Setting expectations: what response actually looks like in the data

One of the things we promise people who come to Music City Ketamine is that we will tell them the truth about how their treatment is going. That includes the part nobody likes: ketamine does not work for everyone, and the data are clear about who tends to respond and when. Ketamine is FDA-approved as an anesthetic; its use for depression, anxiety, PTSD, and chronic pain is off-label.

The foundational study most clinicians point to is a 2013 two-site randomized trial published in American Journal of Psychiatry. Murrough and colleagues compared a single IV ketamine infusion against midazolam in 73 patients with treatment-resistant depression. At 24 hours, 64% of the ketamine arm met response criteria versus 28% on midazolam. That headline number is genuinely encouraging. The quieter finding is the one we make sure to share at intake: roughly a third of the ketamine arm did not respond acutely, even with a properly dosed infusion in a controlled setting.

A 2018 individual participant data meta-analysis by Wilkinson and colleagues, also in American Journal of Psychiatry, pooled patient-level data across multiple single-dose IV ketamine trials. The review confirmed robust acute antidepressant and anti-suicidal effects, while highlighting two things people should hear before they start: the effects diminish over days to weeks, and a meaningful subset of patients simply do not respond. Non-response is not rare. It is part of the distribution.

The induction series and the typical response timeline

Most clinics, ours included, deliver IV ketamine in an induction series. Six infusions over two to three weeks is the standard pattern, drawn from the protocols used in the repeated-dosing studies. The reason for that schedule is not arbitrary. The published evidence suggests that the people who are going to respond tend to start showing it early.

A 2019 study by Phillips and colleagues, published in American Journal of Psychiatry, looked at trajectories of response across repeated IV ketamine for treatment-resistant depression. Most responders showed clinically meaningful improvement within the first few infusions. Patients without measurable response by mid-series rarely converted to responders later in the series. That last sentence is the one we underline. If sessions one through three have produced nothing, sessions five and six are unlikely to suddenly change the picture.

This is why we frame the first three to four infusions as a structured trial, not a commitment to the full series. You are not signing up for six sessions and hoping for the best. You are gathering data with us about whether your particular biology is going to engage with this mechanism.

What we measure: PHQ-9, MADRS, GAD-7, and your own life

You cannot tell whether something is working without a baseline and a measurement. Before your first infusion, we capture standardized scores: PHQ-9 for depression, GAD-7 for anxiety, and depending on the case, additional instruments for PTSD or pain. We repeat these between infusions. The 2017 American Psychiatric Association consensus statement on ketamine, published as a Council of Research Task Force paper in JAMA Psychiatry by Sanacora and colleagues, specifically recommends standardized symptom monitoring across infusions and reassessing the treatment plan when patients fail to show meaningful improvement after an adequate trial. That is the standard we hold ourselves to.

Numbers are not the whole picture, though. We also ask about the texture of your day. Are you sleeping a little more soundly? Did you answer a text you have been avoiding for a month? Did the weight on your chest lift for an afternoon? These signals sometimes show up before the validated scores move. Conversely, scores can drop a few points without much real-world change, and that gap matters.

A useful working rule: if your PHQ-9 has dropped 50% or more, or you have crossed below clinical thresholds, that is a response. If your scores have nudged a little but life feels measurably easier, that is partial response. If neither has happened after four infusions, we have a conversation.

What 'non-response' means in the literature

In the research, response is usually defined as a 50% or greater reduction on a validated rating scale. Remission is a stricter bar, typically a score below a clinical cutoff. Non-response is what is left over: patients whose scores moved less than 50%, or did not move at all.

Across the IV ketamine literature for treatment-resistant depression, non-response rates in the first acute trial sit somewhere in the 30% to 45% range depending on the study and the population. The treatment-resistant depression evidence is the most robust piece of the dataset, and even there, real non-response is a normal and expected outcome for a meaningful minority of patients.

This is worth restating because it gets lost in marketing copy: a clinic that tells you ketamine works for everyone is not telling you the truth the studies tell. Ketamine is not a panacea. It is a real treatment with real response rates, and a real failure rate that we owe you the honesty to name.

Partial response versus no response — they are not the same

One of the most important distinctions we make at session four is between partial response and no response. They look similar on the surface, but they call for different decisions.

Partial response means something is moving. PHQ-9 dropped from 22 to 16. You had two genuinely better days last week. The bottom is a little less close than it was. In that case, we may extend the series, adjust dose within safe ranges, tighten the timing between sessions, or pivot earlier into a maintenance schedule. Partial responders often become more solid responders with the right adjustments.

No response means the needle has not moved. Scores are flat. Daily life is unchanged. You feel the dissociation during the infusion and then you go back to exactly where you started. In that case, more sessions are unlikely to produce a different outcome, and the honest move is to stop and look at other options.

The reason we separate these so carefully is that conflating them produces two bad outcomes: pulling people who would have benefited off treatment too early, and keeping people on a treatment that is not working long past the point it should have stopped.

When to stop: the conversation we have at session four

Around session four, if scores have not moved and your daily life has not shifted, Marla Peterson, CRNA, or a member of the clinical team will sit down with you and walk through the data. We look at your pre-treatment scores, your between-session scores, and your subjective report. We review weight-based dosing, infusion duration, and any concurrent medications that might be blunting response. Sometimes a protocol adjustment is the right call before declaring true non-response.

If, after that review, the picture is genuinely flat, we will say so. We would rather tell you the truth and refer you out than book you for two more sessions you are unlikely to benefit from. That is the part of practice that builds trust, and frankly, it is the standard the APA consensus document expects from clinicians delivering this treatment.

This is also where we coordinate with your prescribing psychiatrist or primary care provider. A real non-response is useful clinical information for them. It tells them something about your neurochemistry and helps them decide what to try next. Never stop or change medications on your own; that conversation belongs with the clinician who prescribed them.

What we recommend instead: TMS, ECT, Spravato, KAP, medication review

Non-response to IV ketamine is not the end of the road. Several treatments have evidence in treatment-resistant depression, and the mechanisms differ enough that ketamine non-responders frequently respond to other options.

If you came to us and ketamine did not work, our job is to send you toward whichever of these is the most reasonable next step for your situation. Not to keep selling sessions.

When ketamine might still be worth a second look later

Non-response now does not always mean non-response forever. There are a few scenarios in which a second look at ketamine, months or a year down the line, can be reasonable. Significant changes in concurrent medications, a meaningful improvement in sleep architecture, completion of a successful TMS or ECT course, or a stretch of psychotherapy that has shifted underlying patterns can all change the biological context. Some patients who were flat on a first trial respond on a second one a year later, when the surrounding picture has changed.

That is a decision that belongs with you, your prescribing provider, and us together. We will not pressure you back through the door. If and when the timing is right, the door is open and the data review starts the same way: baseline scores, structured trial, honest read at session four.

The throughline of this article, and of how we try to practice, is simple. Depression and chronic pain are hard enough without a clinic that hides the ball. The way ketamine works is real but partial, and the kindest version of care names that clearly. If it is working for you, we will keep going. If it is not, we will tell you, and we will help you find what comes next.