Reductions in anxiety and depression symptoms in a subset of outpatients with problematic substance use who received ketamine-assisted psychotherapy: a two-year retrospective chart review

The British Columbia Ministry of Health described a comprehensive paradigm of assessment and response to addictions and substance use in 2004, introducing the terminology of ‘problematic substance use.’ Problematic substance use (PSU) includes individuals who have potentially harmful substance use behaviors or patterns that are not clinical disorders (e.g., driving while impaired, using substances while pregnant) and individuals who have substance use disorders as defined by the Diagnostic and Statistical Manual of Mental Disorders (5th ed.; DSM-5) (1). This particular framework acknowledges that there is a spectrum to substance use instances or patterns, from ‘beneficial use’ to ‘non-problematic use’ to ‘problematic use.’ Substance use disorders are considered the extreme and most damaging end of the spectrum. For the purposes of this study, those with PSU include individuals with clinician-identified substance use disorder and individuals with self-identified or clinician-identified potentially harmful substance use. For the purposes of this manuscript we use ‘high risk’ and ‘problematic’ interchangeably in regard to the spectrum of substance use in this urban outpatient population.

Substance use disorder (SUD) is a prevalent and difficult-to-treat condition that is made more complicated by comorbid mood disorders and biopsychosocial barriers to healing. From 2020 to 2021, an estimated 46.3 million people (16.5 percent of the population) aged 12 or older were living with a diagnosed SUD in the United States. Of this group, 43.7 million (15.6 percent) needed SUD treatment but only 4.1 million people (1.5 percent) received treatment specifically for SUD (2).

In 2021, there were an estimated 26 million persons (9.3 percent of the population) aged 12 or older who had a major depressive episode (MDE) in the past year, for which 14.6 million (5.2 percent) received treatment. Among adults aged 18 or older, 57.8 million persons (22.8 percent) had one or more mental, behavioral, or emotional disorder (classified as any mental illness or AMI) in the past year. It was estimated that 18.3 percent of adolescents aged 12–17 (4.7 million people) and 18.8 percent of adults aged 18 or older (46.5 million people) received mental health services in a specialty setting (2).

Individuals with a SUD may also have other mental disorders and those with mental health disorders may also struggle with substance use (3). Moreover, anxiety, depression, and spiritual distress have been correlated with substance use, where its symptoms are thought to contribute to and result from the complex biopsychosocial process of addiction (4, 5).

Compared to their counterparts who did not have an MDE in the past year, adolescents aged 12–17 with a past year MDE were more likely to use some substances, in the past year or past month (2). In 2021, it was estimated that 935,000 individuals (25.2 percent) among this age group had both an MDE and SUD in the past year.

Mental disorders may contribute to substance use, as individuals with anxiety, depression, or post-traumatic stress disorder (PTSD) may use drugs or alcohol to self-medicate. Other mental health disorders comorbid with SUD can include attention-deficit hyperactivity disorder, bipolar disorder, personality disorders, and schizophrenia. Among adults aged 18 or older, it was estimated that 19.4 million persons (7.6 percent) had comorbidity of AMI and SUD (2).

Among persons under the age of 45, accidental drug overdose is the leading cause of death (6). In the United States, there were 106,699 individuals who died from a drug-involved overdose in 2021, with 70,601 deaths involving synthetic opioids other than methadone (primarily fentanyl). Cocaine-involved deaths rose nearly 54% from 15,883 deaths in 2019 to 24,486 deaths in 2021 (7). Excessive alcohol use contributed to more than 140,000 deaths in the United States each year from 2015 through 2019 (8).

Morbidity and mortality rates are high, biopsychosocial etiologies are complex, and there are mutually sustaining factors between PSU and even subclinical mental distress (3). The treatment of PSU therefore requires comprehensive and individualized treatment plans. The integrative care model at the AIMS Institute includes the principles of prevention, whole-person health (mind–body-spirit), and awareness of whole-systems influence on an individual’s health (9–12). Thus a patient’s mental distress is assessed and treated as one experience, on a spectrum of mortality and morbidity risk, as opposed to siloing one treatment per diagnosis. For example, in-person healthcare delivery for SUD was significantly impacted during the first year of the COVID-19 pandemic. This worsened access to the already limited treatment options for ‘dual diagnoses,’ defined as SUD concurrent with AMI such as depression or anxiety (13). Herein lies the need to explore adjunctive and alternative care models for those with high-risk substance use and comorbid mental distress.

