Digital Enablement of Psychedelic-Assisted Therapy in Non-Clinical Settings: A Systematic Review of Safety, Efficacy, and Implementation Models

Abstract

Psychedelic-assisted therapy offers rapid and profound benefits for treatment-resistant psychiatric conditions but remains constrained by the need for intensive, clinic-based administration. Concurrently, advances in digital health technologies have introduced scalable tools. This systematic review evaluates the safety, efficacy, and implementation of digitally enabled psychedelic-assisted therapy delivered in non-clinical settings. A comprehensive search of five databases, registered in PROSPERO (CRD420251020968) and conducted in accordance with PRISMA guidelines, identified six eligible studies including real-world analyses, clinical trials, qualitative research, and case reports, representing a total of 12,731 participants. Most studies examined at-home ketamine or esketamine therapy supported by telehealth platforms or mobile applications. Data were synthesized narratively given the heterogeneity of designs and outcomes. Digital enablement was associated with high response rates (ranging from 56.4% to 62.8% for depression) and rapid symptom improvement, particularly in depression and anxiety. Remote monitoring and digital tools demonstrated feasibility and acceptability, but serious safety concerns—including psychiatric adverse events and one unintentional overdose—underscore the need for strict oversight. Risk of bias was moderate to serious across non-randomized studies, limiting confidence in the findings. One study on virtual ayahuasca rituals highlighted the sociocultural potential and limitations of online practices. Despite promising preliminary findings, the field is marked by low methodological rigor and absence of controlled trials. Digitally supported at-home psychedelic therapy represents a transformative but high-stakes frontier, requiring robust research and safeguards to ensure safe, equitable, and effective implementation. No funding was received for this review, and the authors declare no conflicts of interest.

1. Introduction

The last two decades have seen a significant resurgence of clinical and scientific interest in psychedelic-assisted therapy for a range of psychiatric conditions, including depression, post-traumatic stress disorder (PTSD), anxiety, and substance use disorders [1,2]. Compounds such as psilocybin, ketamine, and MDMA have demonstrated the potential to produce rapid and profound therapeutic effects, positioning them as a promising new paradigm in psychiatric care [3]. The conventional model for administering these therapies, however, is resource-intensive, typically requiring multiple sessions conducted in a specialized clinical setting under the direct supervision of trained healthcare professionals [4,5]. This model presents considerable barriers to scalability and access, particularly for individuals in rural or underserved communities, and creates significant cost and time commitments for patients [6].

In parallel with the psychedelic renaissance, the field of digital health has expanded dramatically, with telemedicine and remote monitoring technologies becoming integral components of modern healthcare delivery [7,8]. Digital tools, including mobile applications, wearable sensors, and secure video conferencing platforms, now offer novel mechanisms for patient support, data collection, and remote supervision in mental healthcare [9,10].

This systematic review addresses a critical and underexplored intersection between these two advancing fields: the use of digital technologies to enable and support the administration of psychedelic-assisted therapy in non-clinical settings, such as a patient’s home. As interest grows in more accessible treatment models, it is crucial to systematically evaluate the existing evidence regarding the safety, efficacy, and implementation of these digitally enabled, at-home approaches [11].

The objective of this review is to systematically identify, evaluate, and synthesize the evidence on the use of digital technologies to support the administration of psychedelic-assisted therapy in non-clinical settings. Furthermore, we will synthesize the available evidence on the safety and efficacy of these digitally supported interventions, while also assessing their reported feasibility and acceptability for both patients and providers. By addressing these key areas, this review aims to inform best practices, identify critical gaps in the current research landscape, and support the future development of guidelines for the safe and effective digital enablement of psychedelic therapy.

2. Methods

This systematic review was conducted and reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 statement [12]. The review protocol was registered prospectively in the International Prospective Register of Systematic Reviews (PROSPERO) on 28 March 2025 (Registration No. CRD420251020968).

