SS> Time attribute to be explicitly stated in the model. We have one potential candidate for Single Point in Time. DK> The only difference is the specimen type urine sample vs 24 hour urine sample also with rates, needs a time interval to be defined. One solution is that 'Urine Specimen' is underspecified. So what type of urine sample is Pt:Urine? RH> We don't want to conflate all of the technique into the sample type. FA> From LOINC manual: For urine collection, 24H is the “standard” integrated measure and these are almost always reported as mass rates (MRat), substance rates (SRat), or catalytic (CRat) rates. These would contrast with spot or random urine tests that are represented as point (PT) measures in our nomenclature and usually reported as concentrations -- MCnc, CCnc, or SCnc for mass, catalytic, and substance concentrations respectively. However, we can also report the average concentration on a 24-hour specimen – in this case the time aspect value would be 24H but the property would be MCnc/SCnc/CCnc instead of MRat/SRat/CRat.
RH> The Sample type for rates is the same as a total quality measurement. 24 hour timed collection. DK> 24 hour Diastolic Blood pressure is a quantity averaged over 24 hours. FA> We don't have attribute aggregation in the observables model. DK> We don't want to build complex formulas in the Observables model - could put average in the technique, e.g. averaging technique. It would still be subsumed by point in time (PT) DK> So changing the Urine to something like Spot urine would fix the subsumption problem. FA> We have a concept Spot Urine Sample in SNOMED CT. |