Journal of Analytical Toxicology,Vol. 26, October 2002
Evaluation of Freezing Point DepressionOsmolality for
ClassifyingRandom Urine SpecimensDefined as
Substituted Under HHS/DOT Criteria
Janine Denis Cook1, Mark W. Hannon, Sr.2, Tamdan Vo3, and Yale H. Caplan4
I University of Maryland, Baltimore, Maryland; 2Quest Diagnostics Incorporated, Wallingford, Connecticut;
3Montgomery General Hospital, Olney, Maryland; and 4National Scientific Services, Baltimore, Maryland
I Abstract
[
This study evaluates the analytical performance characteristics of
freezing point depression osmolality in urine and osmolality as a
suitable analytical indicator for determining the concentration of
urine specimens submitted for workplace drug testing. Specifically,
this study attempted to determine the utility of urine osmolality to
serve as an indicator of substitution as defined by HHS/SAMHSA
criteria. Urine osmolality was validated by determining the
accuracy, precision, analytical sensitivity, reportable range, and
reference interval for the method. Osmolality was measured in
workplace urine specimens (n -- 66) with creatinine concentrations
-< 5.0 mg/dL. Comparing the results with the lower limit of the
random urine reference intervals for specific gravity (1.002) and
osmolality (50 mOsm/kg), 62% had specific gravities <_1.001,
52% had osmolalities < 50 mOsm/kg, and 47% had both a
creatinine _<5.0 mgldt, specific gravity _<1.001 and an osmolality
< 50 mOsmlkg. Urine specimens (n = 311) were collected from 35
volunteers enrolled in a controlled water loading sludy in which at
least 80 oz (2370 mL) of fluid was ingested over a 6-h period. The
lowest achieved osmolality was 28 mOsmlkg. Polyuria disorders
have produced abnormally low urine osmolalities (lowest reported
18 mOsmlkg) but osmolalities < 23 mOsmlkg have resulted in
death from water intoxication. An osmolality substitution cut-off to
delineate a specimen as inconsistent with normal human urine can
be set at some value < 50 mOsmlkg, when used in a population of
individuals with urine creatinine concentrations _<5.0 mg/dL
Introduction
Workplace drug testing is mandated for federal and U.S.
Department of Transportation (DOT) regulated employees.
Because of possible personnel actions associated with confirmed positive workplace drug-test results, drug abusers may
perceive an incentive to corrupt the drug-testing process
through substitution, dilution, and/or adulteration of their
424
submitted urine specimens. Forensic laboratories have employed several techniques to verify the validity of a specimen
submitted for testing. Biochemical markers such as creatinine
and indicators of concentration, such as specific gravity, are
routinely used in the forensic drug-testing laboratory to assess
the concentration of the urine sample. To ensure the effectiveness of drug testing programs, the Substance Abuse and
Mental Health Services Administration (SAMHSA) has provided guidance to forensic workplace drug-testing laboratories
for assessing specimen validity. One aspect of evaluating specimen validity is by analyzing each for creatinine concentration
and, when creatinine concentration is < 20.0 mg/dL, determining the specific gravity of the specimen (1). Definitions for
two categories of specimens tested for both creatinine concentration and specific gravity are (1) dilute: urine creatinine
concentration < 20.0 mg/dL and specific gravity < 1.003 and
(2) substituted: urine creatinine concentration < 5.0 mg/dL
and specific gravity < 1.001 or > 1.020.
The criteria for classifying a urine specimen submitted for
drug testing as substituted, that is, not consistent with normal
human urine, is based on a published review of four types of
studies (2). The studies included normal random urine reference ranges, published clinical studies involvingthe analysis of
random urines, medical conditions resulting in over hydration,
and published water loading studies. The review also presented
osmolality data for each of the four types of studies. The osmolality data were considered in the selection of the specific
gravity substitution criteria cut-off (2).
Creatinine is a biochemical waste product resulting from
muscle metabolism and is a normal constituent of urine. Both
specific gravity and osmolality are assessments of urine concentration that correlate directly with one another. Specific
gravity is the weight of urine compared with the weight of an
equal volume of distilled water at the same temperature. Osmolality is the measure of the total solute concentration in a
liquid and is directly related to the molar concentration of
Reproduction(photocopying)of editorialcontentof thisjournalis prohibitedwithoutpublisher'spermission.
Journal of AnalyticalToxicology,Vol.26, October 2002
those solutes, most notably waste products such as creatinine
and urea.