2.1. Eligibility Criteria

Studies were included if they met the following criteria: (1) Population: Adult participants (≥18 years) receiving psychedelic-assisted therapy; (2) Intervention: Use of a psychedelic compound (e.g., psilocybin, ketamine, esketamine, MDMA, LSD, ayahuasca) where the administration was supported by a digital technology (e.g., telehealth platform, mobile application, remote monitoring, wearable sensors); (3) Comparator: Studies with any comparator (e.g., standard in-clinic administration) or no comparator were included; (4) Study Design: Both randomized and non-randomized study designs were eligible for inclusion; (5) Context: The intervention was administered in a non-clinical setting, such as a participant’s home, a retreat, or another community-based environment. Studies conducted exclusively in clinical, hospital, or inpatient settings without a digital support component were excluded. Only studies with the full text available in English were included.

2.2. Information Sources and Search Strategy

A systematic search was conducted across the following electronic databases: MEDLINE (via PubMed), PsycINFO, Scopus, Web of Science, and ClinicalTrials.gov. The searches were run from database inception to 15 April 2025. To supplement the database search, reference lists of included articles and relevant reviews were manually scanned to identify additional studies.

The search strategy combined keywords related to psychedelic compounds and digital health technologies. The full search strategy can be found in Supplementary Materials. The following search string is representative of the strategy used for PubMed:

(“psychedelic assisted therapy” OR psilocybin OR ketamine OR esketamine OR MDMA OR LSD OR DMT OR ayahuasca OR mescaline OR hallucinogen) AND (“digital health” OR “digital technology” OR telemedicine OR telehealth OR “remote monitoring” OR “mobile health” OR mHealth OR “wearable electronic devices” OR “mobile app”).

2.3. Study Selection

All records identified through the database searches were imported into Covidence, a systematic review management software, for duplicate removal and screening. Two reviewers (S.L., B.D.) independently screened the titles and abstracts of all unique records against the pre-defined eligibility criteria. The full text of any potentially relevant article was then retrieved and assessed for eligibility by both reviewers independently. Any disagreements at either the title/abstract or full-text screening stage were resolved through discussion and consensus. No formal inter-rater reliability statistic was calculated.

2.4. Data Collection Process

A standardized data extraction form was developed and used to extract relevant information from each included study. Two reviewers independently extracted the data. Any discrepancies in the extracted data were resolved by returning to the source article and reaching a consensus.

2.5. Data Items

The following data were extracted from each included study: author(s), year of publication, title, study design, sample size, participant characteristics, intervention details (psychedelic compound, dose, route of administration, and type of digital technology used), primary and secondary outcomes, efficacy data, and author-reported safety concerns, feasibility, and acceptability.

2.6. Risk of Bias Assessment

The methodological quality of the included studies was independently assessed by two reviewers. The Risk Of Bias In Non-Randomized Studies of Interventions (ROBINS-I) tool was used to appraise non-randomized studies [13]. The quality of qualitative studies was assessed using the Critical Appraisal Skills Programme (CASP) Qualitative Checklist [14]. Case reports were evaluated using a general quality assessment checklist focused on the clarity and completeness of the reporting. Disagreements in the risk of bias assessments were resolved through discussion to reach a consensus. The results of these assessments were used to inform the narrative synthesis and the discussion of the overall limitations of the evidence base. Although formal statistical assessments of reporting bias were not feasible given the heterogeneity and small number of studies, this possibility was incorporated into the interpretation of the strength and generalizability of the findings. The overall certainty of the evidence for each outcome was considered qualitatively, taking into account study design, risk of bias, consistency of findings, and directness of the evidence.

2.7. Synthesis of Results

Given the heterogeneity in study designs, interventions, and outcome measures, we first tabulated the characteristics of each included study and compared them against the planned intervention and outcome groups to determine eligibility for each synthesis. Where relevant data were incomplete, we extracted all available summary statistics directly from the reports without imputation or conversion. Results of individual studies were displayed in tabular format to allow comparison of intervention characteristics, sample sizes, and outcome domains, and were further described narratively in the text. Because of substantial variability in populations, interventions, and outcome measures, a quantitative meta-analysis was not appropriate. Instead, we conducted a structured narrative synthesis, grouping findings by intervention type (e.g., at-home ketamine via telehealth, online ayahuasca rituals) and by outcome domain (efficacy, safety, feasibility, and acceptability). Sources of heterogeneity across studies were qualitatively explored by comparing study settings, delivery formats, and population characteristics. Sensitivity analyses were not performed given the absence of quantitative pooling.

