See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/14186695

Automated selection of statistical quality-

control procedures to assure meeting clinical or
analytical quality...

Article in Clinical Chemistry · March 1997

Source: PubMed

CITATIONS READS

29 52

2 authors, including:

James O. Westgard
University of Wisconsin–Madison
167 PUBLICATIONS 3,645 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Measurement uncertainty in screening immunoassays in blood establishments View project

Measurement uncertainty in medical laboratory View project

All content following this page was uploaded by James O. Westgard on 12 December 2014.

The user has requested enhancement of the downloaded file.


Automated Selection of Statistical Quality-Control Pro-
cedures to Assure Meeting Clinical or Analytical Qual-
ity Requirements, James O. Westgard1,2* and Bernard Stein2
[1 Dept. of Pathol. and Lab. Med., Univ. of Wisconsin
Med. School, Madison, WI 53792 (*address for correspon-
dence; fax 608-263-0910), and 2 Westgard Quality Corp.,
Ogunquit, ME 03907; e-mail www.westgard.com]
Efforts to totally automate laboratory testing processes
must address the issue of how to assure the quality of the
final test result. A recent survey by Tetrault and Steindel
[1] reported that laboratories are using the same quality-
control (QC) procedures today that they used 10 years
ago. Actually, more than half of the laboratories indicated
they are still using control limits set as the mean 6 2s (the
12s rule), a practice that dates to the 1950s and ’60s [2, 3],
when statistical QC was first used with manual methods
and the first generation of automated Technicon Auto-
Analyzer systems. Other laboratories indicated they are
using variations of the multirule type of QC procedure
introduced in the early 1980s [4].
Tetrault and Steindel [1] recommend that “the best
set of control rules will vary from method to method
and cannot be determined through simple algorithms
or formulas. The laboratorian has to balance true error-
detection capabilities against the probabilities of falsely
rejecting a good run.” The information needed about
the error-detection and false-rejection characteristics of
stable QC procedures became available in the clinical
chemistry literature almost 20 years ago [5– 8] and was
extended by Cembrowski et al. [9 –12] to consider
patient data algorithms. This information was incorpo-
rated into laboratory QC texts [13, 14] and clinical
chemistry texts [15, 16] in the 1990s and continues to be
expanded and improved by Parvin [17–20] and by
Smith and Kroft [21].
Guidelines and strategies for selecting QC proce-
dures were described in 1986 [13], and some detailed
applications have been published to demonstrate the
selection and design of QC procedures [22–24]. QC
selection grids were introduced in 1990 [25] to provide
a simple table “look up” approach for selecting QC
procedures on the basis of the size of systematic error
that must be detected and on an assay’s expected
stability or frequency of errors. More quantitative qual-
ity-planning models were introduced in 1991 [26, 27],
and a new planning tool—the OPSpecs chart3—was
developed [28, 29] to show the relation between allow-
able precision and accuracy as well as what QC is
necessary to assure detection of critical-sized errors that
would otherwise cause method performance to exceed
a defined analytical or clinical quality requirement. For
applications involving analytical quality requirements
and commonly used QC procedures, a compilation of
OPSpecs charts is available to support manual applica-
tions for the full range of CLIA proficiency testing
criteria for acceptability [30]. For applications involving
clinical quality requirements and a wider variety of QC
400 Clinical Chemistry 43, No. 2, 1997
Technical Briefs
procedures, a Windows3-based PC program (QC Vali-
dator3, Version 1.1; Westgard Quality Corp., Ogunquit,
ME) was developed to prepare power function graphs,
critical error graphs, and OPSpecs charts [31].
Quality-planning models describe the mathematical
relationships between a quality requirement and the
factors that can cause variation in the final test result.
Some of these factors are analytical, such as the stable
imprecision (smeas, %) and inaccuracy of the method and
the sensitivity of the QC procedure to detect unstable
random and systematic errors (DSEcrit). Others are pre-
analytical, such as the within-subject biological variation
(swsub), which describes the changes in concentration
about the subject’s true homeostatic set point. We have
now developed an automated process to select statistical
control rules and numbers of control measurements (N)
that will assure the quality required by clinical decision
interval (Dint) criteria or analytical total error (TEa) crite-
ria. The quality-planning models used in the automatic
process are essentially the same as those described earlier
[26, 27], except that the effect of replicate measurements
on method performance and QC design are more com-
pletely and directly considered by entering the number of
replicate samples analyzed. Here, we describe the auto-
matic process and demonstrate that it provides results
comparable with those in earlier studies documented in
this journal.
The computer program used, QC Validator Version 2.0,
runs under MicroSoft Windows 3.1 or Windows 95. It
requires an IBM or IBM-compatible computer with a 386
or higher processor, 1 MB hard disk space, 2 MB RAM
memory, a VGA interface and monitor, a printer sup-
ported by Windows, and TrueType fonts. The program
includes an installation utility, a detailed HELP facility,
and a manual that provides tutorials, a reference guide to
program function and operation, and a review of the
technical background for the program. The computer
program allows the user to enter information about the
characteristics of a method’s performance (e.g., smeas,
inaccuracy, expected frequency of errors, swsub) and the
quality required (clinically important change or Dint, TEa).
The user then initiates automatic selection on the basis of
the number of control materials to be analyzed (i.e., 1, 2,
or 3), and the program constructs a chart of operating
specifications (OPSpecs chart) that displays the selected
control rules and N. The automatic QC selection process is
based on user-editable criteria for the types of control
rules that can be implemented by the laboratory, the total
numbers of control measurements that are practical, the
maximum percentage of false rejections that can be toler-
ated, and the minimum percentage of error detection that
is acceptable for detection of medically important DSEcrit
or random errors.
The table of candidate QC procedures contains the
power curves and OPSpecs calculation parameters for 94
3
QC Validator and OPSpecs are registered trademarks of Westgard
Quality Corp. Windows is a registered trademark of Microsoft Corp.
 Clinical Chemistry 43, No. 2, 1997 401