Currently, medication-assisted treatment (MAT) is considered first-line for opioid use disorder. Medications are also approved by the U.S. Food and Drug Administration to treat alcohol use disorder [acamprosate, disulfiram and naltrexone], but overall MAT appears under-utilized in the United States for SUD (14). Depending on the severity of substance use, socioeconomic factors and comorbid mental distress, SUD treatment often involves a time-and resource-intensive combination of medically supervised detoxification and intensive psychotherapy, sometimes with additional group work. While Washington State compares favorably to other states on many health indicators, it has some of the highest rates of mental illness, yet also a severe shortage of mental health professionals including those who can prescribe appropriate medications, those who provide therapy, or providers or facilities that offer both (15).

Ketamine-assisted psychotherapy (KAP) utilizes sub-anesthetic doses of the dissociative compound ketamine, along with psychotherapy, to address psycho-emotional distress. This includes depression, anxiety, and spiritual distress, each of which can lead to, co-occur with, or result from substance use disorders. Interestingly, ketamine and acamprosate have in common the mechanism of glutamatergic NMDA receptor antagonism, warranting evaluation of ketamine’s role in SUD treatment alone, as well as treatment of comorbid mental health concerns (16). While still considered an experimental and off-label use of ketamine, KAP also offers several important novel features to the field of SUD treatment: neurobiologically, the promotion of BDNF (brain-derived neurotrophic factor) and neuroplasticity; a supportive set and setting or ‘container’ for mystical experiences and psychospiritual healing; the potential for sustained benefits with few administrations and low risk profile relative to standard of care; and low to no observed risk for habituation to ketamine (17–19).

Ketamine has demonstrated efficacy in prolonging abstinence among individuals with alcohol and heroin use disorders, including after just a single administration in the context of motivational psychotherapy (18–20). This study evaluates the efficacy of KAP in influencing various psychometric scores in individuals with PSU. Symptoms of anxiety were measured by the Generalized Anxiety Disorder-7 (GAD-7) questionnaire. Depression symptoms were measured by the Patient Health Questionnaire-9 (PHQ-9). The National Institute of Health-Healing Experience of All Life Stressors (NIH-HEALS) questionnaire was used to measure psychosocial and spiritual well being. To further contribute to understanding KAP’s utility, specifically for high risk substance use and harm reduction, participant data was compared to population of enrolled study participants who received KAP for any reason during the same time period. A 2022 systematic review of ketamine for mental health and SUD found only three studies of KAP, none of which covered multiple types of substance use or included those at high risk with a preventive approach to treating comorbid conditions (17). To our knowledge this is the only study of ketamine assisted psychotherapy for outpatients at high risk for mixed substance use disorders assessing their comorbid mental distress in comparison to those without PSU.

Participants received care at the AIMS (Advanced Integrative Medical Science) Institute, an urban community-based integrative medicine practice in Washington State. This study analyzed a subset of data extracted from two concurrent AIMS Institute prospective outcomes studies: the AIMS Medical Outcomes Study (AMOS) and the AIMS Cancer Outcomes Study (ACOS); (NCT04495790 and NCT04512755, respectively). Both are five-year prospective outcome studies of patient outcomes during and after receiving integrative care at the AIMS Institute (9, 10). Retrospective chart reviews were performed for adolescent and adult patients with established care at AIMS Institute who consented to take part in the AMOS or ACOS studies. Both of the studies have been approved by the Institutional Review Board at Seattle University in Seattle, Washington.

Inclusion criteria for the PSU subset were patients who (1) have an active or previous ICD-10 substance use diagnosis per DSM-5 criteria or problematic substance use as defined by the British Columbia Ministry of Health; (2) received at least one KAP experiential session between November 1, 2020-October 31, 2022 at the AIMS Institute; and (3) completed at least one baseline and follow-up psychometric (GAD-7, PHQ-9) or psychosocial and spiritual metrics (NIH-HEALS) questionnaire. Current and former AIMS Institute staff members or volunteer affiliates who received KAP were excluded from this study to reduce the effect of bias upon completing their psychometric assessments.