3. Results

3.1. Study Selection

The systematic search of five electronic databases yielded 99 records. After removing 38 duplicate records, 61 unique titles and abstracts were screened for eligibility. Of these, 50 were excluded as they did not meet the inclusion criteria. The full texts of the remaining 11 articles were retrieved and assessed for eligibility. Five of these full-text articles were excluded for the following reasons: the study was still ongoing (n = 4) or the intervention was conducted in the wrong setting (n = 1). Ultimately, 6 studies met all eligibility criteria and were included in the narrative synthesis. The PRISMA 2020 flow diagram detailing the study selection process is presented in Figure 1 [12].

3.2. Study Characteristics

The 6 included studies were published between 2020 and 2024. The studies were highly heterogeneous in design and scope, comprising two large-scale longitudinal analyses of real-world data [15,16], one multi-center open-label trial [17], one qualitative study [18], and two case reports [19,20]. Sample sizes ranged from a single patient to 11,441 participants. All studies were conducted without a control group.

The primary psychedelic compound investigated was ketamine or its enantiomer esketamine (5 studies), with one study examining ayahuasca. The digital technologies employed included comprehensive telehealth platforms offering video consultations and remote monitoring [15,16], a mobile application for daily self-monitoring [17], remote supervision via telemedicine [20], and public video streaming platforms for group rituals [18]. The characteristics of each included study are summarized in

Table 1. Characteristics of Included Studies.

Study	Study Design	Population (N, Condition)	Intervention Details	Digital Integration	Key Outcomes Measured
[18]	Qualitative Study	N = 12 Santo Daime members	Ayahuasca administered during online group rituals.	Public video streaming platforms (Zoom, YouTube, Facebook Live).	Qualitative experiences, challenges, and benefits of online rituals.
[16]	Prospective, open-label effectiveness trial	N = 1247 Adults with moderate-to-severe depression and anxiety	At-home sublingual ketamine (300–450 mg) over 4 weeks.	Comprehensive telehealth platform (Mindbloom) with video consults, remote monitoring, and behavioral coaching.	Depression (PHQ-9), Anxiety (GAD-7), and Adverse Events.
[19]	Case Report	N = 1 Adult with PTSD	At-home sublingual ketamine (1200 mg).	Telehealth platform for prescribing and instruction.	Clinical presentation of an unintentional overdose.
[17]	Multi-center, open-label, single-arm study	N = 29 Adults with Treatment-Resistant Depression (TRD)	At-home esketamine nasal spray (starting at 56 mg) twice weekly.	“EsCARe” mobile application for daily self-monitoring of symptoms.	Depression (PHQ-9, HAMD), Anxiety (GAD-7), Adverse Events, and App Adherence.
[20]	Case Report	N = 1 Adult with TRD	At-home intranasal ketamine (dose not specified).	Telemedicine for remote supervision by a registered nurse.	Feasibility and safety of remote supervision.
[15]	Longitudinal, real-world data analysis	N = 11,441 Adults with moderate-to-severe depression	At-home sublingual ketamine (avg. dose 590 mg) over 4 weeks.	Comprehensive telehealth platform (Mindbloom) with video consults, remote monitoring, and behavioral coaching.	Depression (PHQ-9), Anxiety (GAD-7), and Adverse Events.

.