The goal of this study was to evaluate freezing point depression osmolality as a marker for classifying a urine specimen
submitted for workplace drug testing as substituted, that is, inconsistent with normal human urine.
Materials and Methods
6-h period, with 40 oz of water consumed during the first 3 h
and 40 oz of water consumed during the second 3-h period.
Urine specimens were collected upon awakening the morning
of the test, just prior to the start of fluid consumption, at the
end of each hour of the 6-h test period and upon awakening the
morning following the test day. Per the protocol, each participant was asked to submit 9 urine specimens for an expected
total of 315 specimens. Four participants were each able to produce only 8 specimens for a total of 311 specimens. The urine
aliquots were analyzed (Kroll Laboratory Specialists, Gretna,
LA) for creatinine by an automated modified Jaff~ method on
the Hitachi 747 (Roche Diagnostic Systems) and for specific
gravity on the Atago model UG-1 (Vee Gee Scientific, Inc.,
Kirkland, WA) electronic refractometer. Specimens were stored
frozen until the performance of the urine osmolality determinations. Urine osmolality analyseswere determined by freezing
point depression using the Advance MicroOsmometer, model
3MO. All urine osmolality results < 100 mOsm/kg were analyzed in duplicate.
Additionally, two urine specimens identified as substituted
were analyzed by freezing point depression osmolality. Statistics were performed using Microsoft Excel software (Microsoft
Corp., Roselle, IL).
The analytical performance characteristics of freezing point
depression osmolality were assessed using two different osmometers, AdvancedMicro-Osmometer models 3MO and 3300
(AdvancedInstruments, Norwood, MA)
A series of studies were conducted to assess the analytical
performance of the urine osmolality method. Both accuracy
and precision studies were performed through the repeated
analysis of the commercially prepared 50, 100, 850, and/or
900 mOsm/kg calibration standards (aqueous sodium chloride solutions) and the Clinitol| 290-mOsm/kg reference solution (aqueous sodium chloride solutions in a serum-based
matrix) (AdvancedInstruments). The analytical sensitivity of
the method was determined by the repeated analysis of deionized water. The reportable range of the urine osmolality
method was determined through the analysis of serial dilutions
Results
of the 50-mOsm/kg calibration standard and of a 2500mOsm/kg sodium chloride stock solution. Four linearly related dilutionswere prepared from each (50, 37.5, 25, 12.5, and
The accuracy and within run precision data for urine osmo0 and 2500, 1875, 1250, 625, and 0, respectively).The reference
lality are presented in TableI. TableII presents the between-run
interval was verified through the analysis of random urine
precision data.
specimens collected from healthy ambulatory volunteers (n =
Through the repeated analysis of deionized water (n = 20,
23). The acceptability of the volunteer specimens was determean 0.0 mOsm/kg, S.D. 0.9 mOsm/kg), the analytical sensimined through dipstick urinalysis (Chemstrip 6, Roche Diagtivity of urine osmolality was determined. The limit of detecnostics, Indianapolis, IN). Method correlations between urine
tion was 1.8 mOsm/kg at 2 S.D. or 2.7 mOsm/kg at 3 S.D. and
creatinine concentration, specific gravity, and osmolality were
the coefficient of variation was 16.8%.
done with urine specimens (n = 294) submitted for workplace
The determined reportable range from the serial dilution
drug testing. The urine creatinine automated screening
studies was 0-2500 mOsm/kg. Over this concentration range,
method involved an alkaline picrate reaction on the Hitachi
the recoveries ranged from 97 to 103%. The linear regression
747 analyzer(RocheDiagnostic Systems). Urine specificgravity
statistics for the determined reportable range were y = 0.99xwas determined with a digital refractometer (UR-1Urine Spe1.8, r = 0.9998 (model 3MO) and y = 1.00x - 1.7, r = 0.9999
cific Gravity Meter, NSG Precision Cells, Farmingdate, NY).
(model 3300).
To compare the diagnostic utility of osmolality in classifyinga
The literature reference interval of 50-1400 mOsm/kg for
urine specimen as substituted, urine specimens submitted for
urine osmotality was demonstrated per the NCCLS protocol
workplacedrug testing meeting the HHS/DOTsubstituted crithrough the analysis of urine specimens collected from 23
teria based on urine creatinine concentration and specific gravity were also analyzed for osmolality.
Table I. Accuracy and Within-Run Precision Studies for the Two Advanced
Urine specimens (n = 311) were colInstrument Micro-Osmometer Instruments
lected from 35 volunteers who particiConcentration Instrument
n
Mean
S.D.
