Significant outcomes
-
Iv ketamine produced a significant reduction in depressive symptoms even when administrated to a heterogenous population under real-world clinical conditions. The results support the view of iv ketamine as an efficacious treatment in a variety of settings.
-
The reduction in depressive symptoms did not translate to a significant reduction in the need for inpatient care. This is in contrast with an earlier similarly designed study and indicate factors other than the antidepressant effect of ketamine being significant in the need for hospitalisation in corresponding populations.
-
The study highlights both treatment specific and extra-pharmacological factors as clinical challenges when implementing iv ketamine treatment in clinical practice and further emphasise the importance of structured treatment and evaluation protocols.
Limitations
-
The study sample was small and heterogeneous, did not include a control group and the patients were recruited from a single study site.
-
The documentation in the medical records was occasionally incomplete or unclear and both the diagnostic and treatment protocol lacked stringency.
-
The rationale for choosing iv ketamine instead of other treatments, and the occurrence of side effects, is unknown. The open-label nature of the study includes the possibility for a significant placebo effect.
Introduction
Ketamine was first developed in the 1960s as a rapidly acting anaesthetic (Corssen and Domino, Reference Corssen and Domino1966). It is a racemic mixture of its two enantiomers, (R)- and (S)-ketamine and derived from phencyclidine (PCP) (Zanos et al., Reference Zanos, Moaddel, Morris, Riggs, Highland, Georgiou, Pereira, Albuquerque, Thomas, Zarate, Gould and Witkin2018). Ketamine is classified as a noncompetitive NMDA-receptor antagonist. However, it exhibits a complex pharmacological profile with interactions at a range of receptors (Zanos et al., Reference Zanos, Moaddel, Morris, Riggs, Highland, Georgiou, Pereira, Albuquerque, Thomas, Zarate, Gould and Witkin2018). Ketamine has since its initial introduction gained prominence as an antidepressant drug (Corriger and Pickering, Reference Corriger and Pickering2019). This was largely initiated by the publication of the first placebo-controlled study following a growing body of preclinical research (Berman et al., Reference Berman, Cappiello, Anand, Oren, Heninger, Charney and Krystal2000). Subsequently, the promising initial results has prompted further investigations, in large part due to the ability of sub-anaesthetic doses of intravenous (iv) ketamine to produce a rapid and robust antidepressant effect in both unipolar and bipolar depression. (Wan et al., Reference Wan, Levitch, Perez, Brallier, Iosifescu, Chang, Foulkes, Mathew, Charney and Murrough2015; Corriger and Pickering, Reference Corriger and Pickering2019; Jelen and Stone, Reference Jelen and Stone2021). Recently, the effectiveness of ketamine has encouraged research into its usefulness as an alternative to electroconvulsive therapy (ECT) (Ekstrand et al., Reference Ekstrand, Fattah, Persson, Cheng, Nordanskog, Akeson, Tingstrom, Lindstrom, Nordenskjold and Rad2022; Anand et al., Reference Anand, Mathew, Sanacora, Murrough, Goes, Altinay, Aloysi, Asghar-Ali, Barnett, Chang, Collins, Costi, Iqbal, Jha, Krishnan, Malone, Nikayin, Nissen, Ostroff, Reti, Wilkinson, Wolski and Hu2023). However, the effectiveness of ketamine treatment on hospitalisation rates in a naturalistic setting remains understudied. With increasing clinical adoption of ketamine in the treatment of depression the need for investigation in a real-world clinical setting increase.
This study aims to explore the outcome with iv ketamine treatment in a real-world clinical setting, primarily measured as days spent as an inpatient.