different QC procedures, including: constant-limit single varied the control rules to match the error detection
rules 12s, 12.5s, 13s, and 13.5s, with N values from 1 to 8; capability to the test method performance. Table 1 shows
multirules such as 13s/22s/R4s/41s/10x# , with N values of 2 the test, total error requirement, observed imprecision,
and 4 applied over runs from 1 to 5; multirules such as calculated ‚SEcrit, and the QC procedure selected manu-
13s/2 of 32s/31s/12x# , with N values of 3 and 6 applied over ally from critical error graphs or selected automatically
runs from 1 to 4; multirules such as 13s/22s/R4s/41s/8x# , from OPSpecs charts. For the automatic selection process,
with N values of 8 applied over 1 run; variable-limit the frequency of errors parameter was set to ,2%, to
single rules such as 10.05 and 10.01, with N values from 2 to permit use of 50% Analytical Quality Assurance (AQA)
8; mean and range rules such as x# 0.05/R0.05, x# 0.01/R0.01, and charts in those situations where 90% AQA could not be
x# 0.002/R0.002, with N values from 2 to 8; and extended limit achieved.
rules such as 14s, 15s, and 16s, with N values from 1 to 8.
For 14 tests, both the manual and automatic ap-
The program includes an editor utility that allows users to
proaches selected the 13.5s rule with N 5 2; for another test
add power curves and OPSpecs calculation parameters
(albumin), both approaches selected the 12.5s rule with
for other rules and N values.
N 5 2. For two tests (chloride and total CO2), the 12.5s rule
OPSpecs charts were prepared for the examples illus-
with N 5 2 had been selected manually, whereas the
trated earlier by spreadsheet calculations of allowable
imprecision and allowable inaccuracy for both the clinical
model [26] and the analytical model [27]. Fig. 1 (top)
shows the OPSpecs chart for a cholesterol example, for
which Dint 5 20% and swsub 5 6.5%. The x-intercepts for
maximum allowable imprecision range from 2.4% to 3.5%,
which agrees with the range of 2.4% to 3.5% observed
previously (Fig. 2a in [26]). For individual QC procedures,
the largest difference was observed for the 13s rule with
N 5 4, for which the current intercept is 2.8% (vs the
previous estimate of 2.95%). Performing duplicate choles-
terol tests and using the average of the two measured
concentrations for diagnostic classification provides a
maximum allowable imprecision of 3.3% to 4.7%, com-
pared with the earlier estimate of 3.2% to 4.9% (Fig. 2b in
[26].
For a cholesterol example in which TEa is 10%, Fig. 1
(bottom) shows the estimated allowable imprecision (x-
intercepts) to be 1.9% to 2.6%, the same as the 1.9% to 2.6%
documented earlier (Fig. 4 in [27]). In the same example,
when error detection is optimized for random error rather
than systematic error, the allowable imprecision is esti-
mated as 1.2% to 2.3%, very similar to earlier estimates of
1.2% to 2.4% (Fig. 6 in [27]).
The differences between the program output and the
earlier spreadsheet calculations are the result of the dif-
ferent calculation parameters for the sizes of systematic
error that can be detected with the specified probability.
These parameters were estimated as the x-values that
correspond to the 0.90 intersection with the power curves
for the different control rules and N values, which in the
spreadsheet calculation were constructed on a point-to-
point basis and then later fitted with smoothed curves to
improve the estimates.
The automatic QC selection process was tested by
comparing the QC procedures selected by the program
that uses the OPSpecs methodology with those selected in Fig. 1. OPSpecs charts for (top) a clinical decision interval quality
an earlier study in which critical error graphs were used requirement of 20% and (bottom) a TEa quality requirement of 10%.
manually to select control rules and N values for 18 Both panels: The operating limits (top to bottom) correspond to the following QC
different analytes on a multitest chemistry analyzer [23]. procedures: 13s/22s/R4s/41s with N 5 4; 12.5s with N 5 4; 13s with N 5 4; 12.5s
with N 5 2; 13s/22s/R4s with N 5 2; and 13s with N 5 2. The operating point
In the earlier study, the design strategy fixed N at 2 per (bottom panel) represents the National Cholesterol Education Program’s recom-
run, considered only single-rule QC procedures, and mended 3% specifications for imprecision and inaccuracy.
402 Technical Briefs

