The fastest-growing pitch in mental health care is convenience: skip the waiting room, do intake on your phone, get treatment mailed to your door. It is a compelling offer, and for this particular treatment it is not what most people say they want.
Our commissioned survey, which closed in June 2026, asked 443 Midwest adults from ten states how they would prefer to receive ketamine or esketamine therapy. Going to a clinic in person won 44 percent. A hybrid, beginning at a clinic and continuing at home, drew 23 percent, and at-home telehealth 22. The remaining 11 percent had no preference.
Combine the clinic and hybrid answers and 67 percent want treatment to begin in a room with staff. The convenience-first model on its own is the preference of roughly one respondent in five.
Why this cuts against the industry's direction
A lot of capital has bet that mental health care will follow every other consumer category, toward the phone and away from the building. For therapy appointments and medication management, that bet has held up.
Dosing is different, and people seem to sense it unprompted. A treatment that changes how you feel while you receive it makes a professional's presence feel less like friction and more like the point. Only the 11 percent of respondents with no preference treat this as pure logistics. Every figure comes from the panel's final validation pass.
Setting and distance are different questions
Two nearly identical figures deserve a flag so nobody misreads them. On a separate question about choosing a provider, 43 percent of respondents ranked proximity among their top two. That figure sits beside the 44 percent who prefer an in-person clinic, but they measure different things: one is the setting, the other is distance from your house. They point the same way, yet no one should merge them, and we cannot say how much they overlap without a cross-tabulation this dataset does not have.
What outranked proximity
Coverage did, and not narrowly. Respondents ranked insurance in their top two 85 percent of the time, nearly double the share who chose nearness. FDA approval drew 27 percent and quick results 24; discretion drew 11, and practices focused on service members and first responders 10.
Another question sharpened it. In deciding whether to pursue the treatment at all, insurance was decisive for 22 percent, while 43 percent more said it would weigh heavily, a combined 65 percent. Forced to choose between a covered option with more paperwork and a quicker cash option, 51 percent took the paperwork, 23 percent took speed, and 26 percent were unsure.
So the clinic preference is not a taste for a premium experience. It belongs to people who will also sit on hold for three weeks to avoid paying cash. Brain Recovery Centers, for example, describes how its in-clinic Spravato program works.
The awareness problem underneath
Spravato, whose FDA approval covers depression that shrugs off standard medicines, was unknown to 73 percent of respondents; 21 percent knew the name without the meaning, and a mere 6 percent could describe it.
That reframes the delivery finding. People are not choosing between models they understand; they are voicing a default instinct about medical care: if something is being given to me, I want professionals nearby. That instinct happens to match how the approved treatment is regulated, since esketamine must be given at a certified site with monitoring and never goes home.
The at-home category is different: generic ketamine prescribed off label through telehealth, with a different legal posture and widely varying oversight. When 22 percent of a general sample prefers home delivery, most are not comparing two known options. They are voting for convenience in the abstract.
The need is broad and the vocabulary is not clinical
Across the sample, 72 percent said routine prescriptions had left depression, anxiety, or PTSD unresolved for them or someone close. Personal experience accounted for 37 percent, a loved one's for 22, and both for 13, while 28 percent had none.
We also gave respondents an open box for the exact search they would run to find help, and 319 filled it in. They did not type pharmacology. They typed "someone please help me," "ptsd help," "therapist near me," and "depression medicine alternatives."
Nobody searches for a delivery model either. The preference in this survey is real, but it surfaces later, once someone is already talking options with a clinician. It is an evaluation criterion, not an entry point.
What we think this means
- Physical locations still matter here. A provider weighing a building against a purely virtual model should not assume virtual is the obvious future when two thirds of the market wants to start in person.
- The hybrid is underbuilt relative to demand. It drew as much support as pure telehealth and follows a sequence most people find intuitive: supervision while the response is unknown, flexibility once it is established.
- Coverage beats everything. In our forced-choice question, an inconvenient option a plan accepts beat a convenient cash one by two to one.
Necessary caution
This is consumer research about preferences, not clinical evidence. It cannot show whether any treatment works, for whom, or how well, and it neither gives medical advice nor promises any outcome. Esketamine is approved for a defined indication and given under supervision; at-home ketamine for mood is off label. Whether either suits a given person is for a clinician who knows that person's history.
If reading about treatment-resistant depression has brought you close to something darker, please do not sit alone with it. Call or text 988 at any hour, free; the Lifeline is there even when you are not in immediate danger.
Methodology
As publisher we commissioned this study and bore its full cost. Pollfish fielded the online survey among its consumer panel members, closing it on June 23, 2026; the finished sample was 443 adults, ages 18 to 64, from ten Midwest states, all of whom completed a consent screener. One open-text question drew 319 responses. Multiple-selection questions are shares of respondents and exceed 100 in total. Results follow the final validation of the panel.