A patient has tried two antidepressants, each at a sensible dose and for a sensible stretch, and is still struggling. You have several doors you can open. Each asks something different of the patient, of your practice, and of the insurer. This comparison lays them out side by side for Midwest primary care physicians and other referrers, drawing on patient-side numbers from our commissioned survey.
The patient-side context
Our survey reached 443 Midwest adults. For 72 percent, depression, anxiety, or PTSD had outlived standard medication, either in their own lives or in someone close, with 37 percent describing it as theirs alone and 13 percent as both theirs and someone else's. Those figures are self-reported, top-line, and not yet final. They describe lived experience rather than a clinical rate, but they suggest the "after two trials" moment is common.
Three other top-line findings shape every option below. Awareness of later-line treatments is low: Spravato was unknown to 73 percent, and roughly a quarter knew what TMS was. Trust runs through the physician: for 74 percent, their own doctor's recommendation outranks every other voice. And coverage is the gate: 85 percent ranked insurance as a top-two provider priority.
Option A: Keep managing in primary care
What it involves. Dose optimization, switching, augmentation, or combination, with closer follow-up.
Patient burden. Low. Familiar setting, familiar clinician, no new travel.
Time to start. Immediate.
Coverage friction. Usually modest, though some agents require prior authorization.
Watch for. Drift. Without a defined checkpoint, a patient can spend a long time in "let's give it a bit longer." Set a date and a measure.
Option B: Collaborative care or psychiatric e-consult
What it involves. A psychiatrist advises on medication strategy while the patient stays with you, often through a care manager.
Patient burden. Low. The patient may never see the psychiatrist in person.
Time to start. Often faster than a full referral where the model exists.
Coverage friction. Varies by payer and whether your practice bills collaborative care codes.
Watch for. Availability. Many Midwest practices, especially rural ones, have no access yet.
Option C: Full psychiatry referral
What it involves. Transfer or co-management with a psychiatrist who can reassess diagnosis and oversee escalation.
Patient burden. Moderate. New relationship, possibly long travel.
Time to start. Frequently the slowest option because of wait lists.
Coverage friction. Depends on network breadth. Among respondents, the Medicaid share (37 percent) nearly matched the commercial share (39 percent), and psychiatry networks for Medicaid can be thin.
Watch for. Patients falling through the gap between referral and first visit. Only 23 percent of respondents said a psychiatrist would be their first stop, so the handoff from primary care needs to be warm.
Option D: TMS referral
What it involves. A course of transcranial magnetic stimulation, typically in frequent sessions over several weeks.
Patient burden. High on time, low on side effects. No medication, no anesthesia, and patients can usually drive afterward.
Time to start. Depends on prior authorization, which commonly requires documented medication trials.
Coverage friction. Moderate to high. Documentation is key.
Patient appeal. Notable. Just over half of respondents had not heard of TMS yet valued a drug-free option, and 13 percent who did know TMS weighted drug-free care heavily.
Option E: Esketamine referral
What it involves. Spravato nasal spray, approved for adult treatment-resistant depression, self-administered under direct observation at a site certified to give Spravato, followed by a monitoring period.
Patient burden. Significant in the early phase, with frequent visits. Patients cannot drive after a session and need a ride.
Time to start. Depends on authorization and site availability.
Coverage friction. Moderate to high. Payers often require evidence of prior antidepressant failure. Just over half of respondents said they would accept extra hoops for covered care.
Patient appeal. For 59 percent, approval by the FDA was a deciding or big consideration. Awareness is the obstacle, since so few have heard of it.
Watch for. Blood pressure, substance use history, and the patient's transport situation. A patient-friendly Spravato overview can help set expectations.
Option F: Becoming the site
What it involves. Some practices choose to certify under the Spravato REMS and administer esketamine themselves.
Patient burden. Lowest of the interventional options, since the patient stays with a clinician they already trust.
Practice burden. Real: certification, space for monitoring, staff time, and buy-and-bill or specialty pharmacy logistics.
Why consider it. Forty-four percent of respondents wanted proximity in their top two, and in-person care was the most preferred format. In parts of the Midwest where the nearest certified center is far away, this may be the only realistic path for some patients.
A note on off-label ketamine
IV ketamine clinics and at-home ketamine programs are not on this list as referral options because they fall outside the FDA-approved framework for depression. Patients will ask about them. A clear explanation of the difference in approval status and oversight serves them better than silence.
Choosing among them
The options are not exclusive. Many patients benefit from psychotherapy alongside any of them, and a psychiatric opinion can guide the choice. A reasonable default is to keep optimizing in primary care with a firm checkpoint, bring in psychiatric input early, and discuss interventional options once the patient meets the criteria their payer uses.
One habit makes every option on this list move faster: a clean, dated record of each antidepressant trial, with the drug, dose, duration, response, and reason for stopping. Payers ask for it before approving TMS or esketamine, psychiatrists ask for it at intake, and patients living with depression are rarely able to reconstruct it from memory. Building it as you go is the cheapest referral tool a practice has.
Whatever the path, pair it with a safety plan. Every patient, in every state, can call or text 988 at all hours for the Suicide and Crisis Lifeline, and veterans can press 1.
Methodology
Patient data come from our Pollfish consumer panel survey; 443 adults aged 18 to 64 responded, and the survey shut on June 23, 2026. They were spread across Oklahoma, Wisconsin, Missouri, Indiana, Minnesota, Kansas, Illinois, Iowa, Ohio, and Nebraska. Top-line figures only, all final after validation. Our publisher commissioned the survey and absorbed the cost. This comparison is informational and does not substitute for clinical judgment.