Across the included studies, the digitally supported interventions varied considerably in frequency, provider involvement, and purpose. In the large-scale Mindbloom studies [15,16], patients completed an initial psychiatric evaluation by a licensed prescriber via video call, followed by structured coaching and periodic video check-ins over a four-week course of sublingual ketamine. These sessions were supplemented by asynchronous digital monitoring and behavioral integration exercises. In Kim et al. (2024) [17], esketamine nasal spray was paired with daily symptom tracking and adherence prompts through the EsCARe mobile application, with prescribers available via telehealth as needed. Longpré-Poirier et al. (2020) [20] employed remote supervision by a registered nurse through real-time telemedicine during at-home intranasal ketamine administration, demonstrating that live monitoring was feasible in a non-clinical setting. By contrast, Hartogsohn (2022) [18] described online ayahuasca rituals conducted through public streaming platforms such as Zoom and YouTube, where digital tools primarily facilitated community cohesion rather than clinical oversight. Together, these findings suggest that digital integration ranged from structured clinical monitoring to purely social facilitation, with video calls most often used for prescribing, safety monitoring, and integration support, and mobile applications for ongoing adherence and symptom tracking.

3.3. Risk of Bias in Included Studies

The methodological quality of the six included studies was assessed using tools appropriate for their respective designs. Overall, the evidence base is characterized by a moderate risk of bias, primarily due to the non-randomized, uncontrolled nature of all included quantitative studies. A summary of the risk of bias assessment for each study is presented below and in

Table 2. Summary of Risk of Bias Assessment for Included Studies.

Study	Study Design	Tool Used	Overall Judgment	Key Sources of Potential Bias
[18]	Qualitative Study	CASP Checklist	High Quality	Lack of researcher reflexivity statement; unclear data analysis process.
[16]	Non-Randomized Study	ROBINS-I	Serious Risk	No control group (confounding); self-selected paying sample (selection bias); >55% missing follow-up data (attrition bias).
[19]	Case Report	Quality Checklist	High Quality	Not applicable (case report); findings are not generalizable.
[17]	Non-Randomized Study	ROBINS-I	Moderate to Serious Risk	No control group (confounding); unstandardized use of concurrent medications (deviation from intervention).
[20]	Case Report	Quality Checklist	Moderate Quality	Dose of ketamine not specified; subjective outcome measure.
[15]	Non-Randomized Study	ROBINS-I	Moderate Risk	Retrospective design (risk of bias from deviations); unclear handling of missing data.

.

3.4. Non-Randomized Studies

The three non-randomized studies were assessed using the ROBINS-I V2 tool.

The studies by Hull et al. and Mathai et al. [15,16] were both judged to be at a serious risk of bias. Both studies shared significant methodological limitations inherent to their design as large-scale analyses of real-world data from a single telehealth provider. The primary sources of bias were the absence of a control group, leading to a risk of confounding, and the use of a self-selected, paying patient population, which introduces a moderate to serious risk of selection bias. Most critically, both studies suffered from a high rate of missing follow-up data, with over 55% attrition in the Hull et al. (2022) study [16], posing a critical risk of bias in that domain. While the interventions were well-defined and validated outcome measures were used (low risk), the lack of pre-registered protocols also created a moderate risk of bias from selective reporting.

The study by Kim et al. (2024) [17] was judged to be at a moderate to serious risk of bias. While the study had low risk of bias in participant selection, intervention classification, missing data, and outcome measurement, it had significant limitations. The lack of a control group introduced a moderate risk of confounding. Most notably, the study was judged to be at a serious risk of bias due to deviations from intended interventions, as the concurrent use of various antidepressants and other psychotropic medications was common and not standardized across the small sample.

3.5. Qualitative Study

The qualitative study by Hartogsohn (2022) [18] was appraised using the CASP Qualitative Checklist and was found to be of high quality. The study presented a clear research aim that was appropriately addressed with a qualitative methodology and a suitable recruitment strategy. The findings were clearly stated and well-supported by data. The primary limitations noted were a lack of a reflexivity statement from the researcher and an unclear description of the data analysis process, which prevented a full assessment of its rigor.

3.6. Case Reports

The two case reports were evaluated using a general quality assessment checklist.

The report by Johnson et al. (2024) [19] was judged to be of high quality. It was exceptionally clear, provided objective clinical data to support its claims, and presented an unambiguous learning point regarding the safety risks of unregulated at-home ketamine therapy.