C.V. (%) Recovery(%)
pated in a Department of Transportation
(DOT) controlled hydration study (3).
23
I00
2
2.4
I00.0
Study participants (25 females, 10 males;
lOOmOsm/kg model 3MO
23
292
4
1.2
100.7
17 whites, 13 Hispanics, 2 Asians, and 3
290 mOsm/kg model 3MO
900 mOsm/kg
model 3MO
23
908
9
0.9
100.9
blacks; aged 19-55 years, mean age 30 +
50 mOsm/kg
model 3300
20
48
1.1
2.3
96.0
9.4 years; body mass index [BMI] range
29o mOsm/kg model 3300
20
290
2.0
0.7
100.0
18-34 kg/m~-,mean BMI 24 + 4.2 kg/m2)
85o mOsm/kg model 3300
20
849
2.6
0.3
99.9
ingested at least 80 oz (2370 mL) over a
425
Journal of Analytical Toxicology, Vol. 26, October 2002
healthy ambulatory subjects (data range: 166--1047 mOsm/kg,
mean 641 mOsm/kg, S.D. 311 mOsm/kg) (4).
The urine creatinine and specific gravity methods were validated as follows: creatinine (accuracy 99-109%, precision
1.5% at 5.0 mg/dL, linearity 1.25-100 mg/dL) and specific
gravity (accuracy 100%, precision 0% at 1.001, linearity
1.000-1.030). The analytical performance of the creatinine
assay at low concentrations was assessed by the inclusion of a
3.2-rag/alL creatinine quality control material (mean 3.222
mg/dL, S.D. 0.391 mg/dL, C.V. 12.13%). The analytical
performance of the refractometer at the low end was assessed
by testing deionized water at 1.000 (mean 1.000, S.D. 0.0,
C.V. 0%).
The correlation of urine osmolality with the conventional
markers of urine concentration was assessed through the analysis of 294 urine specimens submitted for workplace drug
testing. Descriptive statistics for the creatinine, specific gravity,
and osmolality data from these urine specimens is presented in
Table III. The linear regression statistics for creatinine concentration versus specific gravity are y = 0.0001x + 1.0, r =
0.83, for creatinine concentration versus osmolality are # =
5.24x - 100.6, r = 0.81, and for specific gravity versus osmolality are # = 5.70x - 168.8, r = 0.95. The data are presented
graphically in Figures 1-3.
The 66 urine specimens submitted for workplace drug
testing that had urine creatinine concentrations _r 5.0 mg/dL
were analyzed for urine osmolality: 62% (41/66) had specific
gravities ~ 1.001 and thus were classified as substituted per the
HHS/DOT criteria; 52% (34/66) had osmolalities < 50
mOsm/kg; and 47% (31/66) had both a specific gravity _ 1.001
and an osmolality < 50 mOsm/kg, indicating that both specific
gravity and osmolality identified similar percentages of specimens as being less than the lower limit of the respective reference intervals (Figure 4). Two of the specimens with urine
creatinine concentrations ~ 5.0 mg/dL also had specific gravities ~ 1.020, and thus are also classified as substituted. The
creatinine, specific gravity, osmolality, and pH data on these
two specimens are 3 mg/dL, 1.025, 361 mOsm/kg, and pH 4
and 4 mg/dL, 1.033, 522 mOsm/kg, and pH 4, respectively.
Creatinine - Specific Gravity Correlation
1.05
1.04
>
9
1.03
O
..,
.p:,
:-*,.
9
1.02
.
~
9
O
1.01
9
9
+
"
1
i
0
i
=
100
200
Creatinine concentration
300
Figure 1. Urine creatinine and specific gravity correlation of the workplace drug testing specimens (n = 294). The linear regression line is
overlaid on the data.
Creatinine - Osmolality Correlation
2500
=~ ~5oo
..~." ~ . . ~ : : . . : "
'~ 1000
9 " "~ "~
"~.~ I " "
(~ 500 c.. +- :..,.~,,w-.-. . . .
opt7.
0
50
100
150
200
250
300
350
Creatinine concentration
Figure 2. Urine creatinine and osmolality correlation of the workplace
drug testing specimens (n = 294). The linear regression line is overlaid
on the data.
Specific Gravity - Osmolality Correlation
2500
2000
*9
:~
1500
9
9
~
+
+
1000
8
e~
9 ; ~ "
500
0
9
* -~| 9
:.
9
.
1.01
9
9. .
9
,
.
.