automatic selection process identified a multirule proce-
dure with N 5 4. In these latter two cases, if the automatic
QC selection criteria were changed to restrict N to 2, then
the automated process would select a multirule procedure
with N 5 2; if the selection logic for 50% AQA charts was
changed to prefer single rules over multirules, then a 12.5s
rule with N 5 2 would be selected. That is, the automatic
QC process can be modified to implement the same
preferences used earlier in the process of manual selection
from critical error graphs. For one test (calcium), method
performance relative to the quality desired was so mar-
ginal that the assay required a special effort to make
duplicate measurements, which was necessary to improve
method performance and process control. The effect of
making duplicate measurements could be considered
directly with the automatic selection process, which then
identified a 12.5s rule with N 5 4 as necessary to provide
90% detection of medically important errors.
Therefore, although the QC designs and outcomes of
the automatic and manual selections were similar, one
thing was not: The time required to assess the QC designs
for these 18 tests and to document the selection with
printouts of OPSpecs charts was ;30 min with the auto-
matic QC selection process, whereas the earlier study took
several months.
In the future, one can envision an automatic QC
selection process integrated into laboratory instrument
software, on-line QC monitors, data management work-
stations, laboratory information systems, and automated
laboratory process control software, thereby permitting
laboratory scientists to focus on defining the quality
needed for a test, rather than worrying about the control
rules and numbers of control measurements that should
be used. When coupled with on-line data acquisition and
appropriate software to estimate method performance
characteristics, the information needed for QC design will
itself be automatically available. When linked to software
for on-line QC monitoring, the selected rules can be
implemented automatically. Thus, an integrated system
for analytical quality management is possible when an
automatic QC selection process is linked with method
performance data and on-line QC monitoring, allowing
the management of analytical quality with an appropriate
QC design based on current information on method
performance.
Such an integrated quality-management system pro-
vides the opportunity for a dynamic QC system that can
automatically adjust to changes in method performance,
e.g., changing to more-sensitive control rules and higher
N values if performance deteriorates and relaxing the
rules and reducing N if performance improves. Changes
in method stability can be factored into the dynamic QC
process and can also be used to adjust run length, as
suggested by recent design work on the application of
“average of normals” patient data algorithms to measure
process stability and determine when to requalify the
testing process [32]. Thus, technology for the next-gener-
ation system for analytical quality management can be
envisioned now that automatic QC selection is a reality.
Westgard Quality Corp. supported the development of
this software. Robert Kennedy performed the coding, and
Sten Westgard developed the HELP facility and program

Table 1. QC procedures selected by automated QC selection process using OPSpecs charts compared with those selected
manually by using critical error graphs.
Control rules and N selected