The report by Longpré-Poirier et al. (2020) [20] was judged to be of moderate quality. While it presented a clear and novel solution to a clinical problem during the COVID-19 pandemic, it lacked key details, most notably the specific dose of intranasal ketamine administered, and the subjective “positive experience” outcome was not supported by any standardized measures.

3.7. Synthesis of Results

The findings are organized by the type of intervention and explore the key outcome domains of efficacy, safety, and feasibility/acceptability.

3.8. At-Home Ketamine Therapy via Telehealth Platforms

Two large longitudinal studies [15,16] investigated at-home, sublingual ketamine therapy for depression and anxiety using the same telehealth platform (Mindbloom).

• Efficacy: Both studies reported significant and rapid reductions in symptoms. Hull et al. (2022) [16], in a sample of 1247 patients, found that 62.8% achieved a response (≥50% reduction) on the PHQ-9 for depression, with a large effect size (d = 1.61). Similarly, 62.9% responded on the GAD-7 for anxiety (d = 1.56). Mathai et al. (2024) [15] replicated these findings in a much larger cohort of 11,441 patients, reporting a 56.4% response rate and a 28.1% remission rate for depression.
• Safety: The safety profile was generally favorable, with a low overall rate of adverse events (AEs) reported (4.7–4.8%). However, both studies noted the occurrence of serious adverse events (SAEs), which were exclusively psychiatric in nature. Mathai et al. [15] reported six SAEs, including psychosis and suicidal behavior, and 46 patients (0.4%) discontinued treatment due to AEs.
• Feasibility/Acceptability: The large-scale implementation and high treatment completion rates reported in both studies suggest that this model is a feasible and acceptable method for increasing access to care.

3.9. At-Home (Es)ketamine with Mobile or Telemedicine Support

Three studies examined the use of intranasal ketamine or esketamine in a non-clinical setting.

• Efficacy: Kim et al. (2024) [17] found that esketamine nasal spray produced a significant reduction in depressive symptoms (mean PHQ-9 score decreased from 19.69 to 14.14; p < 0.001) in 29 patients with TRD. Notably, their mobile app detected a statistically significant improvement in depressive symptoms just one day after the first dose (p = 0.049). The case report by Longpré-Poirier et al. (2020) [20] did not measure efficacy quantitatively but noted a “positive experience” for a 61-year-old patient receiving remotely supervised intranasal ketamine.
• Safety: The safety findings were mixed. The case report by Longpré-Poirier et al. (2020) [20] reported no AEs. In contrast, Kim et al. (2024) [17] reported one death by suicide and one instance of self-harm, though both were attributed to the exacerbation of pre-existing symptoms rather than the drug itself. The case report by Johnson et al. (2024) [19] detailed the most severe safety event in this review: a life-threatening, unintentional overdose of 1200 mg of sublingual ketamine at home, which resulted in hospitalization.
• Feasibility/Acceptability: These studies suggest the model is feasible. Kim et al. (2024) [17] reported high patient adherence (89.6% in the first 3 days) to the daily monitoring app. Longpré-Poirier et al. (2020) [20] concluded that remote supervision was a feasible strategy to continue care during the COVID-19 pandemic. However, the overdose reported by Johnson et al. (2024) [19] raises critical concerns about the feasibility of ensuring patient safety without robust regulation and clear provider instructions.

3.10. Online Ayahuasca Rituals

One qualitative study [18] explored the experiences of 12 Santo Daime members participating in ayahuasca rituals online via platforms like Zoom.

• Efficacy: Efficacy was not measured clinically. The study found that the online format fulfilled an important social function by allowing the community to maintain its practice and foster a sense of global connection during a time of social isolation.
• Safety: The primary concerns were contextual rather than medical, including a higher potential for distractions, social anxiety from being on camera, and an “in-built tension between the social and spiritual dimensions of ritual.”
• Feasibility/Acceptability: The online format was a feasible ad-hoc alternative during the pandemic. However, acceptability was mixed. While it provided a crucial social lifeline, participants also reported an “impoverished ritual experience” and expressed concerns about the commodification of the sacred ritual.