In this retrospective chart review, 24 months of data was manually sorted and abstracted from two electronic health record (EHR) systems for patients who received ketamine-assisted psychotherapy at the AIMS Institute in Seattle, Washington within the time frame of November 1, 2020 to October 31, 2022. One platform serves as the primary EHR, storing patient data, demographics, and associated diagnoses. A second EHR platform tracks various psychometric scores over time; this system regularly prompts patients to submit responses every 90 days to self-report questionnaires. Questionnaires utilized were the PHQ-9 for severity of depression symptoms, the GAD-7 for severity of anxiety symptoms, and the National Institute of Health – Healing Experiences from All Life Stressors scale (NIH-HEALS), a validated 35-item scale to assess the patient’s state of psychosocial and spiritual elements of health resilience and well being. Improvement is measured with decreased PHQ-9 and GAD-7 ratings and with increased NIH-HEALS ratings.

The following data was collected: basic demographics; dates of KAP sessions; type of ketamine treatment received; diagnosis of AMI, SUD, or conditions that may benefit from KAP (e.g., chronic pain, insomnia); type of substance used; psychiatric medications 1 month leading up to the first KAP session; pre-medications utilized on day of KAP; adverse side effects; and various questionnaire scores (ACE, Resilience, GAD-7, PHQ-9, NIH-HEALS). Due to the low volume of data of documented SUD, clinician recall was additionally used to identify patients with histories of SUD or active PSU of any severity.

Baseline GAD-7, PHQ-9, and NIH-HEALS scores were established using the most recent questionnaires that preceded a participant’s first KAP session. Changes were calculated by comparing the difference between the baseline questionnaire score and the last post-KAP questionnaire score completed. Statistical analyses were performed on paired data. In assessing the relationship between time elapsed since KAP (or prior to KAP) to psychometric outcomes, the number of days prior to or after KAP were calculated for each participant and correlated to GAD-7-PHQ-9, and NIH-HEALS scores.

The participant results were compared to the entire study-enrolled population of KAP recipients at AIMS during the same two-year time period. In this comparison, the participant group is termed the participant group is termed PSU Subpopulation.

Microsoft Excel software was used for statistical analysis. A majority of the data was analyzed utilizing descriptive statistics (counts, percentages, means, and standard deviation). Due to ssmall sample size of the PSU Subpopulation (n = 18), Wilcoxon rank-sum tests were performed between baseline and final metric scores with an established significance level of ɑ < 0.05. For the larger KAP population, paired t-tests examined differences between baseline and final metric scores with an established significance level of p < 0.05. An independent samples t-test compared average changes between the PSU Subpopulation and KAP Population. Linear regression analysis was performed on psychometric scores. The correlation between receiving KAP and psychometric score changes was analyzed with Pearson correlation analysis.

Participants ranged in age from 16 to 67 years. The mean participant age at the time of their first ketamine session was 39.56 years. A majority of participants were in their 30s (n = 7; mode = 48 years; median = 38 years). Participants identified their gender as follows: female (n = 11), male (n = 6) and genderqueer (n = 1).

Within the context of problematic substance use, a majority of participants used alcohol (n = 10, 55.56%), followed by nicotine (n = 5, 27.78%), and opioids (n = 2, 11.11%). Other substances used include benzodiazepines, cocaine, heroin, zolpidem, and an unknown psychedelic substance (each with n = 1, 5.56%). Most participants reported using one substance (n = 15, 83.3%) to their AIMS provider. Two participants (11.11%) were identified to have used two substances. One participant (5.56%) was identified to have used three substances.

The average Adverse Childhood Experiences (ACE) and Resilience scores were 3.89 ± 2.47 and 10.17 ± 2.96, respectively. The number of KAP experiential sessions attended by each participant ranged from 1 to 6 sessions, with a median of 3 sessions. Table 1 outlines each participant’s characteristics.

At the time of the first KAP treatment, participants ranged from two to eight comorbid conditions with a mode of four conditions. The most common comorbid conditions were a type of depression (n = 14, 77.8%), a type of anxiety (n = 10, 55.6%), and post-traumatic stress disorder (n = 10, 55.6%). Conditions such as insomnia, chronic pain, traumatic brain injury, seizure disorders, and migraines were also noted. Table 2 outlines the PSU population’s comorbid conditions.

