Data Model that includes ICF

[bennidi]

ICF provides a (disappointing) model of the medical domains used to describe functioning and impairments of the human body. It is designed without any considerations of algorithmic processing. It provides only a single hierarchical structure and is not decomposed into smaller reusable concepts.

One example is the category b28011:Pain in chest=Sensation of unpleasant feeling indicating potential or actual damage to some body structure felt in the chest. This is quite dumb for various reasons:

• A sensation of pain can be induced from other sources (e.g. damage to nervous system in spinal cord).
• The description mixes medical finding with diagnostic indications.
• The category pain does not enumerate the different possible types (stinging, numb, pressure-like), nor does it define relevant other dimensions of pain like frequency (constant, pulsating ...), nor is it a sensible combination of the reference to a body structure (chest) and a symptom (pain).

It is almost unbelievable that this is supposed to be the work of experts - officially endorsed by all 191 WHO Member States!

This can be done a lot better by decomposition into more fundamental concepts allowing for flexible reassembly of its parts into descriptions of malfunctioning (symptoms). Consult our draft of a more flexible medical data model.