What Is Medical Schema?

Medical schema is the group of labels in the schema.org vocabulary that were built for health businesses and clinicians, such as MedicalClinic for a facility and Physician for a doctor or a doctor's office. For a health practice, these labels let your site say plainly that this is a clinic, here is its specialty and here is who works there. Schema.org is a shared vocabulary that Google and other search engines read, and it lists these types as official options.

How medical schema works

Schema markup in general labels the facts on a page. Medical schema picks health-specific labels from the same vocabulary. Generic labels describe any business. Medical labels describe yours.

Schema.org defines MedicalClinic as a facility, often tied to a hospital or medical school, that is devoted to specific diagnosis or healthcare. Its family tree matters. MedicalClinic sits under MedicalBusiness, which sits under LocalBusiness, so it inherits the standard fields for a local business, such as address, telephone, opening hours and URL. It adds two health-specific properties, medicalSpecialty and availableService. It also inherits isAcceptingNewPatients from the MedicalOrganization branch of its family tree.

Schema.org defines Physician as an individual physician or a physician's office, treated as a medical organization. It carries medicalSpecialty, hospitalAffiliation, availableService and a usNPI property, which holds the 10-digit National Provider Identifier issued to US health care providers by CMS. That field is the bridge to the public record in the NPI registry.

Schema.org lists 26 more specific types under MedicalBusiness, including Dentist, Dermatology, Physiotherapy, PlasticSurgery, PrimaryCare and Psychiatric, along with MedicalClinic and Physician. I do not see a chiropractor or med spa type on that list, so those practices need to choose the closest honest fit.

Google's guidance applies on top of the vocabulary. Its LocalBusiness page says to use the most specific subtype possible, and that name and address are the required properties for that markup. Its introduction to structured data says markup is a standardized way to give Google information about a page, and that following its guidelines is a condition of eligibility for rich results. Google also states that no additional requirements or special optimizations exist for AI Overviews or AI Mode, and that no special structured data is needed there.

A practice with several locations repeats the clinic layer once per location, so each address, phone number and set of hours sits on its own page and is labeled there. A parent organization can tie the locations together, which keeps the brand and the individual offices from looking like unrelated businesses to a machine.

You do not need to learn the property names. Your developer or plugin handles the code. What you need to supply is a list of true facts, written once, so the code can be generated from the same source as your website and listings.

Putting it together, a well-built health site uses three layers.

• The clinic page. MedicalClinic markup with name, address, phone, URL, hours, specialty and services, placed on your entity home.
• Each provider page. Physician or an equivalent person-level label with credentials, specialty and NPI, linked to the clinic.
• Matching visible text. Every labeled fact appears on the page where a visitor can read it.

Example: medical schema at a health practice

This example is illustrative, not a transcript. A hormone clinic has one physician, two nurse practitioners and a clinic manager. The clinic page carries a short factual block and MedicalClinic markup listing the address, phone, hours, the specialty of hormone therapy and the services offered. The physician has a page that carries Physician markup with her specialty, her hospital affiliation, her NPI and a link back to the clinic.

The two nurse practitioners are a different case, since schema.org lists no type named for nurse practitioners under MedicalBusiness. The clinic labels them with a general person-level type, sets an accurate job title and links each to the clinic. It does not label them as Physician, because that would state something untrue. The manager gets no markup at all, since the page does not need one.

Then the owner runs the clinic page and the physician page through the Schema Markup Validator at validator.schema.org. The tool shows the types it detects and any errors. She checks the address, phone and name in the results against the visible page and against her NPI record, and finds one old phone number in the markup, left by the plugin's default setting. She fixes it at the source.

Real data tells you what to expect. The markup helps Google and other tools read facts about the practice directly. Google says it may make a page eligible for richer results, with no guarantee, and says special markup is not required for AI features. For other AI tools, I have not found a vendor statement that medical schema is a ranking or citation factor. In the MGMA analysis of 33,340 ChatGPT citation events, practice-owned websites accounted for about 23% of citations in that analysis. The author founded a commercial service and the study is correlational. So treat medical schema as hygiene that pairs with a clear page.

The FTC adds a legal angle. Its health products guidance says health claims need competent and reliable scientific evidence. Markup that lists a treatment is a statement about your services, so keep the labeled services consistent with what you actually offer and can support.

How to check yours in 60 seconds

1. Run your home page and one provider page through the Schema Markup Validator at validator.schema.org.
2. Look at the detected types. Do you see MedicalClinic, Physician or another health type, or only a generic one?
3. Open the detected item and read the name, address, phone and specialty.
4. Compare those values to the visible page and to your NPI record.

A generic type with the right facts is acceptable. A health type with wrong facts needs fixing first.

What to do about it

1. Label the clinic page as MedicalClinic, or the closest honest subtype, with name, address, phone, URL, opening hours and medical specialty.
2. Label each physician page as Physician, with specialty and the NPI that matches the public record. Use a general person-level label for clinicians without a matching health type.
3. Link each provider to the clinic they work at inside the markup.
4. List only services you offer and can support with the visible page.
5. Add the markup as JSON-LD, then test it in the Schema Markup Validator.
6. Retest after edits, and when a provider joins, leaves or changes a phone number.
7. For several locations, give each location its own page and its own clinic markup, and link them to the parent organization.
8. Keep one shared fact sheet with every name, address, phone, specialty, service and NPI, and generate the markup from it so the site, the listings and the code say the same thing.