Methods
Setting
The use of ketamine to treat psychiatric illness was not approved by the Swedish Medical Products Agency and there were no national or clinic-specific treatment guidelines during the period covered in this study. A survey conducted in 2016 by the Swedish Agency for Health Technology Assessment and Assessment of Social Services estimated that in Sweden in 2015 fewer than 75 patients received iv ketamine to treat depression ((SBU) 2017). The survey also reported that iv ketamine was viewed as an alternative to ECT and mainly offered to patients with treatment-resistant depression (TRD) and chronic suicidal ideation. TRD was in this instance defined as inadequate response to a minimum of two adequate antidepressants treatments. This is in line with the definition adopted by The US Food and Drug Administration (FDA) and the European Medicines Agency (EMA), that is, inadequate response to a minimum of two antidepressants despite adequacy of the treatment trial and adherence to treatment (McIntyre et al., Reference McIntyre, Alsuwaidan, Baune, Berk, Demyttenaere, Goldberg, Gorwood, Ho, Kasper, Kennedy, Ly-Uson, Mansur, McAllister-Williams, Murrough, Nemeroff, Nierenberg, Rosenblat, Sanacora, Schatzberg, Shelton, Stahl, Trivedi, Vieta, Vinberg, Williams, Young and Maj2023). However, a consensus definition of TRD does not currently exist (McIntyre et al., Reference McIntyre, Alsuwaidan, Baune, Berk, Demyttenaere, Goldberg, Gorwood, Ho, Kasper, Kennedy, Ly-Uson, Mansur, McAllister-Williams, Murrough, Nemeroff, Nierenberg, Rosenblat, Sanacora, Schatzberg, Shelton, Stahl, Trivedi, Vieta, Vinberg, Williams, Young and Maj2023).
The clinical setting was a University Hospital psychiatric clinic in the city of Umeå in northern Sweden. The clinic manages both inpatient and outpatient specialised care. Iv ketamine had been routinely used at the clinic during the years leading up to 2018 after which the treatment with iv ketamine was terminated by administrative decision. The clinic had no established dosing protocol for iv ketamine.
Study design
This retrospective register based observational study was conducted to assess the outcome of iv ketamine on the need for inpatient care in a real-world population. All participants enrolled in the study were at least 18 years of age and were included based on having received iv ketamine on a psychiatric treatment indication between 2015 and 2018. As there were no national or local clinical guidelines for ketamine-treatment at the time, treatment was initiated based on clinical assessment at the discretion of the treatment-responsible doctor and off-label. No clinic-specific inclusion criteria were present. Study participants were screened and selected according to anaesthesia records by matching psychiatric treatments and anaesthesia procedures at the hospital care register between the years 2015 and 2018 since ketamine treatment was administrated via the anaesthesia department. In total, 46 patients were found eligible for inclusion. No exclusion criteria were used. The participants psychiatric medical records were then examined. Ethical approval for this study was provided by The Swedish Ethical Review Authority.
Outcome measures
Two primary outcomes were chosen. First, the total number of days hospitalised, before and after treatment with iv ketamine, was measured. Second, the number of hospitalisations, before and after ketamine treatment, was measured. These measures were chosen because they reflected the study participants overall functioning and severity of illness. Our expectation was that iv ketamine would have been reserved for a patient population characterised by treatment resistance and chronicity, and that iv ketamine would be used as an alternative treatment to ECT. To investigate the impact of previous hospitalisation on our primary outcomes analysis was performed both on the total sample as well as the sub-group including only patients with at least one day’s pretreatment hospitalisation.
Based on the participants medical records, a starting point of iv ketamine treatment was defined. The starting point was set as the first day of the first index series of iv ketamine. From this starting date, a timespan of 365 days pre- and post-treatment was identified. Thereby, a consistent number of days pre- and post-ketamine treatment was obtained for each patient. This timespan was then examined regarding our primary outcomes for each patient.
To describe the study participants severity of depressed symptoms, self-reported Montgomery-Åsberg Depression Rating Scale score (MADRS-S; (Svanborg and Åsberg, Reference Svanborg and Åsberg1994; Svanborg and Åsberg, Reference Svanborg and Åsberg2001)) was registered when obtained, before and after the first index series with ketamine. In patients with no recorded MADRS-S scores (10 at baseline, 19 at endpoint) this data was replaced with a retrospectively estimated MADRS-S score through a review of the medical records by the clinical investigator (KS). The MADRS-S scale has been shown to have high concordance with the original MADRS scale (Mattila-Evenden et al., Reference Mattila-Evenden, Svanborg, Gustavsson and Asberg1996; Svanborg and Ekselius, Reference Svanborg and Ekselius2003). The clinical diagnosis used as an indication for ketamine treatment, sex, age and psychiatric comorbidity were registered. The total number of index series, as well as the number of individual ketamine treatments given, were also recorded. Response was determined as a symptom reduction greater or equal to 50% by the index-series of ketamine as measured by MADRS-S score (Nierenberg and DeCecco, Reference Nierenberg and DeCecco2001) and remission as a MADRS-S score less than 13 points (Svanborg and Ekselius, Reference Svanborg and Ekselius2003).