1.02
.
1.03
1.04
1.05
Specific gravity
Figure3. Urine specific gravity and osmolality correlation of the work-
Table II. Between-Run Precision Studies for the Advanced
Instruments MicroOsmometer model 3300
Concentration
n
Mean
S.D.
50 mOsm/kg
290 mOsm/kg
850 mOsm/kg
10
15
14
50
293
849
2.9
3.0
5.0
place drug testing specimens (n = 294). The linear regression line is
overlaid on the data.
C.V. (%) Recovery(%)
5.8
1.1
0.6
100.0
101.0
99.9
Substituted Specimens
1,001
>
E
Table III. Summary Statistics of the Data from the 294
Urine Specimens Submitted for Workplace Drug Testing
o
0
I
if)
Mean
S.D.
Range
426
Creatinine
(mg/dL)
Specific gravity
Osmolality
(mOsm/kg)
77
77
0-340
1.016
1.012
1.000-1.046
672
519
0-2404
1.0~
10
20
30
40
50
Osmolality
Figure 4. Graphical representation of those specimens submitted for
workplace drug testing that meet the criteria of creatinine _<5.0, as a
function of specific gravity <_1.001 and osmolality < 50 mOsm/kg.
Journal of AnalyticalToxicology,Vol. 26, October 2002
The urine specimens collected from the controlled hydration
Creatinine Concentration vs. Specific Gravity
study were analyzed for creatinine concentration, specific
1.040
gravity, and osmolality. There was better tracking and correla1.035
tion between creatinine concentration, specific gravity, and
1.030
osmolality (creatinine concentration vs. specific gravity, r =
1.025
9
. 9 9
0.9317; creatinine concentration vs. osmolality, r = 0.8925;
9
;.,
9
2 1.020
, :..
and specific gravity vs. osmolality, r = 0.9596) in this con1.015
trolled study when compared with the data on those speci1.010
mens submitted for workplace drug testing. Linear regression
1.005
1.000
equations for the respective analyte pairings were creatinine
0.0
100.0
200.0
300.0
concentration versus specific gravity (y = 0.0001x + 1.002),
Creatinine concentration (mg/dL)
creatinine concentration versus osmolality (y = 4.00x + 103),
and specific gravity versus osmolality (y = 35389.7x- 35367.9).
Figure5. Correlation of creatinine concentration with specific gravity
Graphical representations of the correlation data are given in
in the urine specimens from the hydration study (n = 311). The linear
the following charts (Figures 5-7).
regression equation is overlaid on the data points.
A statistical summary of the hydration study data by collection time is presented in Table IV. The lowest osmolality result
Creatinine Concentration vs. Osmolality
was achieved at the end of the second hour of the hydration
1500
protocol. Levels generally returned to initial values at the end
of the 24-h testing period (a.m.: mean 736 mOsm/kg, SD 253
E
:"'-,
1000
mOsrn/kg; a.m. + 24 h: mean 665 mOsm/kg, SD 216 mOsm/kg;
p = 0.086). A plot of the osmolality means for each protocol collection time is given in Figure 8.
-~ 500
:"-"
Out of 311 total urine specimens from the hydration study,
103 samples had a urine creatinine concentration < 20.0
~
0
,
,
mg/dL, 68 samples had specific gravities < 1.003, and 66 sam0
0.0
100.0
200.0
300.0
ples met the dilute criteria with both a creatinine result
Creatinine concentration
< 20.0 mg/dL and a specific gravity result < 1.003. The specific
Figure6. Correlationof creatinineconcentrationwith osmolalityin the
gravity/osmolality pairs for those specimens with a creatinine
urine specimensfrom the hydrationstudy(n = 311).The linear regresconcentration < 20 mg/dL is presented in Figure 9. The
sion equation is overlaidon the data points.
number of specimens classified as dilute at each given time
point in the hydration protocol is given in Table IV. None of the
Specific Gravity vs. Osmolality
311 specimens met the paired criteria for substitution, a cre1600
atinine concentration ___5.0 mg/dL and a specific gravity
1400
5 1.001 or ___1.020. Only six samples (2%) from five different
1200
participants (14%) were < 50 mOsm/kg, the lower limit of the
1000
reference interval for urine osmolality (pairings: creatinine,
g
800
specific gravity, osmolality: 15.3, 1.002, 28; 7.8, 1.001, 33; 6.6,
600
1.001, 33; 7.5, 1.001, 36; 8.7, 1.001, 47; 8.7, 1.002, 47) (4).