Test TEa, % Smeas, % DSEcrit Manually Automatically

Sodium 3.08 0.52 4.27 13.5s, N 5 2 Same
Potassium 10.0 1.17 6.90 13.5s, N 5 2 Same
Chloride 4.0 1.04 2.20 12.5s, N 5 2 MR, N 5 4
Total CO2 10.0 2.50 2.35 12.5s, N 5 2 MR, N 5 4
Glucose 8.0 1.20 5.02 13.5s, N 5 2 Same
Urea N 10.0 1.33 5.87 13.5s, N 5 2 Same
Creatinine 30.0 3.00 8.35 13.5s, N 5 2 Same
Calcium 5.0 1.68 1.33 Special case MR, N 5 4
Phosphorus 10.0 1.28 6.16 13.5s, N 5 2 Same
Uric acid 10.0 1.10 7.44 13.5s, N 5 2 Same
Cholesterol 10.0 1.35 5.76 13.5s, N 5 2 Same
Total protein 12.0 1.84 4.87 13.5s, N 5 2 Same
Albumin 10.0 2.13 3.04 12.5s, N 5 2 Same
Total bilirubin 20.0 2.20 7.44 13.5s, N 5 2 Same
GGT 10.0 1.17 6.90 13.5s, N 5 2 Same
ALP 10.0 1.17 6.90 13.5s, N 5 2 Same
AST 20.0 3.0 5.02 13.5s, N 5 2 Same
LD 20.0 3.0 5.02 13.5s, N 5 2 Same
MR, multirule; GGT, g-glutamyltransferase; ALP, alkaline phosphatase; AST, aspartate aminotransferase; LD, lactate dehydrogenase.
 Clinical Chemistry 43, No. 2, 1997
documentation. Additional information about the pro-
gram is available at www.westgard.com.
References
1. Tetrault GA, Steindel SJ. Qprobe 94 – 08. Daily quality control exception
practices. Chicago: College of American Pathologists, 1994.
2. Levey S, Jennings ER. The use of control charts in the clinical laboratory.
Am J Clin Pathol 1950;20:1059 – 66.
3. Henry RJ, Seagalov M. The running of standards in clinical chemistry and the
use of the control chart. J Clin Pathol 1952;5:305–11.
4. Westgard JO, Barry PL, Hunt MR, Groth T. A multi-rule Shewhart chart for
quality control in clinical chemistry. Clin Chem 1981;27:493–501.
5. Westgard JO, Groth T, Aronsson T, Falk H, de Verdier C-H. Performance
characteristics of rules for internal quality control; probabilities for false
rejection and error detection. Clin Chem 1977;23:1857– 67.
6. Westgard JO, Falk H, Groth T. Influence of a between-run component of
variation, choice of control limits, and shape of error distribution on the
performance characteristics of rules for internal quality control. Clin Chem
1979;25:394 – 400.
7. Westgard JO, Groth T. Power functions for statistical control rules. Clin Chem
1979;25:863–9.
8. Westgard JO, Groth T. Design and evaluation of statistical control proce-
dures: applications of a computer “quality control simulator” program. Clin
Chem 1981;27:1536 – 45.
9. Cembrowski GS, Westgard JO, Kurtyzc DFI. Use of anion gap for the quality
control of electrolyte analyzers. Am J Clin Pathol 1983;79:688 –96.
10. Cembrowski GS, Chandler EP, Westgard JO. Assessment of “average of
normals” quality control procedures and guidelines for implementation.
Am J Clin Pathol 1984;81:492–9.
11. Cembrowski GS, Westgard JO. Quality control of multichannel hematology
analyzers: evaluation of Bull’s algorithm. Am J Clin Pathol 1985;83:337–
45.
12. Lunetsky ES, Cembrowski GS. Performance characteristics of Bull’s multi-
rule algorithm for the quality control of multichannel hematology analyzers.
Am J Clin Pathol 1987;88:634 – 8.
13. Westgard JO, Barry PL. Cost-effective quality control: managing the quality
and productivity of analytical processes. Washington, DC: AACC Press,
1986:240pp.
14. Cembrowski GS, Carey RN. Laboratory quality management. Chicago: ASCP
Press, 1989:264pp.
15. Westgard JO, Klee GG. Quality assurance. Chapter 17 in: Burtis CA,
Ashwood ER, eds. Tietz textbook of clinical chemistry, 2nd ed. Philadelphia:
WB Saunders, 1994:548 –92.
16. Cembrowski GS, Sullivan AM. Quality control and statistics. Chapter 4 in:
Bishop ML, Duben-Engelkirk JL, Fody EP, eds. Clinical chemistry: principles,
procedures, correlation, 3rd ed. Philadelphia: Lippincott, 1996:61–98.
17. Parvin CA. Estimating the performance characteristics of quality-control
procedures when error persists until detection. Clin Chem 1991;37:
1720 – 4.
18. Parvin CA. Comparing the power of quality-control rules to detect persistent
systematic error. Clin Chem 1992;38:358 – 63.
19. Parvin CA. Comparing the power of quality-control rules to detect persistent
increases in random error. Clin Chem 1992;38:364 –9.
20. Parvin CA. New insight into the comparative power of quality-control rules
that use control observations within a single analytical run. Clin Chem
1993;39:440 –7.
21. Smith FA, Kroft SH. Exponentially adjusted moving mean procedure for
quality control: an optimized patient sample control procedure. Am J Clin
Pathol 1996;105:44 –51.
22. Linnet K. Choosing quality-control systems to detect maximum clinically
allowable analytical errors. Clin Chem 1989;35:284 – 8.
23. Koch DD, Oryall JJ, Quam EF, Feldbruegge DH, Dowd DE, Barry PL, Westgard
JO. Selection of medically useful quality-control procedures for individual
tests done in a multitest analytical system. Clin Chem 1990;36:230 –3.
24. Westgard JO, Oryall JJ, Koch DD. Predicting effects of QC practices on the
cost-effective operation of a multitest analytical system. Clin Chem 1990;
36:1760 – 4.
25. Westgard JO, Quam EF, Barry PL. Quality control selection grids (QCSGs) for
planning QC procedures. J Clin Lab Sci 1990;3:271– 8.
26. Westgard JO, Hyltoft Petersen P, Wiebe DA. Laboratory process specifica-
tions for assuring quality in the US National Cholesterol Education Program.
Clin Chem 1991;37:656 – 61.
27. Westgard JO, Wiebe DA. Cholesterol operational process specifications for
assuring the quality required by CLIA proficiency testing. Clin Chem 1991;
37:1938 – 44.
28. Westgard JO. Assuring analytical quality through process planning and
quality control. Arch Pathol Lab Med 1992;116:765–9.
29. Westgard JO. Charts of operational process specifications (“OPSpecs
403