As the medical records did not give a complete and satisfactory registration of the given treatment doses of ketamine for every individual treatment a mean dose for the sample could not be calculated. Therefore, a typical treatment dose was estimated from a random sample of 20 participants.
Data analysis
Normality of the continuous variables (age, MADRS-S score, the number of hospitalisations and hospital days) were analysed with QQ plots. Wilcoxon signed rank test was used to compare pre- and posttreatment hospital days and hospital admissions. Paired samples t-test was used to compare pre- and posttreatment MADRS-S scores. Concordance between self-reported and estimated MADRS-S scores were analysed with independent samples t-test.
To explore associations for multicollinearity (r > 0,8) between background (sex and age) and clinical variables (diagnosis of mood disorder, number of ketamine treatments and -series, pre- and posttreatment hospitalisations, hospital days and MADRS-S scores) Spearman’s correlation coefficients were calculated. A logistic regression analysis with the presence of posttreatment hospital days as the dependent variable was performed including examining collinearity.
Statistical analysis was performed using IBM SPSS Statistics, version 25 (Armonk NY). The level of statistical significance was set at p < 0.05.
Results
Baseline characteristics of the sample are presented in Table 1. Forty-six patients were identified as having received iv ketamine treatment at the clinic between 2015 and 2018. Of these, 37 had complete, retrievable medical records, and 9 patients were referred from a different region with only partial medical records available.
Table 1. Baseline characteristics
Pretreatment = 365 days prior to start of index series.
Posttreatment = 365 days after start of index series.
SD = Standard Deviation; IQR = Interquartile range.
A wide diagnostic spread was observed. Thirty-three patients (72%) had a mood disorder ICD-10 diagnosis, of which 20 (43%) had a diagnosis of depression and 13 (28%) a diagnosis of bipolar disorder. Of the latter, 46% had bipolar depression and 54% hade an unspecified bipolar disorder. Of those with anxiety disorders as a main diagnosis, 50% had a F41 diagnosis, 38% had a F43 diagnosis and 12% had a F45 diagnosis. Those with other diagnoses included 40% with an eating disorder, 40% with a neuropsychiatric disorder and 20% with a personality disorder as the main diagnosis. In total, twelve patients (26%) had registered psychiatric comorbidities, with personality disorders being the most common (9%).
Thirty-one per cent had received ECT pretreatment and 14% received ECT posttreatment during follow-up.
The estimated given mean ketamine dose per treatment was 45 mg (SD 12.7) and roughly in line with the widely adopted dose of 0,5 mg/kg (Correia-Melo et al., Reference Correia-Melo, Leal, Carvalho, Jesus-Nunes, Ferreira, Vieira, Magnavita, Vale, Mello, Nakahira, Argolo, Cardoso, Souza, Fontes, Ferreira, Araujo-de-Freitas, Tuena, Echegaray, Cavalcanti, Lucchese, Bandeira, Telles, Lima, Sampaio, Silva, Marback, Del-Porto, Abreu, Sarin, Paixao, Carvalho, Machado, Turecki, Lacerda and Quarantini2018). Noteworthy was the occasional inclusion of esketamine. When encountered, esketamine was converted to an equivalent dose of racemic ketamine using accepted ratios (Correia-Melo et al., Reference Correia-Melo, Leal, Carvalho, Jesus-Nunes, Ferreira, Vieira, Magnavita, Vale, Mello, Nakahira, Argolo, Cardoso, Souza, Fontes, Ferreira, Araujo-de-Freitas, Tuena, Echegaray, Cavalcanti, Lucchese, Bandeira, Telles, Lima, Sampaio, Silva, Marback, Del-Porto, Abreu, Sarin, Paixao, Carvalho, Machado, Turecki, Lacerda and Quarantini2018).
Primary outcomes
Thirty-seven patients were eligible for the main analysis regarding our primary outcomes. There was no significant difference between pre- and posttreatment hospital days (p = 0.170, t-test), or number of hospitalisations (p = 0.740). Neither was there a difference between pre- and posttreatment hospital days (p = 0.230), or hospital admissions (p = 0.943) in the sub-group (n = 23) including only patients with at least one day’s pretreatment hospitalisation.
