Electronic health record (EHR) information can be grouped by how it is organized: structured, unstructured, or semi-structured. This is a useful way to discuss data and analytics, but it is not a single official classification that every EHR uses. Real records combine these forms, and the same information may be stored differently across products.
What is EHR data?
An EHR is a digital record of health information maintained by healthcare providers. Depending on the organization and system, it can include demographics, diagnoses, medications, allergies, notes, test results, and other information relevant to care. “EHR” and “EMR” are sometimes used casually as if they mean the same thing, but ONC describes EHRs as broader records that can support information being shared across providers and care settings; an EMR is often used for a more limited record within a practice.
1. Structured data
Structured data is entered into defined fields or selected from a controlled list. Examples may include a coded diagnosis, medication order, date, vital sign, or standardized demographic field. Because values follow an expected format, systems can often sort, filter, report, and exchange them more consistently—provided the data are entered accurately and the relevant standards are used.
Example: A blood pressure recorded in designated systolic and diastolic fields is easier to trend than the same value mentioned only in free text.
2. Unstructured data
Unstructured data does not fit neatly into a fixed set of fields. Examples can include narrative progress notes, dictated text, scanned documents, or images. These materials may preserve context and detail that are important to clinicians, but they can be harder to search, compare, or use in automated reports without additional review or processing.
Example: A clinician’s narrative describing a patient’s symptoms may contain important context that is not represented by a diagnosis code alone.
3. Semi-structured data
Semi-structured data combines predictable organization with flexible content. A note template with labeled sections and free-text responses is one example. A document may also include tags or metadata that identify its type and date while retaining narrative text. The exact boundary between structured and semi-structured depends on the system and the use case.
Example: A templated visit note may have separate sections for history, assessment, and plan, with a mixture of selected options and clinician-written narrative in each section.