charts”) for assessing the precision, accuracy, and quality control needed to
satisfy proficiency testing criteria. Clin Chem 1992;38:1226 –33.
30. Westgard JO. OPSpecs manual, expanded ed. Operating specifications for
precision, accuracy, and quality control. Ogunquit, ME: Westgard QC,
1996:240pp.
31. Westgard JO. A program for evaluating QC procedures. Med Lab Observ
1994;26(2):55– 60.
32. Westgard JO, Smith FA, Mountain PJ, Boss S. Design and assessment of
average of normals (AON) patient data algorithms to maximize run lengths
for automatic process control. Clin Chem 1996;42:1683– 8.
Rapid Screening for a1-Antitrypsin Z and S Mutations,
Christopher W.K. Lam, Chi-Pui Pang,* Priscilla M.K. Poon,
Chang-Hong Yin, and Geetha Bharathi (Dept. of Chem.
Pathol., Chinese Univ. of Hong Kong, Prince of Wales
Hosp., Shatin, N.T., Hong Kong; *author for correspon-
dence: fax 852 26365090, e-mail cppang@cuhk.hk)
a1-Antitrypsin (A1AT) is a serine protease inhibitor re-
quired for the prevention of proteolytic tissue damage,
principally in the lung, by neutrophil elastase released by
inflammatory cells [1]. While severe A1AT deficiency is
the major factor leading to emphysema and related pul-
monary diseases, it is also associated with neonatal hep-
atitis and cirrhosis [1, 2]. A1AT deficiency is an autosomal
codominant disorder with a prevalence of about 1:3000 in
Caucasians [3]. The A1AT gene has 7 exons spanning ;12
kb. The most common gene defect resulting in A1AT
deficiency is that of a protease inhibitor (PI)-system Z
mutation Glu342 to Lys, which is a single base substitution
of G to A in exon 5 [4, 5]. The S mutation, a Glu264 to Val
change, is caused by an A to T substitution in exon 3 [6].
Individuals with SS are unaffected, SZ may be symptom-
atic, and ZZ results in the most severe clinical symptoms.
In Caucasians the prevalence of the S allele ranges from
5% to 10% and Z allele 2% to 5% depending on geograph-
ical location [7, 8]. Although the frequencies are unknown
in the Chinese, geographical variability of the A1AT
alleles is evident by phenotypic analysis of the PI variants
[9]. We have established a rapid screening procedure
involving multiplex PCR to detect the Z and S mutations
in Chinese in Hong Kong who came from southern China.
EDTA–whole-blood specimens were obtained from lo-
cal Chinese in Hong Kong who attended the Prince of
Wales Hospital for routine checkup or for treatment of
diabetes mellitus. Genomic DNA was extracted from the
blood specimens by the salting-out method [10]. Our
procedure for mutation analysis was modified from the
PCR-mediated site-directed mutagenesis method of Taze-
laar et al. [11] with their primers for the Z and the S
mutations, labeled as primers ZF and ZR and primers SF
and SR respectively. Each PCR mixture, in a final volume
of 25 mL, contained 0.2 mmol/L deoxynucleoside triphos-
phates (Boehringer Mannheim, Mannheim, Germany), 1.5
mmol/L magnesium chloride, 0.1 g/L gelatin, 2.5 pmol
each of primers ZF, ZR, SF, and SR (synthesized by Gibco
BRL, Gaithersburg, MD), 0.2 mg of DNA, 0.5 U of Taq
polymerase (Gibco BRL), and 13 PCR buffer from Gibco
404 Technical Briefs
BRL. After an initial denaturation at 94 °C for 5 min, a
35-cycle PCR program was carried out on a Perkin-Elmer
(Norwalk, CT) thermal cycler: 94 °C for 1 min, 55 °C for 1
min, and 72 °C for 2 min; final extension at 72 °C for 10
min. After PCR, the restriction digestion mixture was
prepared: 10 mL of PCR product, 1 U of Taq I restriction
endoclease (Gibco BRL) [12], and buffer to a final volume
of 15 mL. Restriction digestion was completed after 2 h of
incubation at 65 °C. The digested PCR products were
analyzed by 3% agarose electrophoresis at constant volt-
age of 200 V for 1 h. Control samples of known genotype
were electrophoresed in each gel along with a size cali-
bration mixture. For fast screening, 0.2 mg of DNA from
five individuals were added to a PCR mixture for ampli-
fication and subsequent Taq I restriction digestion. When
an abnormal result appeared, the individual DNA speci-
mens were analyzed separately to identify the DNA
sample carrying the mutant allele. To validate this proto-
col, all the DNA specimens in this study were analyzed
individually and in groups of five. Identical results were
obtained.
We obtained DNA specimens from 2005 unrelated
Chinese subjects free of primary lung and liver diseases.
Two individuals were found to be heterozygous for Z, i.e.,
MZ, and another two heterozygous for S, i.e., MS (Fig. 1).
No SS, SZ, or ZZ were found. There were therefore two Z
and two S alleles from a total of 4010 alleles, giving a
frequency of 0.05% for both the Z and S mutations and
0.1% for the MZ and MS genotypes in the Chinese
population. Combined with simultaneous phenotyping
observation that showed absence of other mutations, our
study suggests that 99.8% of the Chinese are of the MM
genotype.
Fig. 1. Taq I endonuclease restriction analysis of A1AT alleles after
amplification with Z and S primers in the same reaction mixture.
Lanes (a) to (c) were PCR products of 5 DNA templates amplified simultaneously:
(a) only M alleles were detected, (b) a Z allele was detected, and (c) an S allele
was detected. Lanes (d) to (h) were PCR products of single DNA templates
amplified individually: (d) an MS control, (e) an MM homozygote, (f) an MM
control specimen, (g) an SS control, and (h) a ZZ control. (M) is a DNA marker of
pBR HaeIII digest.
Such a dominance of the MM genotype in Chinese is
unexpected, although people of Asian origin are known to
have fewer Z mutations than Caucasians. While northern