MADRS score
Estimated MADRS-S scores exhibited equal variance and no significant difference to the self-reported scores (p = 0.657 pretreatment, p = 0.442 posttreatment, t-test). MADRS-S score was significantly reduced after the initial index-series of ketamine for both self-reported (p = 0.001, mean difference 9 points, SD 10.9, t-test) and combined, that is, self-reported and estimated scores (p < 0.001, mean difference 9 points, SD 11.3, t-test).
Thirteen patients (31%) achieved a response and nine (21%) remission.
Post hoc analysis
The results of the Spearman correlation matrix are presented in Table 2. A moderate negative correlation between age and MADRS-S score pretreatment can be noted (r = −0.553, p < 0.000).
Table 2. Spearman’s correlation for MADRS-S pre- and posttreatment scores and background variables
Logistic regression, with the dependent variable as posttreatment days hospitalised was performed to investigate the impact of socio-demographic and clinical predictors on treatment outcome (Table 3). The independent variables in the model included sex, age, binary mood disorder diagnosis (F3 diagnosis yes/no), the number of received ketamine treatments and MADRS-S score change during index treatment. None of the predictors showed statistical significance in the model.
Table 3. Logistic regression with the presence of posttreatment hospital days as the dependent variable
Discussion
Iv ketamine treatment produced a significant reduction of depressive symptoms. This is consistent with findings seen in other studies (Marcantoni et al., Reference Marcantoni, Akoumba, Wassef, Mayrand, Lai, Richard-Devantoy and Beauchamp2020; McIntyre et al., Reference McIntyre, Rosenblat, Nemeroff, Sanacora, Murrough, Berk, Brietzke, Dodd, Gorwood, Ho, Iosifescu, Lopez Jaramillo, Kasper, Kratiuk, Lee, Lee, Lui, Mansur, Papakostas, Subramaniapillai, Thase, Vieta, Young, Zarate and Stahl2021) supporting the view of iv ketamine as an efficacious treatment in a variety of settings. Compared to a 2018 retrospective clinical study including TRD patients assessed as ultra-resistant (Thomas et al., Reference Thomas, Baker, Lind and Dursun2018) and a 2020 meta-analysis examining the efficacy of ketamine on TRD (Marcantoni et al., Reference Marcantoni, Akoumba, Wassef, Mayrand, Lai, Richard-Devantoy and Beauchamp2020) remission rates were higher and response rates lower. Both the response and remission rates were higher in our sample compared to a Canadian, retrospective, naturalistic study examining the effectiveness of iv ketamine on depressive symptoms, suicidal ideation and functional disability in major depressive disorder and bipolar disorder (McIntyre et al., Reference McIntyre, Rodrigues, Lee, Lipsitz, Subramaniapillai, Gill, Nasri, Majeed, Lui, Senyk, Phan, Carvalho, Siegel, Mansur, Brietzke, Kratiuk, Arekapudi, Abrishami, Chau, Szpejda and Rosenblat2020). A 2024 retrospective study examining clinical outcomes of iv ketamine for depression found both lower response and remission rates after 6 weeks of treatment compared to our sample (Pfeiffer et al., Reference Pfeiffer, Geller, Ganoczy, Jagusch, Carty, Festin, Gilmer, Martis, Ranganathan, Wiechers and Hosanagar2024). However, in the KetECT and ELEKT-D studies, examining the comparative effectiveness of iv ketamine to ECT, remission rates in the iv ketamine sample was distinctly higher than seen in our study sample (Ekstrand et al., Reference Ekstrand, Fattah, Persson, Cheng, Nordanskog, Akeson, Tingstrom, Lindstrom, Nordenskjold and Rad2022; Anand et al., Reference Anand, Mathew, Sanacora, Murrough, Goes, Altinay, Aloysi, Asghar-Ali, Barnett, Chang, Collins, Costi, Iqbal, Jha, Krishnan, Malone, Nikayin, Nissen, Ostroff, Reti, Wilkinson, Wolski and Hu2023). The same was found in a 2022 retrospective analysis of iv ketamine for depression in a real-world setting (McInnes et al., Reference McInnes, Qian, Gargeya, DeBattista and Heifets2022). These inconsistencies might reflect the diagnostic heterogenicity and the severity of depressed mood in our study sample. It underlines the importance of maintaining a strict treatment protocol in the clinical setting to reliable produce expected results. Lack of a strict protocol increases the risk of uncertain treatment outcomes.
