The Wikipedia page on data integrity lists 4 types of integrity constraints: entity integrity, referential integrity, domain integrity, and user-defined integrity.

Is there a name for integrity constraints that stem from common sense / knowledge?

Example :

I have one table with 6 columns containing some data about hospital admissions:

  • hospital_admission_id (integer field)
  • patient_id (integer field)
  • patient_date_of_birth (date field -> I am aware that it should be in some patient table but ignore this in this example)
  • date_of_hospital_admission (date field)
  • date_of_hospital_discharge (date field)
  • hospital_admission_note (text field: a relatively long text by the medical staff summarizing some data about the patient when he was admitted at the hospital)

One integrity constraint stemming from common sense / knowledge is that the patient more than one year old at the time of hospital admission (i.e. date_of_hospital_admission - patient_date_of_birth > 1 year), hospital_admission_note shouldn't say that the patient is a neonate.

Another integrity constraint from common sense / knowledge is that a patient cannot be admitted in the hospital if he already is in the hospital (i.e. some checks would have to be done around date_of_hospital_admission and date_of_hospital_discharge).

I feel that:

  • It's not an entity integrity constraint since it's not an issue around primary keys.
  • It's not a referential integrity constraint since it's not an issue around foreign keys.
  • It's not a domain integrity constraint since it's not about checking whether values are in some list.
  • I guess it could be qualified as a user-defined integrity constraint but I thought they might be some more precise technology to specifically designate integrity constraints stemming from common sense / knowledge and not just a specific user's preferences.

You're going beyond the scope of referential constraints with this (good) question. What you are looking for is more to do with Ontologies and Archetypes than RDBMS constraints. You could also check out this page which is on "Intelligent Databases" and gives as an example, medical records.

Check out the OpenEHR (Open Electronic Health Record here) whereby the system goes beyond data storage (essentially AIUI, your question). It's a fascinating area, touching on machine learning and artifical intelligence. It is far from a trivial problem - and in particular with medical records with all sorts of different regulatory frameworks and privacy concerns. Doctors, surgeons, dentists and pharmacists have different "views" of a patient - how to tease all these out is a major issue in the field. It will come as no surprise that the OpenEHR system is one of many competing in this fascinating and challenging area.

From the OpenEHR (what is page)

What is openEHR?

Abstract architecture image openEHR is a virtual community working on interoperability and computability in e-health. Its main focus is electronic patient records (EHRs) and systems.

The openEHR Foundation has published a set of specifications defining a health information reference model, a language for building 'clinical models', or archetypes, which are separate from the software, and a query language. The architecture is designed to make use of external health terminologies, such as SNOMED CT, LOINC and ICDx. Components and systems conforming to openEHR are 'open' in terms of data (they obey the published openEHR XML Schemas), models (they are driven by archetypes, written in the published ADL formalism) and APIs. They share the key openEHR innovation of adaptability, due to the archetypes being external to the software, and significant parts of the software being machine-derived from the archetypes.

The essential outcome is systems and tools for computing with health information at a semantic level, thus enabling true analytic functions like decision support, and research querying.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.