Europeans have a higher prevalence of the Z genotypes
than southern Europeans, the Z and S mutations range
from 1% to 4% and from 5% to 10% respectively in most
Caucasian populations [7, 13]. In the Chinese, even the
MZ and MS heterozygotes are less prevalent, at 0.5%,
similar to the SS or ZZ homozygotes, at 0.25% and 0.3%
respectively, of the British population [13]. To ascertain
the association between the Z mutation and A1AT phe-
notypes in the Chinese population, we are analyzing the
genotypes and phenotypes of normal Chinese subjects
and of patients with emphysema. Meanwhile, we have
shown our protocol of a double PCR with 5 DNA tem-
plates to be rapid, reliable, and economical for screening a
large number of specimens directly for the Z and S
mutations. There is, however, a limitation in this ap-
proach of batch analysis. If a single DNA sample in the
batch of five samples failed to amplify by PCR, it would
not be recognized. If this sample happened to be from a
patient with a mutation, it would have been missed. In
our individual analysis of 2005 samples, we did not have
any sample that failed to amplify, showing the PCR
protocol to be robust.
We thank Kathy Piesse of the Royal Prince Alfred Hospi-
tal, Sydney, Australia and N.A. Kalsheker of the Univer-
sity of Nottingham, UK, for their supply of specimens of
confirmed M, S, and Z genotypes.
References
1. Cox DW. a1-Antitrypsin deficiency. In: Scriver CR, Beaudet AL, Sly WS, Valle
D, eds. The metabolic and molecular bases of inherited disease, 7th ed.
New York: McGraw-Hill, 1995:4125–58.
2. Blank CA, Brantly M. Clinical features and molecular characteristics of
a1-antitrypsin deficiency. Ann Allergy 1994;72:105–20.
3. Sveger T. Liver disease in alpha1-antitrypsin deficiency detected by screen-
ing of 200,000 infants. N Engl J Med 1976;294:1316 –21.
4. Long GL, Chandra T, Woo SLC, Davie EW, Kurachi K. Complete sequence of
the cDNA for human a1-antitrypsin and the gene for the S variant. Biochem
1984;23:4828 –37.
5. Lomas DA, Evans DL, Stone SR, Chang WSW, Carrell RW. Effect of the Z
mutation on the physical and inhibitory properties of a1-antitrypsin. Biochem
1993;32:500 – 8.
6. Jeppsson J-O. Amino acid substitution Glu–Lys in alpha-1-antitrypsin PiZ.
FEBS Lett 1976;65:195–7.
7. Hjalmarsson K. Distribution of alpha-1-antitrypsin phenotypes in Sweden.
Hum Hered 1988;38:27–30.
8. Dykes DD, Miller SA, Polesky HF. Distribution of alpha1-antitrypsin variants
in a US white population. Hum Hered 1984;34:308 –10.
9. Ying QL, Zhang ML, Liang CC, Liu XP, Huang YW, Wang RX, et al.
Geographical variability of alpha-1-antitrypsin alleles in China: a study on six
Chinese populations. Hum Genet 1985;69:184 –7.
10. Miller SA, Dykes DD, Polesky HF. A simple salting out procedure for
extracting DNA from human nucleated cells. Nucleic Acids Res 1988;16:
1215.
11. Tazelaar JP, Friedman KJ, Kline RS, Guthrie ML, Farber RA. Detection of
a1-antitrypsin Z and S mutations by polymerase chain reaction-mediated
site-directed mutagenesis. Clin Chem 1992;38:1486 – 8.
12. Dry PJ. Rapid detection of alpha-1-antitrypsin deficiency by analysis of a
PCR-induced TaqI restriction site. Hum Genet 1991;87:742– 4.
13. Hutchison DCS. The epidemiology of a1-antitrypsin deficiency. Lung
1990;168(Suppl):535– 42.
 Clinical Chemistry 43, No. 2, 1997
Quantitative Immunological Detection of Total Estro-
gen Receptor (Cytosolic and Nuclear) in Term Decidua
of Preeclampsia: a Preliminary Study, Sanaa Eissa,1*
Mohamed M. Mostafa,2 Alaa A.E.A. El-Gendy,3 and Ibrahim
A. Senna3 [1 Oncology Diagnostic Unit, Biochem. Dept.
(*address for correspondence: fax 202-285-9928), 2 Dept. of
Gyn., Ain Shams Faculty of Med., Abbassia, Cairo, Egypt,
and 3 Fayoum General Hospital, Egypt]
Although progesterone and estrogens are essential for
maintaining human pregnancy after implantation, few
reports have investigated the localization of their specific
receptors in different uterine cell types throughout preg-
nancy [1, 2]. Wu et al. detected estrogen receptors (ER) in
low concentrations in decidua in early pregnancy but not
in term decidua [2]. ER concentrations in term decidua of
pregnancies with complications (in particular, preeclamp-
sia) have not been investigated.
Classical biochemical assay methods estimate ER either
in cytosol fraction or in the high-salt extracted nuclei
[3–5]. These assays need two subcellular fractionation
steps and two ER assays in two fractions: cytosol and
nuclear extract. To measure total ER (cytosolic and nu-
clear) in one fraction using one single assay in decidual
samples, we modified the assay method by using a
homogenization buffer containing KCl. The effective KCl
concentration for maximum ER extraction without inhibi-
tion of the ER enzyme immunoassay (EIA) ranged from
0.4 – 0.8 mol/L KCl.
All steps of sample preparation in our laboratory were
carried out at 4 °C, all reagents were purchased from
Sigma Chemical Co. (St. Louis, MO), and all women
included in the study gave informed consent. We used
405
scrapped decidual samples from placental bed from
women with uncomplicated pregnancies who underwent
cesarean section for obstetrical causes (n 5 10) and from
women with preeclampsia (n 5 20): mild (n 5 4), mod-
erate (n 5 6), and severe (n 5 10). Tissues were immedi-
ately washed in ice-cold saline and homogenized at 1
g/10 mL in ice-cold homogenization buffer (10 mmol/L
Tris buffer, pH 7.5, containing 10 mmol/L K2EDTA, 100
mL/L glycerol, 5 mmol/L benzamidine, 10 mmol/L
2-mercaptoethanol, 0.39 mmol/L phenylmethylsulfonyl
fluoride, and 5 mg/L aprotinin) with Ultraturax T-25