3.11. Certainty and Bias from Missing Data

Across the included studies, the certainty of the evidence was generally low to moderate. This was primarily due to limitations in study design, small sample sizes, and inconsistent reporting of outcomes. Several studies lacked complete outcome data or provided only partial results, raising the possibility of bias due to missing information. In particular, selective reporting of favorable findings and the absence of long-term follow-up data limited confidence in the overall conclusions. Because of these issues, the strength of the evidence for efficacy outcomes was judged as tentative, while evidence on safety and feasibility outcomes was more consistent but still constrained by incomplete reporting.

Across the (es)ketamine telehealth studies [15,16,17], pooled descriptive data suggest response rates for depression between 56–63%, with adverse events occurring in 4–5% of participants. Although exploratory and not derived from a formal meta-analysis, these findings indicate a consistent clinical effect and a relatively low but notable rate of adverse outcomes.

4. Discussion

4.1. Summary of Principal Findings

This systematic review is the first to synthesize the evidence on the digital enablement of psychedelic-assisted therapy in non-clinical settings. Our findings reveal a rapidly emerging therapeutic model that holds significant promise for expanding access to care but is accompanied by critical safety and implementation challenges. The included studies, predominantly focused on at-home ketamine and esketamine, consistently demonstrate rapid and substantial reductions in depressive and anxiety symptoms. This approach appears highly feasible, with evidence of large-scale implementation and high patient adherence to digital monitoring tools. However, this promise is tempered by significant safety concerns, including the occurrence of serious psychiatric adverse events and a reported case of a life-threatening overdose in an at-home setting. Furthermore, a qualitative exploration of online ayahuasca rituals suggests that while digital platforms can fulfill a crucial social function, they may fail to replicate the immersive, sensory richness of in-person spiritual practice.

4.2. Interpretation of Findings

The efficacy data from the large-scale, real-world studies of at-home sublingual ketamine are a key finding of this review. The high response rates (>50%) and large effect sizes reported are broadly consistent with the established literature on in-clinic intravenous ketamine for treatment-resistant depression [15,16]. The study by Kim et al. (2024) [17] further reinforces this, demonstrating a statistically significant improvement in depressive symptoms just one day after the first dose of esketamine. These findings suggest that the rapid antidepressant effects of (es)ketamine can be successfully translated from a supervised clinical setting to a remotely monitored, at-home model. This has profound implications for scalability, potentially allowing this potent treatment to reach a much broader patient population. The pooled descriptive findings reinforce the consistency of effects observed across studies, despite methodological limitations.

Beyond its role as an outlier, the ayahuasca ritual study [18] highlights important sociocultural dimensions of digital enablement. While online rituals facilitated community connection and continuity of practice during the pandemic, participants also reported an “impoverished ritual experience” and concerns that digital migration risks commodifying sacred traditions. These findings underscore the ethical complexity of applying commercial digital infrastructures to practices rooted in cultural and spiritual contexts.

However, the interpretation of these efficacy findings must be cautious. All included studies lacked a control group, making it impossible to disentangle the pharmacological effects of the psychedelic compound from the significant non-specific effects of the intervention, such as patient expectancy, the supportive context of coaching, and the novelty of the treatment. The participants in these studies were also self-selected and often paid out-of-pocket, likely representing a more motivated and less complex patient population than might be seen in public health systems.

Notably, both of the largest studies [15,16], which together represented the majority of participants in this review, were conducted exclusively through the Mindbloom telehealth platform. While these large real-world datasets provide valuable insights, reliance on a single commercial provider introduces potential biases related to patient selection, treatment protocols, and organizational practices. As such, the generalizability of these findings to other digital platforms or healthcare contexts is limited.

The most critical findings of this review relate to safety. While the overall incidence of adverse events in the large ketamine studies was low, the nature of the reported serious adverse events—including psychosis, suicidal behavior, and one death by suicide—is alarming. These events underscore that even with remote monitoring, the profound psychological effects of psychedelics carry inherent risks. The case report by Johnson et al. (2024) [19] of an unintentional overdose highlights the most extreme and preventable of these risks. This incident demonstrates a critical vulnerability in the at-home model: the potential for patient or provider error in a setting without immediate medical support. Together, these findings create a clear tension between the therapeutic promise of expanded access and the ethical imperative to ensure patient safety.

