Urine Sampling - Normalization

[eh_carbaugh@ccmail.pnl.gov]

     One can normalize urine samples based on time, volume, creatinine, 
     specific gravity, or other more unusual parameters.  The issue really 
     is which one is most appropriate?
     
     At Hanford, we normalize based on time, using the last voiding before 
     retiring through the first voiding upon rising in the morning, for two 
     nights running.  We refer to this as a "simulated 24-hour period" 
     though the term could reply to any normalized result.  Medley, 
     Kathren, and Miller (Health Physics 67(2)122-130) argued that this 
     underestimates total excretion by 15-20%.  That conclusion is not 
     surprising if you believe that the true time represented by the sample 
     interval more likely runs from 7-10 hours, rather than 12 hours.  
     Check yourselves on this point!
     
     Normalizing based on volume is a real problem because the volume does 
     vary substantially, depending on the liquid consumption, activity 
     level, etc.  It's my opinion that volume variations are a lot more 
     erratic and unpredictible than the time variations.
     
     Adjustment based on specific gravity was briefly investigated by Graul 
     and Stanley in 1982 (Am.Ind.Hyg.Assoc.J, 43(11):863).  They concluded 
     that it wasn't a good method for normalizing.  
     
     Recently, Esther Kim at Savannah River conducted a study, 
     "Determination of Time Interval of Urine Sample by Measuring 
     Creatinine in Urine," (ESH-HPT-950007, Westinghouse Savannah River 
     Company). Esther concluded that creatinine was no better than the 
     other normalizing techniques- it was not nearly as constant as tribal 
     memory seems to believe.  Contact Esther or Tom LaBone for a copy - I 
     don't think they have published it in the peer-reviewed literature.
     
     Bottom line- I still like the simulated 24-hour sample for convenience 
     or data interpretation and requiring the least manipulation for use.  
     Granted, it's a little inconvenient for the worker, but not as 
     inconvenient as a true 24-hour sample. It provides for collection of 
     enough urine that you can get a reasonable detection capability.  (As 
     Liz Brackett noted, a single void urine sample for Pu is not likely to 
     give you a satisfactory MDA.)  I believe that trying to use a single 
     void urine sample for routine Pu bioassay monitoring is essentially 
     worthless.  Single voids are great for tritium, and may have a place 
     for soluble uranium - but not for Pu!
     
        Gene Carbaugh
        Internal Dosimetry
        Pacific Northwest National Laboratory
        eh_carbaugh@pnl.gov