homogenizer for five bursts of 1 min each, separated by a
1-min pause. The homogenate was filtered and divided
into two parts. The first part was centrifuged in a Beck-
man CS-6R centrifuge (Brea, CA) at 800g for 15 min at 4 °C
to obtain the crude nuclear pellet. The supernatant fluid
was recentrifuged at 100 000g for 1 h with a Beckman L7
ultracentrifuge to obtain the cytosol. The crude nuclear
pellet was dissolved in 10 mmol/L ice-cold phosphate
buffered saline (pH 7.2) by sonication for three 30-s bursts
and recentrifuged at 800g for 15 min at 4 °C. The washed
nuclear pellet was incubated with 5 volumes of ice-cold
high-salt extraction buffer (homogenization buffer con-
taining 0.4 mol/L KCl) on ice for 30 min with vortex-
mixing every 10 min. Then it was ultracentrifuged in a
Beckman L7 ultracentrifuge at 100 000g for 30 min at 4 °C.
The nuclear extract (supernatant) was obtained.
The second part of the homogenate was incubated with
0.4 mol/L KCl on ice for 30 min with vortex-mixing every
10 min, then ultracentrifuged in a Beckman L7 ultracen-
trifuge at 100 000g for 30 min at 4 °C. The supernatant was
isolated. After quantifying the protein concentration in
the cytosol, nuclear extract, and tissue extract by using
Fig. 1. (A) Total ER in term
decidua in normal pregnancy
and mild, moderate, and se-
vere preeclampsia; (B) cytoso-
lic and nuclear ER in term de-
cidua in severe preeclampsia.
Dotted line represents the lower
detection limit for immunological
detection of ER.
406 Technical Briefs
Bradford’s method [6] with bovine serum albumin as the
calibrator, we applied the samples directly to the ELISA
plate for assay of ER with EIA kit from Abbott Laborato-
ries (Chicago, IL) [7].
ER concentrations estimated by our modified method
strongly correlated to those calculated as a sum of ER
concentration in cytosol plus those in nuclear extract (y 5
20.98 1 1.128x, r 5 0.91, P 5 0.005). Cytosolic, nuclear,
and total ER were not detectable (lower detection limit, 1
fmol/mg protein) in term decidua from women with
uncomplicated pregnancies or from women with mild or
moderate preeclampsia, but were detected in term de-
cidua from women with severe preeclampsia (Fig. 1)
(cytosolic ER, 1.4 –9 fmol/mg protein, mean 4.98 fmol/mg
protein; nuclear ER, 1.5–11 fmol/mg protein, mean 4.68
fmol/mg protein; total ER 2.6 –19.2 fmol/mg protein,
mean 9.1 fmol/mg protein). This difference was statisti-
cally significant by Student’s t-test (P 5 ,0.01).
Although our results suggest a significant association
between ER detection in term decidua and severe pre-
eclampsia, the cause– effect relationship cannot be deter-
mined in this study. Establishing a cause– effect relation-
ship may improve understanding of the biological
relationship and may have potential clinical implications
through development of diagnostic or therapeutic mark-
ers for these lesions. We recommend the routine use of
this modified assay for quantification of total ER in the
clinical laboratory.
We thank A. Khalifa for his unlimited support.
References
1. Perrot-Applanat M, Deng M, Fernandez H, Lelaidier C, Meduri G, Bouchard P.
Immunohistochemical localization of estradiol and progesterone receptors
in human uterus throughout pregnancy: expression in endometrial blood
vessels. J Clin Endocrinol Metab 1994;78:216 –24.
2. Wu WX, Brooks J, Millar MR, Ledger WL, Glasier AF, McNeilly AS. Immuno-
localization of estrogen and progesterone receptors in the human decidua in
relation to prolactin production. Hum Reprod 1993;8:1129 –35.
3. Kokko E, Janne O, Kauppila A, Vihko R. Effects of tamoxifen, medroxy
progesterone acetate and their combination on human endometrial estrogen
and progestin receptor concentrations. Am J Obstet Gynecol 1982;143:
382–5.
4. Clarck JH, Peck EJ Jr. Nuclear retention of receptor– estrogen complex and
nuclear receptor sites. Nature 1976;260:635–9.
5. Thorpe SM, Lykkesfelt AE, Vinterby A, Lonsdorfer M. Quantitative immuno-
logical detection of estrogen receptors in nuclear pellets from human breast
cancer biopsies. Cancer Res 1985;46(Suppl):4251.
6. Bradford MM. A rapid and sensitive method for quantitation of microgram
quantities of protein utilizing the principle of protein dye binding. Anal Bioch
1976;72:248 –51.
7. Greene GL, Nolan C, Engler JP, Jensen EV. Monoclonal antibodies to human
estrogen receptor. Proc Natl Acad Sci U S A 1980;77:5115–9.
Speciation of Arsenic in Serum, Urine, and Dialysate of
Patients on Continuous Ambulatory Peritoneal Dialysis,
Xinrong Zhang,1 Rita Cornelis,1* Jurgen De Kimpe,1 Louis
Mees,1 and Norbert Lameire2 (1Lab. for Anal. Chem., Inst.
for Nuclear Sci., Univ. of Gent, Proeftuinstraat 86, Gent,
Belgium; 2Renal Div., Dept. of Med., Univ. Hosp., De
Pintelaan 185, B-9000 Gent, Belgium; *author for corre-
spondence: fax 32 (0)9 2646699, e-mail Cornelis@
inwchem.rug.ac.be)
Renal replacement therapy is currently achieved by hemo-
dialysis (HD), hemofiltration, hemodiafiltration (HDF), con-
tinuous ambulatory peritoneal dialysis (CAPD), or renal
transplantation. The concentration of trace elements in se-
rum of patients could be influenced by these treatments. A
significant increase of As concentrations in serum of patients
on HD treatment has been reported [1, 2]. As concentrations
higher than the reference value were also observed in serum
of patients on HDF [3]. Although there are several studies
describing the status of trace elements in serum, plasma, and
dialysate in CAPD patients [4, 5], no data on As serum
concentrations of CAPD patients are available. The aim of
this work is to determine total As concentrations and to