The most common reported adverse reactions to ketamine were nausea and dizziness (each at n = 4, 22%), followed by fatigue (n = 2, 11%). More uncommon adverse effects were vomiting, dry mouth, and slower movements (each at n = 1, 5.6%).

Average baseline score, last completed score, change in scores, and comparison of average changes for corresponding questionnaires and populations are outlined in Table 4. The scores for the problematic substance use participant population (named ‘PSU Subpopulation’) can be compared against the score for the KAP population. Figures 1A,B illustrate the average score changes when compared to baseline for both the participant population and the KAP population.

Participants who completed at least one post-KAP GAD-7 questionnaire (n=14, 77.78%) had baseline GAD-7 scores ranging from 1 to 21 points. After at least one KAP experiential session, scores ranged from 0 to 18 points. The score range for GAD-7 is 0–21 with higher scores indicating worsening severity of anxiety symptoms. The Wilcoxon rank-sum test calculated U = 42. This rejects the null hypothesis (ɑ = 0.05 of 55). The enrolled KAP population paired t-test result for GAD-7 scores was statistically significant at p = 6.64 x 10⁻⁵ (n = 104).

There were 88.89% (n = 16) of participants who completed at least one post-KAP PHQ-9 questionnaire. Baseline PHQ-9 scores ranged from 4 to 23 points. After at least one KAP experiential session, scores ranged from 0 to 17 points. The score range for PHQ-9 is 0–27 with higher scores indicating worsening severity of depression symptoms. The Wilcoxon rank-sum test calculated U = 52, rejecting the null hypothesis (ɑ = 0.05 of 75). The enrolled KAP population paired t-test result for PHQ-9 scores was statistically significant at p = 6.25 x 10⁻⁷ (n = 111).

There were 83.33% (n = 15) of participants who completed at least one post-KAP NIH-HEALS questionnaire. Baseline NIH-HEALS scores ranged from 94 to 146 points. After at least one KAP session, scores ranged from 79 to 146 points. The score range for NIH-HEALS is 35–175, with higher scores indicating greater positive transformation in response to challenging life events. The Wilcoxon rank-sum test calculated U = 139. This did not reject the null hypothesis (ɑ = 0.05 of 64). The enrolled KAP population paired t-test result for NIH-HEALS scores was statistically significant at p = 0.037 (n = 85).

Participants’ psychometric scores and calculated changes between the baseline score and the last post-KAP questionnaire score completed are described in Table 5. Figures 2A–C illustrate respective psychometric scores plotted by the number of days prior to or after a KAP session. In terms of correlations, the models illustrate that there is partial predictability to outcomes with medium effects on post-KAP psychometric scores. The R² values for GAD-7, PHQ-9, and NIH-HEALS were calculated as 0.098, 0.143, and 0.106, respectively.

A majority of participants with PSU had comorbid anxiety (n = 10) and/or depression (n = 12). Overall, data showed improvements to psychometric scores after having at least one KAP session. The average reductions in GAD-7 scores by 6.71 ± 9.15 points and PHQ-9 scores by 7.44 ± 5.42 points from baseline are statistically significant. Additionally, Pearson’s correlation indicates that there is practical significance of having at least one KAP session influence psychometric scores with medium effects (GAD-7: R² = 0.098; PHQ-9: R² = 0.143). GAD-7 scores show reductions to anxiety symptoms, where the average baseline scores in the moderate range (10–14) shifted to the mild range (5–9) post-KAP. The same trend was seen with depression symptoms on the PHQ-9, where the average baseline scores changed from the moderate range (10–14) to the mild range (5–9) post-KAP.

All participants who completed follow-up psychometric testing demonstrated improvement in at least one domain. For example, when referring to Table 5, participants #1, #12, and #14 had increases to their GAD-7 scores, implying worsened anxiety symptoms. However, these same participants also had decreases in PHQ-9 scores, implying improved depression symptoms.

