Bruno Cadonna wrote: > Hi all, > > After a thread on the openehr-clinical list about associated symptoms > (cf. subject "Associated symptoms" on openehr-clinical) some technical > questions about computation of time came to my mind. > > In openEHR the time data can be found in different places in different > ENTRY types. > > OBSERVATIONs have HISTORYs for attributes "data" and "state". Time > data in HISTORY represent the timing of observations, they may be used > for computations. Time data in the ITEM_STRUCTURE of the attribute > "data" of an EVENT in a HISTORY should not be used for computation. > An example for the latter is archetype > openEHR-EHR-OBSERVATION.symptom_check.v1 which includes CLUSTER > archetype openEHR-EHR-CLUSTER.symptom.v3.adl. > openEHR-EHR-CLUSTER.symptom.v3.adl defines time data at archetype node > at0063 "Date / time of onset". This time data should not be used for > computation to avoid contradiction with time data recorded in the > HISTORY of the OBSERVATION. > > Is this right? Any timing information that is correct in the EHR should be usable for computation - there is no reason not to. BTW times like 'date/time of onset' will usually occur in Evaluation archetypes, not Observation archetypes. > > > EVALUATIONs have time data somewhere in the ITEM_STRUCTURE of their > attribute "data". This time data may be used for computation. > For example archetype openEHR-EHR-EVALUATION.problem-diagnosis.v1 > defines time data at archetype node at0003 "Date of initial onset" > which may be used for computation. > > Is this right?
same meaning as above. > > > ACTIONs have time data in attribute "time" which states when the > action completed and in the ITEM_STRUCTURE of the attribute > "description". Both time data may be used for computation. > An example for the latter is archetype > openEHR-EHR-ACTION.medication.v1 which includes archetype > openEHR-EHR-ITEM_TREE.medication.v1. > openEHR-EHR-ITEM_TREE.medication.v1 defines time data at archetype > node at0019 "Date (time) of first administration". This time data may > be used for computation. > > Is this right? > What about archetype node at0032 "Date (time) of last administration" > of openEHR-EHR-ITEM_TREE.medication.v1? Does it contradict time data > in attribute "time" in ACTIONs? I assume that this wouldbe more properly named 'date/time of previous administation' - I assume it is there to provide an idea of the period between administrations - useful e.g. in the case of painkiller medication. > > > INSTRUCTIONs have time data in "expiry_time" and in the ITEM_STRUCTURE > of the attribute "description". Both time data may be used for > computation. > An example for the latter is archetype > openEHR-EHR-INSTRUCTION.medication.v1 which includes archetype > openEHR-EHR-ITEM_TREE.medication.v1. > openEHR-EHR-ITEM_TREE.medication.v1 defines time data at archetype > node at0019 "Date (time) of first administration". This time data may > be used for computation. > > Is this right? > Here archetype node at0032 "Date (time) of last administration" of > openEHR-EHR-ITEM_TREE.medication.v1 could also contradict the > attribute "expiry_time" of INSTRUCTIONs, but the semantics of these > time data is quite different, thus this should not be a problem, right? * all of the times you have mentioned occur somewhere on a notional real world timeline concerning the individual patient. Some are built-in to the openEHR reference model because they are universal, e.g. the times in History and Action (it dosn't matter what the data being collected are, these times always make sense). Others are completely dependent on the data being collected and are defined in archetypes. The latter are no less meaningful on the patient timeline however. * - thomas beale