Regarding our primary endpoints it´s noteworthy that even though a significant reduction in symptom score and hence treatment effect was observed, this did not translate to a significant reduction in the need for inpatient care. This is in contrast with an earlier similarly designed study where treatment with oral ketamine significantly reduced the need for inpatient care for patients with TRD and post-traumatic stress disorder (Hartberg et al., Reference Hartberg, Garrett-Walcott and De Gioannis2018). The lack of significant effect regarding our primary endpoints may therefore be due to the presence of factors other than the pharmacological antidepressant effect of ketamine such as potential misuse, diagnostic variability and expectancy bias.
In the multivariate analysis, neither sociodemographic factors such as age and sex nor the treatment-specific variable of number ketamine treatments received were found to be significantly affecting the posttreatment days hospitalised. This is in line with previous investigations regarding clinical predictors for effectiveness of iv ketamine (Rong et al., Reference Rong, Park, Rosenblat, Subramaniapillai, Zuckerman, Fus, Lee, Pan, Brietzke, Mansur, Cha, Lui and McIntyre2018) and brings into focus the need to develop clinically useful markers to predict the effectiveness of ketamine treatment. Efforts investigating clinical variables, biomarkers and neuroimaging results are ongoing (Rong et al., Reference Rong, Park, Rosenblat, Subramaniapillai, Zuckerman, Fus, Lee, Pan, Brietzke, Mansur, Cha, Lui and McIntyre2018; Meshkat et al., Reference Meshkat, Ho, Cao, Teopiz, Rosenblat, Rhee, Di Vincenzo, Ceban, Jawad and McIntyre2023). However, further research is needed before these findings can be reflected in the development of feasible clinical tools to personalise ketamine treatment strategy.
Interestingly, nor had a mood disorder diagnosis or reduction of depressive symptoms an association on hospitalisation in the multivariate analysis. This indicates that extra-pharmacological factors, such as social isolation, might contribute to the need for inpatient care in the study sample. Previous research has shown patients with strong social support being less likely to be admitted to psychiatric hospital (Albert et al., Reference Albert, Becker, McCrone and Thornicroft1998; Van Veen et al., Reference Van Veen, Wierdsma, van Boeijen, Dekker, Zoeteman, Koekkoek and Mulder2019) and a 2022 meta-analysis examining the social dimension of suicidal aetiology emphasised social isolation as a risk factor for suicide (Motillon-Toudic et al., Reference Motillon-Toudic, Walter, Séguin, Carrier, Berrouiguet and Lemey2022).
Worth highlighting is the wide diagnostic spread seen in the study sample. It is possible that in some cases treatment with ketamine was initiated based on a clinical presentation of depressed mood, which was not reflected in an ICD-10 diagnosis encoded in the medical records. In this context it is to be considered that the study aimed to explore the outcome with iv ketamine in a real-world clinical setting, without strict inclusion criteria commonly seen in randomised trails and that the diagnoses were based on clinical evaluations and not on structured diagnostic interviews, which could have provided a higher diagnostic accuracy. Furthermore, the prevalence of comorbidity is low. Other studies have shown a high degree of comorbidities in TRD populations receiving iv ketamine (Thomas et al., Reference Thomas, Baker, Lind and Dursun2018). In particular, the frequency of comorbid personality disorders is low in our study sample compared to what is to be expected (Tyrer et al., Reference Tyrer, Reed and Crawford2015). Taken together, this emphasises the importance of maintaining a meticulous diagnostic process in the clinical setting.
