Reading Ketamine Response Rates: Denominators, Comparators, and Follow-Up

A response rate can sound precise while leaving the most important questions unanswered. If a clinic says that most patients improve with ketamine, readers need to know who was counted, what improvement meant, and how long it lasted. The goal is not to reject percentages. It is to give them enough context to become useful evidence rather than a persuasive number detached from the study that produced it.

Start with the denominator

The denominator is the group used to calculate the percentage. It might include everyone randomized, everyone who received a first session, or only those who completed a treatment course. These groups are not necessarily equivalent. If people leave because of cost, adverse effects, or lack of benefit, an outcome among completers may look more favorable than an outcome among all starters. Ask how missing assessments were handled rather than assuming every unreported patient improved.

Eligibility criteria matter as much as sample size. A trial of adults with nonpsychotic unipolar depression does not automatically estimate outcomes for bipolar depression, adolescents, or people in an acute medical crisis. A larger sample improves some aspects of precision, but it cannot correct an inappropriate population match. Before comparing a personal situation with a study, identify the diagnosis, prior treatment history, and major exclusions.

Response is not the same as remission

A response generally means a specified reduction on a symptom scale, often at least half of the baseline severity. Remission means symptoms fall below a defined threshold. A person can meet a response definition while still having substantial depression. Neither label necessarily means that functioning has fully recovered or that improvement will persist. A report should specify the exact scale and criterion rather than substituting terms such as success or cure.

In the Murrough et al., single-infusion ketamine randomized trial, 2013, 24-hour response rates were 64% with ketamine and 28% with midazolam in a 73-person randomized trial. The comparison is meaningful because it identifies a control treatment and a time point. It is not a prediction that 64% of all clinic patients will recover permanently. Nor should the 28% control response be ignored when judging how much benefit can reasonably be attributed to the intervention.

Ask what the comparator contributes

Midazolam was an active placebo intended to improve the comparison with a treatment that produces noticeable effects. No control design is perfect, particularly when participants may recognize an altered state. Still, a randomized comparison answers a stronger causal question than asking only whether people felt better after attending a clinic. Improvement over time can reflect treatment, expectation, concurrent care, natural fluctuation, and other influences.

An active-treatment comparison asks a different question again. The Anand et al., ketamine versus ECT, 2023 tested ketamine against ECT using a noninferiority design in nonpsychotic treatment-resistant depression. Noninferiority means the result met a prespecified margin under that study's conditions. It does not prove equal effects, equal risks, or equal suitability in every clinical circumstance. The trial should not be stretched into a verdict about conditions it excluded.

Separate early outcomes from durability

A 24-hour outcome, an end-of-course outcome, and a six-month outcome should not appear as though they represent the same kind of success. Later assessments also need information about additional treatments and who remained in follow-up. If a reported benefit required ongoing sessions, that is relevant to cost, burden, and safety. The word lasting should be accompanied by an actual observation period.

The Daly et al., esketamine relapse-prevention randomized trial, 2019 illustrates another important distinction: maintenance participants were selected after benefiting from initial esketamine treatment with an oral antidepressant. A relapse-prevention result in that group cannot be used as the chance that any new patient will improve. Read our maintenance guide for the practical implications of this enriched design.

Translate the percentage into consultation questions

Ask a clinic whether its advertised figure comes from a published trial or its own records. Request the treatment route, outcome definition, assessment time, denominator, and handling of dropouts. If those details are unavailable, treat the number as uncertain rather than filling in the gaps optimistically. A reputable explanation can acknowledge uncertainty without implying that treatment is useless.

Use the consultation decision guide to place those answers alongside safety, alternatives, and follow-up arrangements. For personal monitoring, the treatment-record article explains how to track outcomes without equating every daily fluctuation with success or failure. The most useful percentage is not the highest one on a website. It is the one whose population, definition, comparator, and follow-up are clear enough to inform a realistic decision.

Medical disclaimer: This article provides general education, not diagnosis, individualized treatment advice, or emergency care. Evidence and regulatory statements refer to the dated sources cited, not an exhaustive verification of all changes through October 30, 2026.

Sources

1. Murrough et al., single-infusion ketamine randomized trial, 2013
2. Daly et al., esketamine relapse-prevention randomized trial, 2019
3. Anand et al., ketamine versus ECT, 2023