4.3. Limitations of the Evidence

The findings of this review must be interpreted in the context of several significant limitations: the absence of randomized controlled trials, the reliance on uncontrolled study designs, high attrition rates in some datasets, and the concentration of data from a single commercial provider. These issues collectively reduce confidence in efficacy estimates and limit the generalizability of findings.

Our formal risk of bias assessment confirms these concerns. The three non-randomized quantitative studies were all judged to be at a moderate to serious risk of bias. The studies by Hull et al. (2022) and Mathai et al. (2024) [15,16], while large, were particularly susceptible to bias from the lack of a control group (confounding), self-selected paying patient populations (selection bias), and high or unclear rates of missing data. One study [16] was judged to be at a serious risk of bias due to an attrition rate exceeding 55%. Similarly, the study by Kim et al. (2024) [17] was at a moderate to serious risk of bias due to the unstandardized use of concurrent psychotropic medications across its sample.

Furthermore, the included studies are highly heterogeneous in their design, interventions (ketamine vs. ayahuasca), and digital tools (comprehensive platforms vs. simple monitoring apps), which precludes direct comparisons and limits the generalizability of the findings. While the qualitative study and case reports were assessed as being of moderate to high quality in their reporting, their study designs do not allow for causal inferences. Consequently, the promising efficacy signals reported in this review should be considered preliminary and interpreted with significant caution until they can be substantiated by more rigorous, controlled research.

4.4. Limitations of This Review

This systematic review has its own limitations. First, the search was restricted to studies published in English, which may have introduced a language bias and resulted in the omission of relevant studies published in other languages. Second, due to the rapid emergence of this field, some relevant data may exist in the grey literature (e.g., conference proceedings, white papers) which was not systematically searched. Finally, as with any systematic review, there is the potential for missed studies despite a comprehensive search strategy.

4.5. Implications for Clinical Practice and Future Research

The findings of this review have immediate implications for clinicians and digital health leaders. While at-home, digitally supported psychedelic therapy is a promising model for expanding access, it cannot be implemented without a robust framework for ensuring patient safety. This must include: (1) rigorous patient screening to exclude individuals at high risk for adverse psychiatric events [21,22,23,24]; (2) clear and unambiguous protocols for medication administration to prevent errors like the overdose reported [25,26,27,28]; and (3) a well-defined system for remote monitoring and immediate crisis response [29,30].

A critical gap across all studies is the absence of long-term follow-up data. Without evidence on sustained efficacy or cumulative risks, the ethical and clinical appropriateness of digitally supported, at-home psychedelic therapy remains uncertain. Given the occurrence of serious psychiatric adverse events and a documented overdose, the lack of longitudinal safety monitoring represents a significant limitation that must be addressed before widespread dissemination.

For future research, the most urgent need is for high-quality RCTs that directly compare at-home, digitally supported models to standard in-clinic administration or an active control. Such studies are essential for isolating the true effect of the at-home model and understanding its non-inferiority or superiority to established care. Furthermore, long-term safety and efficacy data are completely absent from the current literature and are urgently needed. Research should also aim to disentangle the “active ingredients” of these complex interventions—is the primary benefit from the drug, the digital support, the human coaching, or the combination?

5. Conclusions

The digital enablement of psychedelic-assisted therapy in non-clinical settings is a rapidly advancing frontier in psychiatry that holds the potential to revolutionize access to a powerful class of treatments. The current evidence, though limited, suggests that these models are feasible and can produce rapid and significant antidepressant and anxiolytic effects. However, this potential is shadowed by serious safety concerns that must be addressed through the development of rigorous clinical guidelines, best practices for remote monitoring, and further high-quality research. Without a framework that prioritizes patient safety, the promise of this innovative treatment model may be undermined by preventable harm.