Dear colleagues,
I looked at the page url provided and have difficulty to inspect the figures because of a low resolution.
I agree that a more patterned way is necessary in order to have EHR systems that have access to more uniformly defined data needed for processing.
This requirement I call: Interpretability.
When all kinds of Observations need to be expressed in a uniform way we need ideally:
1- We need uniform patterns for each possible observation and all the various ways in which observations are expressed
2- We need to cater for ways to express the metadata of Observations that are part of a meaningful list
3- We need to cater for ways to express the metadata of the observation, itself
4- We need to cater for ways to express the metadata around the data subject
5- We need to cater for ways to express the metadata of the treatment process of the topic
6- We need to cater for ways to express the metadata of the clinical process
7- We need to cater for ways to express the metadata of the documentation process
8- We need to cater for ways to express the metadata of all the contextual information
ad1: e.g. quantitative-, semi-quantitative-, qualitative observations expressed using: numbers, text, codes, and many units of measurement,
ad2: e.g. Blood pressure, Lab panels, …
ad3: e.g. seeing, hearing, touching, smelling, tasting, …
ad4: e.g. body position, before, during or after exercise, etc.
ad5: e.g. reason for encounter, data collection/observation, history, examination, evaluation, planning, ordering, execution
ad6: e.g. intake, investigation, treatment, referral,
ad7: e.g. de novo recording of a fact, re-use of previously, recorded facts, clinical data, administrative data, preliminary data/unprocessed data, data admitted to the record
ad8: e.g. localisations in time and space in absolute and relative terms
In my way of thinking I start with the documentation process in the ENTRY with two CLUSTERS. One for the context data and one for the Panel. The Panel consists of a CLUSTER for each Panel component and one for the context of all. And then per Panel component two CLUSTERS: one for data and one for its context.
Gerard Freriks
+31 620 34 70 88
+31 182 22 59 46
gfrer@luna.nl
Kattensingel 20
2801 CA Gouda
the Netherlands