400
Table V presents the data pairings representing the lowest
200
results for each analyte.
0
Critics of the HHS/DOT substitution criteria have argued
o
1.000
1.010
1.020
1.030
1.040
that creatinine is unduly influenced by muscle mass, espeSpecific gravity
cially in females and in those with low BMIs. Results from the
Figure7. Correlation of specific gravity with osmolality in the urine
water-loaded patient with the lowest BMI (17.58 kg/m2 [BMI
specimens
from the hydration study (n = 311). The linear regression
healthy reference interval: 18.5-24.9 kg/m2]) were graphed to
equation is overlaid on the data points.
depict the relative trending of creatinine, specific gravity, and
osmolality with water ingestion (Figure
10). The correlation coefficient of subject
Table IV. Statistical Summary of Subject Hydration Study Data by Collection Time
BMI to the osmolality of his/her first
lh
2h
3h
4h
5h
6h
A.M.+
morning urine collection is r = -0.47.
Osmolality
0h
post
post
post post post post
24 h
Based on the two-tailed paired t-test cal(mOsm/kg) A.M.
culation, there was no significant differ736
756
463
261
152
182
207
170
665
Mean
ence between subjects on first morning
253
290
400
261
112
193
190
144
216
osmolality by subject age, gender, BMI,
S.D.
or race up to the 0.05 level of significance.
Range 239-1148 126-1134 59-1602 28--1217 47-703 47-756 33-698 33-440 208-1083
Two forensic random urine specimens
Dilute
0
0
3
11
14
14
11
13
0
reported as substituted per the Federal
.
9
9
.
.
9
427
Journal of Analytical Toxicology, Vol. 26, October 2002
guidelines were analyzed for urine osmolality. Results from
these specimens are presented in Table VI.
In the DOT hydration study, the lowest urine osmolality
achieved was 28 mOsm/kg. This is consistent with previous
water-loading studies where the lowest osmolality result seen
OsmolalityMeans
Conclusions
Urine specific gravities _< 1.001 can be produced during
states of excessive hydration; however, there is limited numerical flexibilityin urine specific gravity numbers when compared with osmolality values. Below the lower limit of the reference intervals for these two tests, specific gravity provides a
range of numbers from 1.000 to 1.001, and osmolality gives a
range of numbers from 0 to 49 mOsm/kg. For instance, in the
present hydration study, 24 urine specimens had specific gravities of 1.001 and none had a specific gravity of 1.000. The individual results for the hydration study urine specimens with
osmolalities < 50 mg/dL are 28, 33, 33, 34, 36, 47, and 47
mOsm/kg. Thus, osmolality is well suited to differentiate between an overly dilute and a substituted urine sample.
Osmolality is not readily automated and requires special instrumentation. The test has become less labor intensive,
though, as manufacturers have enhanced instrument capabilities. The instrumentation is relatively inexpensive, as are the
necessary reagents and supplies.
It has been previously established that creatinine concentration _<5.0 mg/dL and specific gravity ___1.001 are valid criteria for classifying a submitted urine specimen as substituted,
that is, inconsistent with normal human urine (2). Method
evaluation experiments reported here have proven that freezing
point depression osmolality has acceptable accuracy and precision and an acceptable reportable range and reference interval. Urine freezing point depression osmolality correlates
well with urine creatinine concentration and specific gravity,
indicating that all three markers behave similarly as markers
of urine concentration.
800 7
700
600
5O0
.~ 400
300
200
100
O
0
1
2
3
4
5
6
7
8
9
Specimen
Figure8. Plot of the mean osmolality results at each of the nine collection time points. Specimen 1 was the first morning void, specimen
2 was the baseline sample collected just prior to the beginning of the
hydration protocol, specimens 3-8 were collected at the end of each
hour of the 6-h hydration protocol, and specimen 9 was the first
morning void collected the morning after.
Creatinine < 20 mg/dL
3oo
E
u)
250
0
200
9
.!
E
-~
150
100
50
;I il,
v
o
O
1.000
1.005
1.010
1.015
Specific gravity
1.020
Figure 9. The specific gravity/osmolality pairs for those specimenswith
a creatinine concentration < 20 mg/dL (n = 103).
Hydration Trending of Subject with BMI of 17.5 kg/m 2
Table V. Data Pairings for the Lowest Hydration Results
Creatinine Concentration
(mg/dL)
Specific
Gravity
Osmolality
(mOsm/kg)
5.2*
6.6-25.7(n=24)
15.3
1.002
1.001'
1.002
112
33-211 (n= 24)
28*
140
1,030
120 '
1.025
100
1,020
,,~
5o
1,015
6O
1.010
t~
4O
* The lowest resultof the pair.
