Two findings from our survey of 443 adults sit uncomfortably side by side, and together they describe a referral problem that lands in primary care.
First, Spravato was a blank for most of them. The drug, an esketamine spray carrying FDA approval for stubborn depression, meant nothing to 73 percent of our panel. Twenty-one percent knew the name alone; a thin 6 percent could say what it does.
Second, when we asked whose advice would tip them toward ketamine or esketamine care, 74 percent pointed to their own physician. Advertising managed 2 percent, podcasts a lone 1. Friends and family earned 18 percent, and 4 percent would heed someone they follow online who wore a uniform, military or first responder.
Patients will not arrive asking for this by name. When the subject comes up, the deciding voice is yours.
The population is larger than the label suggests
We asked whether depression, anxiety, or PTSD that first-line prescriptions left unresolved had shown up in their own life or the life of someone close, and 72 percent said yes. Thirty-seven percent meant their own life, 22 percent meant someone near them, and 13 percent answered both. The other 28 percent had no such history.
This was a general sample of adults 18 to 64, not a screened psychiatric cohort. Whatever the precision of any single figure, the direction is plain: inadequate response to first-line medication is a near-majority experience in this region, and your panel is not an exception.
Where patients say they will present
We asked respondents where they would start if they wanted this kind of care. Primary care physicians took 56 percent. Psychiatrists and other mental health clinicians took 23 percent. Online searching took 12 percent and a friend 1 percent; the leftover 5 percent simply did not know.
For most patients, then, the funnel starts in your exam room, including many who will never reach specialty psychiatry on their own. Where psychiatric waits run long, a primary care visit is less the first step toward evaluation than the only step that reliably happens.
What patients will actually say
One free-text item invited each person to write the search they would run when looking for help, and 319 wrote something. The answers were bare and symptomatic: ptsd treatments, therapist near me, help with depression, best ways to handle depression, depression medicine alternatives, how to help someone with depression. One simply wrote "depressed." Another wrote "someone please help me." No respondent typed a drug name or a clinical term.
A patient who has never heard the vocabulary cannot raise the subject. In a fifteen-minute visit, a missing question is not a missing need.
The screening implication is practical. A patient who says the medication "helps a little" may be describing a partial response with no framework for escalating it, because they do not know escalation exists.
Attitude is not the obstacle
Clinician reluctance here is sometimes framed as sparing patients an unwelcome suggestion. Our data does not support that framing. Respondents reacting to the idea of ketamine-based treatment for depression or PTSD most often chose cautious but open, at 34 percent, with hopeful or curious at 18 percent. Skeptics were 21 percent, the unaware another 18 percent, and 9 percent were negative. About half leaned open, and fewer than one in ten were firmly against.
What moves the cautious majority is institutional legitimacy. FDA approval would be decisive for 19 percent of respondents and weigh heavily for another 40 percent, a combined 59 percent; 14 percent said it made no difference.
That argues for precision. The approved product is esketamine, dosed only at certified sites with a supervised observation window after every session. IV ketamine for depression and ketamine dispensed for home use through telehealth are off-label arrangements with different oversight. Patients hear all three as one word. Drawing the line for them is one of the few things a referring physician can hand over that materially improves their decisions, and a patient-facing explainer such as Brain Recovery Centers' description of supervised Spravato treatment can reinforce it after the visit.
Coverage decides more of these cases than clinical fit does
Asked to rank what matters in picking a clinic, respondents placed insurance in their top two at a rate of 85 percent. A short drive ranked next, well back, at 43 percent. Speed of results trailed FDA approval, 24 percent to 27, while privacy sat at 11 and a specialty in service members and first responders at 10.
Insurance would decide or heavily sway whether 65 percent pursued treatment at all. Given a direct tradeoff, 51 percent would accept more administrative steps to stay covered, against 23 percent who would self-pay for speed and 26 percent unsure.
With multiple answers allowed, TRICARE covered 5 percent of the sample and Medicare 23, but the two big blocs were commercial insurance at 39 and Medicaid at 37; 9 percent had nothing. A pathway built only around commercially insured patients misses most of the people in front of you.
A referral is more likely to convert when it carries the coverage answer. Knowing which nearby sites take Medicaid, and which handle prior authorization in-house, changes outcomes more than more patient education does.
What this suggests for practice
- Treat partial response as a prompt to discuss the full menu, including options the patient cannot name.
- Name the approved product and its setting requirements, and separate it from off-label models; patients cannot do that alone.
- Keep a short, current list of referral sites with payer participation noted, and know who handles prior authorization at each.
- Expect the conversation to begin with you.
None of this is clinical guidance or a judgment about any individual patient. Indication, eligibility, contraindications, and risk are governed by the labeling and the patient's history. The survey speaks only to what patients know, what they want, and where they will show up.
For patients in crisis, give them 988. Calling or texting those digits anywhere in the country reaches trained Lifeline staff; it costs nothing, never closes, and takes calls from people worried about someone else as readily as from those at risk themselves.
Methodology
As publisher, we commissioned this first-party study and footed the bill. Pollfish fielded it through its consumer panel to adults 18 to 64 in ten states, among them Missouri, Kansas, Iowa, Ohio, and Minnesota; by the June 23, 2026 close there were 443 completes. Respondents consented at the outset. We report top-line percentages only; on multi-select items the shares exceed 100 in total. Pollfish has finished validating the panel, so every figure here is final.