Telehealth-supported ketamine for depression and anxiety: A systematic review.

A systematic review of telehealth-supported ketamine for depression and anxiety in adults found that, across three US commercial provider studies, symptom response rates ranged from 49.5% to 62.8% on the PHQ-9 and 47.6% to 62.9% on the GAD-7, with remission rates of 20.7% to 32.6% and 23.9% to 31.3%, respectively. However, all studies were at critical risk of bias, the evidence certainty was very low, and follow-up data were available for only a minority of the up to 16,876 patients. The review concludes that the evidence is provider-generated and insufficient to guide clinical practice, though it may inform future hybrid service design in Australasia.

Study at a glance

Abstract

ObjectivesTo systematically review the effectiveness, safety, and care models of telehealth-supported ketamine for depression and/or anxiety in adults and consider implications for Australasian clinical practice.MethodsFollowing PRISMA 2020, we searched for randomised or observational studies of adults receiving at-home telehealth-supported ketamine for depression and/or anxiety, to August 2025. Primary outcomes were ≥50% symptom reduction and remission. Programme characteristics were synthesised narratively. Risk of bias and certainty of evidence were assessed. Protocol was not registered, and the study was not funded.ResultsOf 3,857 records screened, three met criteria, all from US commercial providers. Programmes differed in dosing schedules, clinical staffing, supervision, and safety monitoring. Across up to 16,876 patients, follow-up outcome data were available for a minority. Response ranged 49.5-62.8% (PHQ-9) and 47.6-62.9% (GAD-7), and remission 20.7-32.6% and 23.9-31.3%, respectively. All studies were at critical overall risk of bias. GRADE certainty for effectiveness and safety was very low.ConclusionsTelehealth-supported sublingual racemic ketamine was associated with symptom improvements in selected populations. Yet, the evidence base is provider generated, and at critical risk of bias. Its main contribution to Australasia is to inform consideration of model components for a future hybrid service design.