Opened 3 years ago

Last modified 19 months ago

#165 new question

UOM available on UCUM document not in CDC PHIN VADS

Reported by: dmerriner Owned by: christof
Priority: major Milestone:
Component: Keywords: Tranlation Failure
Cc:

Description

Hello, We are working with our Department of Health (DOH) and we are struggling with several of the Units of Measure. These are reported by our reference laboratory and they are as follows:

log cps/mL log IU/mL

I find the UCUM reference of (respectively)

{Log_copies}/mL {Log_IU}/mL

Because these are not found on the PHIN VADs database, we are unable to send these results until resolved. It is impacting our ability to move foward. Can you share with me what the equivalent would be on the PHIN VADS v3 UOM database? Would it be possible these were omitted as an oversight?

Change History (3)

comment:1 Changed 23 months ago by gschadow

These copies and log copies are a problem. We need expertise or a good metrologically oriented discussion paper on this.

#90, #143, #165, #173

comment:2 Changed 19 months ago by christof

  • Owner set to christof

The problem that Guther mentions is related to the fact that a log can only be calculated from a numerical value (i.e. a "dimensionless" quantity). See a discussion of this for pH here: http://www.iupac.org/publications/pac/1985/pdf/5703x0531.pdf

In the case of pH this is resolved by introducing a denominator that is formally 1 mol/kg-l or 1 mol/ dm3 respectively. This denominator represents the rule that the measurement is to be taken at "standard conditions" (pressure, temperature). Meaning that a measurement taking place at non-standard conditions would yield different numerical values.

Coming back to the "units" in question here: {Log_copies}/mL {Log_IU}/mL are probably missing the point. It should rather be something like {Log}{copies/mL}, because the numerical values will NOT be proportinal to 1/mL here.

In a future release, we might come up with a formalized notation for log-units. They are are used frequently (dB, Neper), described as "levels". The formalism would require the introduction of a reference quantity, examples are dB[mV] or dB[10.nV] where the reference quantity would be given in square brackets. In our case here, this could be something like log[mL-1] or log[{copies}/mL] or log[[IU]/mL]. This is to be discussed and thoroughly checked against the UCUM formalism. This is a candidate for a future extension.

comment:3 Changed 19 months ago by christof

Note: Some "faithful adoption" of such a future formalism is seen here: https://rtmms.nist.gov/rtmms/index.htm#!units (go to Tab Units and search for decibel)

Last edited 19 months ago by christof (previous) (diff)
Note: See TracTickets for help on using tickets.