Participants experienced more improvements to psychometric scores when they also completed more KAP sessions in a shorter time interval. One participant (#11) who completed the greatest number of in-office KAP sessions during the study time period (6 sessions), also participated in a KAP session at the most frequent rate: every other week. This participant had the highest improvements to GAD-7 and PHQ-9 scores. These outcomes corroborate a 2022 systematic literature review, which found that higher doses of ketamine, more frequent KAP sessions, and longer durations of psychotherapy increase the efficacy and durability of improvements in conditions such as depression, anxiety, and SUD (25). However, there are participants who experienced benefit after fewer KAP sessions. Two participants (#3, #9) reported significant reductions in anxiety and depression after only one session and one participant’s (#14) depression score significantly decreased after two sessions (Tables 1, 5).

Little is known about ketamine’s long-term efficacy in addressing anxiety and depression(21, 23). Trends in PHQ-9, GAD-7 and NIH-HEALS show sustained improvement since the most recent ketamine administration. Psychometric responses were gathered by self-report. As such, there is a wide variety of time intervals between KAP sessions and repeat questionnaire completion (Figures 2A–C). Self-reports up to 269 days (8.8 months) after KAP, with an average of 59 days between responses, indicate that the statistically significant changes to anxiety and depression scores are not attributable to only the acute or subacute effects of ketamine, and may represent a durability of effect or new mutually sustaining protective features in a population with PSU. Given the months-years of development and complex etiology of PSU, investigating appropriate interventions with persisting effects without repeated administrations is valuable (26). Outcome intervals in integrative medicine research frameworks are often months long to reflect durable change in foundations of health, with the Bravewell Collaborative’s National Integrative Medicine Database (PRIMIER) collection of self-report data at three 2 month intervals, followed by three 6-month intervals (27). The data captured in our subset analysis, while the sample size and R² values are modest, are preliminary suggestions of durability of benefit in depression, anxiety and psycho-spiritual health for those with PSU. Further commentary on the trend of improved psychometric measures as a patient’s ketamine experience was approaching is provided in Supplementary material (6.3.1).

A limitation of this study includes possible non-response bias since data collected is dependent on patients consistently submitting self-report measures. The standard practice at the AIMS Institute is to regularly collect PHQ-9, GAD-7, and NIH-HEALS every 90 days after establishing baseline; however, these psychometric survey responses vary widely in their proximity to KAP experiential sessions (and integration visits). Minimizing non-response and regularizing psychometric intervals was attempted by sending out automated questionnaire reminders via the secondary EHR platform, collecting psychometrics through the primary EHR platform, and handing out paper questionnaires where necessary. Despite this, response rates remain low (50–56%). Participants who chose not to repeat psychometric scores may have done so for a number of reasons, including a poor or neutral outcome. This also introduces selection bias to the pool of available results.

Another limitation is that there is no consistent data collected to definitively address whether KAP is effective at reducing the frequency of PSU. As data was collected retrospectively for those with a history of high risk substance use, we did not consistently collect substance use data prior to and after KAP. This study did not include a matched control group, nor are we able to control for other various interventions for psycho-emotional health utilized by participants during the study time period, given the outpatient clinical context of the study. Though there are statistically relevant findings demonstrating improvements to anxiety and depression scores with KAP to a medium effect, a small sample size limits the power of results reported here. Variations in the number of KAP sessions patients received between baseline and psychometric follow up measurements limit our observations regarding KAP for PSU.

With a greater sample size, future research could assess differences among high-risk substance use in those with and without clinical depression or anxiety and provide a matched-control by including those with PSU who receive integrative care without KAP.

Ketamine-assisted psychotherapy should be considered early in the development of an integrative and individualized treatment plan for patients with known comorbid PSU of any kind and symptoms of anxiety or depression. Thorough screening for symptoms of anxiety and depression should be performed for any patient-or clinician-identified problematic substance use. Practices for including KAP in an integrative model could include thorough assessment of comorbid PTSD and the presence or history of suicidality as these symptoms could influence the psychosocial and spiritual effects of KAP.

Additional strategies may be necessary for increasing the rate of follow-up psychometric reporting for outpatient research studies in integrative mental health. For this clinic, only around 50–56% of the participant population completed at least one repeat psychometric questionnaire. Barriers to collecting this data may exist in the clerical systems and research protocols at this institute and in other urban outpatient community-based clinics. Additional patient reminders during post-ketamine care (including integration sessions) or including clinician-administered psychometrics may close some gaps in data collection.