It has previously been noted that in a sample of TRD patients characterised by high chronicity, treatment resistance and chronic suicidal ideation, the established treatment regime with 0,5 mg/kg may not be sufficient to produce symptom improvement (Ionescu et al., Reference Ionescu, Bentley, Eikermann, Taylor, Akeju, Swee, Pavone, Petrie, Dording, Mischoulon, Alpert, Brown, Baer, Nock, Fava and Cusin2019). This finding could be applicable to some patients who could have received lower doses during treatment. Unfortunately, due to lack of data we were not able to estimate the number of patients with suboptimal dosage. Furthermore, both iv racemic ketamine and iv esketamine were used in this study sample. A 2021 meta-analysis comparing iv racemic ketamine to intranasal esketamine indicate that racemic ketamine could be superior to esketamine for treating depression (Bahji et al., Reference Bahji, Vazquez and Zarate2021). To our knowledge there is no evidence indicating that iv ketamine and iv esketamine can be used interchangeably when treating psychiatric illness. However, a 2016 multi-centre, randomised, placebo-controlled trial exploring the efficacy of iv esketamine on 30 patients with TRD observed an antidepressant effect (Singh et al., Reference Singh, Fedgchin, Daly, Xi, Melman, De Bruecker, Tadic, Sienaert, Wiegand, Manji, Drevets and Van Nueten2016). The interchangeable use of ketamine and esketamine might contribute to the observed lack of congruity regarding some of our findings compared to prior research.
The open-label nature of this kind of study comes inherently with biases. As is often the case with these kinds of clinical studies, the study sample was small and heterogeneous, did not include a control group and the patients were recruited from a single study site. These factors limit the generalisability of the study’s findings. Furthermore, it lacks the stringency in treatment protocol found in randomised clinical trials, evident in both the patient selection and treatment regime. The documentation in the medical records was occasionally incomplete or unclear. The study lacks specific data on adverse events and drop-out frequency, areas crucial to be able to correctly judge the risk–benefit profile of a given treatment. This is of particular importance in off-label treatment, especially when patient inclusion seemingly is expanded into diagnostic areas containing limited or no scientific evidence. The precise rationale for choosing iv ketamine over established treatments or whether patients were deemed resistant to ECT pretreatment, is not included in the study data. As ketamine was administered off-label without established treatment guidelines, no clear directives whether ECT was to be administered prior to ketamine were in place, further highlighting a problem with off-label use of ketamine in such a setting. Further management strategies post-treatment is not included in the study data.
As was noted previously, the comparative effectiveness of iv ketamine to ECT have been addressed in two previous studies (Ekstrand et al., Reference Ekstrand, Fattah, Persson, Cheng, Nordanskog, Akeson, Tingstrom, Lindstrom, Nordenskjold and Rad2022; Anand et al., Reference Anand, Mathew, Sanacora, Murrough, Goes, Altinay, Aloysi, Asghar-Ali, Barnett, Chang, Collins, Costi, Iqbal, Jha, Krishnan, Malone, Nikayin, Nissen, Ostroff, Reti, Wilkinson, Wolski and Hu2023). The results indicate ECT being superior to iv ketamine in hospitalised patients. As ketamine was mainly regarded as an off-label option for TRD during the years included in this study, we expected to find a population characterised by treatment resistance and chronicity with a high degree of hospitalisation. Instead, a large variation in the study sample was found. Given the sample characteristics, a more precise description of the treatment related considerations when initiating iv ketamine at that time would have been an asset. The comparable effectiveness of iv ketamine to treatments such as second-generation antipsychotics and combined antidepressants in TRD has neither been established (McIntyre et al., Reference McIntyre, Rosenblat, Nemeroff, Sanacora, Murrough, Berk, Brietzke, Dodd, Gorwood, Ho, Iosifescu, Lopez Jaramillo, Kasper, Kratiuk, Lee, Lee, Lui, Mansur, Papakostas, Subramaniapillai, Thase, Vieta, Young, Zarate and Stahl2021).
The open-label nature also suggests the possibility of a large placebo effect. A recent, triple-masked study, administrating sub-anaesthetic doses of iv ketamine to depressed patients undergoing anaesthesia for routine surgery showed no difference in antidepressant effect when comparing iv ketamine to saline-placebo (Lii et al., Reference Lii, Smith, Flohr, Okada, Nyongesa, Cianfichi, Hack, Schatzberg and Heifets2023). This suggests significant influence of expectancy bias and extra-pharmacological effects on treatment results. These findings might be relevant in explaining the results in this study showing a reduction in depressive symptoms but not in the need for inpatient care. However, a 2015 meta-analysis examining the effects of iv ketamine on major depressive disorder and bipolar disorder did not demonstrate a significant difference in effect size between open-label and participant-blind infusions (Coyle and Laws Reference Coyle and Laws2015). These inconsistencies further emphasise the need for additional research and clinical treatment guidelines.
