SNOMED Documentation Search

 Other Documents
Skip to end of metadata
Go to start of metadata

Current Version - Under Revision

When querying for absence of a finding the most likely motivation is to establish the absence of a finding. In the absence of evidence to the contrary the normal assumption is that an abnormal finding is absent. Furthermore, in most cases a point in time assertion of absence does not imply the finding was never true, nor that could not be true at a future point in time. Thus in most cases, a query for absence is more concerned with checking that there is no statement indicating the presence of the finding rather that searching for a statement of presence.

There are exceptions to this:

  • If the abnormal finding is usually found in associated with a confirmed finding:
    • E.g. "absence of chest pain" in a patient with a confirmed "myocardial infarction").
  • If the abnormal finding is obscure and may easily have been overlooked:
    • E.g. "Kopliks spots".
  • If an assertion of presence of a finding was made by an informer of unknown reliability:
    • E.g. Bystander asserts that patient had a "heart attack" but clinical assessment excludes this.

If these exceptions apply the presence of a statement of an absent finding may be of interest. However, this depends of specific thinking around the question being posed so that the query criteria achieve the desired result. It is not enough to simply search for a specific absence and its subtypes . The assumptions about presence or absence of a finding must be considered.

Example: determine the number of road accident victims who have been admitted to hospital but have no fractured bones. In practical terms the best approach would just be to exclude those with known presence of fracture. The assumption is that, unless a fracture is mentioned, they are not known to have a fracture.

Another possible motivation for looking for absence findings is to monitor or audit the delivery of care and check that appropriate questions have been asked, tests done, possibilities considered, etc. In these cases, the query needs to search for both presence and absence ... or possibly for a procedure code representing the appropriate examination or investigation.

Example: Were all patients admitted for routine surgery asked if they had any allergies.