'E
o
Table Vl. Creatinine Concentration, Specific Gravity, and
Osmolality Results on Urine Specimens Reported as
Substituted per SAMHSA Criteria
Specimen
Creatinine
Concentration
(mg/dL)
Specific Gravity
Osmolality
(mOsm/kg)
1
2
0.2, 0.0
0.3
1.000, 1.000
1.000
17, 17
2, 2
428
1.OO5
20
tO
0
I
1
I
2
I
3
I
4
[
5
I
6
Specimens
I
7
I
8
1.0OO
9
I~ o~.I,r2,~
t--~- specific gravity
Figure10. Hydration resultsfrom a studyparticipant with a BMI of 17.5
kg/m2. Specimen 1 was the first morning void, specimen 2 was the
baseline sample collected just prior to the beginning of the hydration
protocol, specimens 3-8 were collected at the end of each hour of the
6 h hydration protocol, and specimen 9 was the first morning void collected the morning after.
Journalof AnalyticalToxicology,VQI.26, October 2002
was 32 mOsm/kg (2). Clinical case studies of disorders that
manifest with polyuria have produced some very low (< 50
mOsm/kg the lower limit of the random urine reference interval) urine osmolality results (18, 23, 32, 36, 37, 38, 40, 45,
and 49 mOsm/kg) (3). Interestingly, the patient associated
with the two lowest results (18 and 23 mOsm/kg) died of water
intoxication shortly after the collection of those random urine
specimens (2). Additionally, the analysis of two forensic specimens previously identified as substituted produced urine osmolality results of 2 and 17 mOsm/kg, respectively.
These results can be combined with the previously published extensiveliterature review of normal urine reference intervals (50-1200 mOsm/kg), clinical studies involvingthe analysis of random urine specimens, theoretical renal dilutional
limits (7 mOsm/kg theoretical limit with a 29 L/d ingestion),
medical conditions resulting in overhydration (psychogenic
polydipsia,water intoxication, diabetes insipidus, nephrogenic
diabetes, and iatrogenic diabetes; lowest case study result is 18
mOsm/kg), and water loading studies (lowest achieved is 28
mOsm/kg) (3). Therefore, an osmolality substitution cut-off to
delineate a specimen as inconsistent with normal human urine
can be set at some value < 50 mOsm/kg, when used in a population of individuals with urine creatinine concentration
< 5.0 mg/dL.
Acknowledgments
and specific gravity results used in the analytical performance
studies; Terrie Kowalewskiof Quest Diagnostics for the use of
the Model 3MO osmometer and supplies; Ed Mehrlust of Advanced Instruments, Inc. for the use of the model 3300 osmometer; Kenneth Edgell, U.S. Department of Transportation, Washington, D.C. for providing the water loading study
specimens; and Leon R. Glass, Ph.D., Kroll Laboratory Specialists, Gretna, LA for the creatinine concentration and specific gravity data on the water loading study samples.
Portions of this paper were originally presented as posters
at the 2000 and 2001 American Association for Clinical
Chemistry national meetings in San Francisco, CA and
Chicago, IL, respectively.Abstracts of these studies were published in Clinical Chemistry [46:A19 6 (2000) and 47:
A74 (2001)].
References
1. NLCP: State of the Science Update #1. Subject: Urine specimen
validity testing Evaluation of the scientific data used to define a
urine specimen as substituted. (February 14, 2000). http://workplace.samhsa.gov/ResourceCenter/resource.asp?RCategorylD=8&String=Regulations/Guidance.
2. J.D. Cook, Y.H. Caplan, C.P. LoDico, and D.M. Bush. The characterization of human urine for specimen validity determination
in workplace drug testing: a review. J. Anal. ToxicoL 24" 579-588
(20o0).
K.C. Edgell, L.R. Glass, and Y.H. Caplan. Paired measurements of
creatinine and specific gravity after water loading. Abstract.
J. Anal. ToxicoL 25:367 (2001).
4. C.A. Burtis and E.R. Ashwood, Eds. Tietz Textbook of Clinical
Chemistry, 3rd ed. WB Saunders, Philadelphia, PA, 1999.
3.
We would like to thank Ross Lowe, Ph.D. of Quest Diagnostics for providing the workplace urine specimens, the two
substituted urine specimens, and the creatinine concentration
429