The reinforcing and rewarding properties of ketamine together with its potential toxicity have brought concerns about potential misuse into focus (Liu et al., Reference Liu, Lin, Wu and Zhou2016). A 2022 review highlighted that the evidence is insufficient to dependably demonstrate the abuse liability of ketamine in depressed patients and that high quality RCTs have excluded patients with high abuse potential (Le et al., Reference Le, Cordero, Jawad, Swainson, Di Vincenzo, Jaberi, Phan, Lui, Ho, Rosenblat and McIntyre2022). Case reports have observed ketamine dependance in depressed patients when not treated in a proper setting (Bonnet, Reference Bonnet2015; Schak et al., Reference Schak, Vande Voort, Johnson, Kung, Leung, Rasmussen, Palmer and Frye2016). It has previously been acknowledged that evidence is lacking regarding dose optimisation, treatment frequency and long-term efficacy of ketamine treatment in TRD patients (McIntyre et al., Reference McIntyre, Rosenblat, Nemeroff, Sanacora, Murrough, Berk, Brietzke, Dodd, Gorwood, Ho, Iosifescu, Lopez Jaramillo, Kasper, Kratiuk, Lee, Lee, Lui, Mansur, Papakostas, Subramaniapillai, Thase, Vieta, Young, Zarate and Stahl2021). Altogether, this suggests a risk of treatment continuation based on factors other than treatment efficacy. These factors could have contributed to the administrative decision to terminate ketamine treatment at the clinic. Our study sample lack data on substance abuse, dissociation and euphoric effects during treatment and therefore cannot address this question further. To ensure that treatment continuation is based on increased patient functioning and to reduce potential abuse liability, treatment evaluation using objective measures, such as the need for hospitalisation, and not subjective experiences is prudent and a cautionary approach would be advisable.
The study highlights both treatment specific and extra-pharmacological factors as clinical challenges posed in a real-world study sample of patients found eligible for iv ketamine in psychiatric practice. Considering the multiple clinical questions waiting to be elucidated by further research the results in this study emphasise the importance of a determined treatment plan, governed by established guidelines and considering the diagnostic and social perspective as well as treatment specific factors. The study underlines the importance of maintaining a strict inclusion and treatment protocol with set evaluation criteria even in the clinical setting to facilitate the evaluation of treatment effects more precisely. Highlighting the difficulties associated with undefined inclusion and evaluation criteria in conjuncture with promising, but sometimes conflicting, scientific evidence. The main strength of this real-world study design is its potential for strong external validity in a clinical setting, which in part is due to the absence of exclusion criteria. However, the study results also highlight this possible strength as a limitation. Since initiating an off-label treatment without strict inclusion criteria might lead to suboptimal outcomes, as indicated by the lack of significant effect on our primary outcomes.
We believe that our primary outcome measures as such are suitable in conjunction with commonly used symptom rating scales in evaluation of treatment response in corresponding patient populations. Especially given the many clinical questions awaiting clarification, suggesting a cautionary approach. Otherwise, there is a risk of poorly focused treatments which might not provide the greatest benefit for the individual patient nor a best possible risk–benefit proposition.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/neu.2024.18.
Author contribution
Karl Sandström, MD, Senior Consultant Physician, Department of Clinical Sciences (Psychiatry), Umeå University, Umeå, Sweden.
Substantial contributions to conception and design, acquisition of data, analysis and interpretation of data.
Olli Kampman, MD, PhD, Professor, Senior Consultant Physician, Department of Clinical Sciences (Psychiatry), Umeå University, Umeå, Sweden.
Contributions to analysis and interpretation of data as well as drafting the article and revising it critically for important intellectual content.
Peter Asellus, MD, PhD, Associate Professor, Senior Consultant Physician, Department of Clinical Sciences (Psychiatry), Umeå University, Umeå, Sweden.
Contributions to conception and design and acquisition of data. Contributions to drafting the article and revising it critically for important intellectual content.
Financial support
The study was financed by grants from the Swedish state under the agreement between the Swedish government and the county councils, the ALF-agreement (RV-983148).
Competing interests
Peter Asellus has previously received consulting fees from Janssen Pharmaceutical, unrelated to this project.
Ethical statement
Ethical approval for this study was provided by The Swedish Ethical Review Authority.
Open access
