Who Is the Best Pain Doctor in Dallas? What ChatGPT Said, and Why.

A patient with a bad back does not start by calling a referral line. More and more, she opens an app and types a question.

So we asked it for her. On September 29, 2026, we typed one question into ChatGPT and into Google: “Who is the best pain management doctor in Dallas?” No follow-up, no extra detail. Only the question a person types late at night after another bad day.

The answers are worth reading closely, because this is where more patients now begin. In rater8’s 2026 Patient Choice Report, 36 percent of patients searching for a doctor said AI tools like ChatGPT influenced their choice, up from 17 percent a year earlier. Adoption was highest among patients aged 45 to 60, the group many established practices most want to see.

In a single year, the share of patients who said AI tools shaped their choice of doctor rose from 17 to 36 percent.

What ChatGPT said

It opened with a caution: there is no single best doctor, because the right choice depends on the condition. Then it named five physicians and explained why each one belonged on the list.

The explanations are the interesting part.

For the first name, a physician who has been on the faculty of a major academic medical center since 1991, it cited board certification, a subspecialty in pain medicine, three decades of tenure and repeated recognition by D Magazine, Castle Connolly and Texas Monthly. For the second, it cited double board certification and fellowship training at Cleveland Clinic.

Only after those came patient ratings. ChatGPT called them useful and then placed them “secondary to board certification, relevant procedural expertise, hospital affiliation, and experience with your particular condition.”

Its sources were telling too: the medical center’s own physician page, a physician profile on a scheduling platform and the WebMD directory.

What Google said

Google did not produce an AI Overview for this question. It showed a map with three practices, all with strong ratings and large review counts, followed by Yelp, one physician’s own website, two Reddit threads and U.S. News. Google has said its local results weigh relevance, distance and prominence, and ratings and reviews feed that last one.

The gap between the two

The practice at the top of Google’s map, with 770 reviews, did not appear anywhere in ChatGPT’s answer. Neither did the practice with more than 1,200 reviews.

Only one physician appeared prominently in both: a doctor with a 4.9 rating, more than 500 reviews and a personal website strong enough to rank on its own.

Two systems, two sets of rules. Google’s map leaned on proximity, ratings and review volume. The AI assistant leaned on something much closer to what a referring physician would weigh: credentials, tenure, institutional affiliation and recognition from people outside the practice, all stated plainly somewhere it could read them.

What this means for an established practice

Your credentials need to be written where machines can read them. A board certification framed in the waiting room is invisible to an AI assistant. It belongs on your website, on every directory profile and in the public certification record, and it should read the same everywhere. This is the work we call the Foundation.

Recognition compounds. Each named mention by a city magazine, a ranking body or a professional society gives an assistant something to cite. The physician ChatGPT led with had decades of it.

Reviews keep you in the conversation. They earned places on Google’s map and served as supporting evidence in ChatGPT’s answer. Patients still care a great deal: 75 percent will not book below four stars. A strong rating is the entry ticket. On its own, it rarely decides the answer.

The answer narrows when the question does. ChatGPT closed by offering to refine the list by condition: sciatica, neuropathy, post-surgical pain. A practice known for something specific, and described that way in public, has a far better chance of being the answer to that follow-up. It is one reason we work with one practice per discipline in each market.

A ten-minute exercise

Open ChatGPT. Type the question a new patient would ask about your specialty in your city, and read what comes back.

Are you named? If you are, is the reason given the one you would want a patient to hear? If you are not, who is, and what did the assistant say about them?

For many practice owners, those ten minutes are the most honest reading of their reputation they will get all year.

A note on method: this was one query, on one day, from one account. AI answers vary by session, location and over time, so treat this as a snapshot.

For physician-owned practices, here is how we work with physicians.