We expected a list of keywords. We got something closer to a stack of letters.

Near the end of our survey of 443 adults across the Midwest, we left a blank box and asked a simple question: if you were looking for help, which words would you put into Google? We thought we might learn which treatment names people knew. Instead, 319 people wrote answers that read like notes to themselves.

What was in the box

Here are some of the phrases, exactly as they were typed:

Read them slowly and you notice what is absent. There are no brand names, no drug classes, no clinical terms. People typed what they felt and what they wanted. The most technical phrase in the set was "depression medicine alternatives," and even that is a wish more than a query.

Why the missing drug names matter

The absence is not random. Elsewhere in the same survey, we put the name Spravato in front of people. It is esketamine in nasal spray form, FDA-approved for depression that standard treatment could not lift, and 73 percent of respondents had never encountered it. Another fifth knew the name but not the medicine, and a small 6 percent could actually describe it. Readers who want the basics can start with a plain-language guide to Spravato.

Line those two findings up and the story is simple. People are searching for relief in their own words, and some of the newer options for hard-to-treat depression are described online in words they have never encountered. The two vocabularies pass each other in the dark.

Who was doing the searching

These were not specialists or patients recruited from a clinic. They were ordinary adults aged 18 to 64 from ten states between Ohio and Nebraska. And a large share of them had a personal stake. Our question on hard-to-shift depression, anxiety, or PTSD, the kind standard medication did not help, drew a yes from 72 percent of respondents, for themselves, someone close, or both. Adding the 37 percent who said themselves to the 13 percent who said both, 50 percent have been through it personally.

That context changes how the search phrases read. "How to cope with my anxiety" is not hypothetical for most of the people who could have written it.

What people said they would actually do

Searching and deciding turned out to be different acts. When respondents chose where to begin with a newer depression treatment, an online search got just 12 percent. The primary care doctor drew 56 percent and a psychiatrist 23 percent; 5 percent were unsure where to even begin, and a friend was the first call for 1 percent.

On whose recommendation would push them to try something new, the doctor won by a landslide at 74 percent. Ads drew 2 percent and podcast hosts 1 percent.

So the search box is a place of private admission. The exam room is where the decision gets made. Most people, it seems, already know that.

The feeling underneath

We also asked people how ketamine therapy for depression or PTSD sounded to them on first hearing. The largest group, 34 percent, landed on cautious but open. Next came 21 percent who were skeptical, then two groups of 18 percent each, one hopeful or curious and one that had never heard of it, and finally 9 percent who reacted badly.

The spread fits the search phrases well. These are people who want help and are wary of hype. They are not looking for a miracle. They are looking for a way through.

If you have typed something like this

If any of the phrases above could have come from your own search history, a few things may be useful to know.

If what you typed was a plea

"Someone please help me" was in our data too. If that is where you are, please reach out now rather than later. Wherever you are in the country and whatever the hour, the Suicide and Crisis Lifeline is waiting at 988, by phone or text; veterans can press 1. A real person will answer.

What this is and is not

This article reports what survey respondents told us about how they search for help. It is not medical advice, and it makes no claim that any particular treatment will work for you. Talk options through with the clinician who treats you.

Methodology

The survey ran through the consumer panel at Pollfish and stopped collecting on June 23, 2026, once 443 adults, none younger than 18 or older than 64, had finished it, in Oklahoma, Missouri, Nebraska, Ohio, Iowa, Kansas, Wisconsin, Indiana, Illinois, and Minnesota. Quoted phrases come from the 319 respondents who filled in the open search question, reproduced as typed. The figures we cite are top-line and final, taken after Pollfish's validation. Paying for the study, and commissioning it, was this article's publisher.