and altered anterior cingulate activity. The authors acknowledge several limitations: lack of contemporaneous alternative measures of subjective experience, absence of clinical outcome data linking acute phenomenology to antidepressant response, use of convenience sampling for both quantitative and qualitative components which may limit generalisability, and the relatively low mean CADSS score in the sample which could influence factor structure. They note the possibility that a different factor solution might emerge in samples with higher CADSS scores. Building on these results, the authors propose that a ketamine-specific revision of the CADSS could (A) add items assessing disinhibition and a sense of peace, (B) remove items with low yield in this context, and (C) reframe certain bodily-sensation items to better match patients' descriptions. They recommend further research to validate a revised instrument that more accurately quantifies acute ketamine experiences, especially because reliable measurement matters for understanding links between acute effects and clinical response, and for maintaining blinding in active-comparator trials.