speciate As species in serum, urine, and dialysate of CAPD
patients.
Fourteen CAPD patients were studied. Serum, urine,
and dialysate samples were collected from the University
Hospital. Patients gave their informed consent before
blood sampling. To decrease the influence of As intake
from the diet, patients were requested to refrain from
ingesting seafood during the 3 days before blood and
urine collection. The reagents and apparatus for the
separation and measurement of As species and for the
measurement of total As have been described [6]. Briefly,
two types of HPLC columns were used for the separation
of anionic and cationic As species: an anion exchange
column (Supelcosil LC-SAX, 250 3 4.6 mm; Supelco,
Bellefonte, PA) and a cation exchange column [Dionex
(Sunnyvale, CA) Ionpac® CS 10, 250 3 4 mm]. A Perkin-
Elmer (Norwalk, CT) 3030 atomic absorption spectrome-
ter was used throughout for the detection of As signals.
Analysis of serum creatinine was performed according to
the Jaffe method [7].
The accuracy of the total As measurements was tested
by simultaneously analyzing a Certified Freeze-Dried
Reference Serum (CRM) of the University of Gent, Bel-
gium. The accuracy of the As speciation measurements
was tested by analyzing a BCR candidate Reference
Material CRM 526 tuna tissue, as no serum reference
material is yet available for the As speciation study. No
significant differences were established between the ana-
lytical results and the certified values. The precision (CV)
of the method for 10 replicate analyses of aqueous solu-
tion of As species at a concentration of 10.0 mg/L was
always better than 5% for each form of As. The 3-day
run-to-run precision of measurement of 10.0 mg/L As
species added to serum was better than 10%.
The mean total As concentration in the serum of 14
CAPD patients is 4.67 6 5.41 mg/L, significantly higher (P
,0.001) than the reference values of 0.96 6 1.52 mg/L in
serum of healthy subjects previously obtained in our
laboratory [8], indicative of an accumulation of As in
serum of CAPD patients. The main As species in the
serum of these patients are dimethylarsinic acid (DMA)
and arsenobetaine (AsB), respectively carrying 15.2%
 Clinical Chemistry 43, No. 2, 1997
(0.71 6 1.05 mg/L) and 76.2% (3.56 6 4.27 mg/L) of total
As. No significant differences were observed between
CAPD patients and previously reported nondialyzed ure-
mic patients (0.82 6 1.05 mg/L DMA and 3.55 6 4.58
mg/L AsB) [6]. The DMA concentrations of CAPD pa-
tients are, however, significantly lower than those of HD
patients (1.93 6 1.51 mg/L, 29.8% of total As, P ,0.001)
[6], although no significant difference of AsB concentra-
tion was observed between these two groups (3.56 6 4.27
vs 3.47 6 2.89 mg/L, P 5 0.5658). Fig. 1(top) shows the
comparison of As species concentration in the serum of
the three groups of patients (uremic nondialyzed, CAPD,
and HD). The AsB data imply that the main source of As
accumulation in serum of CAPD patients comes from the
diet, as AsB is particularly present in fish and seafood.
Significantly lower DMA concentration in CAPD patients
than in HD patients indicates that the CAPD treatment is
probably more efficient in removing the toxic inorganic
As species from the blood and better avoids the in vivo
methylation in the body since this treatment is a contin-
uous procedure compared with the intermittent proce-
dure of HD.
The mean concentrations of DMA, AsB, and total As in
urine of 10 CAPD patients with residual urine production
are 2.24 6 1.80, 9.09 6 8.50, and 12.28 6 10.66 mg/L,
respectively. These concentrations are significantly higher
407
than the serum values of these patients (P 5 0.0191, 0.009,
and 0.040, respectively). However, considering the rela-
tively small amount of urine produced by these patients
(mean 758 mL/24 h) vs the overall blood volume (esti-
mated mean 5000 mL/per patient, mean hematocrit
34.1%) and the volume of drained dialysate (mean 8680
mL/24 h), the absolute urinary excretion of As is very
low. The mean amounts 6 SD in the total urine are: DMA
1.34 6 0.98 mg, AsB 4.62 6 2.57 mg, and total As 6.30 6
3.37 mg, and represent only 10% of total As distributed in
the four compartments (serum 23%, packed cells 14%,
dialysate 53%). Calculation of the ratio of DMA/AsB
showed no significant difference between urine and se-
rum (P 5 0.1456) or dialysate (P 5 0.0845). This suggests
nonselectivity of renal removal of DMA, a more toxic As
species than AsB.
A very low As concentration was found in fresh dialy-
sate (0.04 mg/L). The mean concentrations are, however,
increased to 3.98 6 4.91 (total As), 0.59 6 0.87 (DMA), and
3.06 6 3.96 (AsB) mg/L in 24-h dialysate. No significant
differences in As concentrations were found between
serum and drained dialysate after 24-h dialysis (P 5
0.4482, 0.5984, and 0.5770 for total As, DMA, and AsB,
respectively), indicating a nonselective accumulation or
removal of different As species. These results can be easily
understood because CAPD means a slow equilibrium
Fig. 1. Comparison of (top)
mean total As and As species
concentrations in serum of
healthy subjects (HS) (n 5 23),
uremic nondialyzed (ND) (n 5
19), CAPD (PD) (n 5 14), and
hemodialysis (HD) (n 5 18) pa-
tients, and (bottom) mean total
As and As species concentra-
tions in serum (SE) (n 5 14),
dialysate (DL) (n 5 14), and
urine (UR) (n 5 10) of CAPD
patients.
408 Technical Briefs
dialysis. The low-molecular-mass species of AsB and
DMA are easily transferable across the peritoneal mem-
brane, and equilibrium in concentrations is readily
achieved. The As species in dialysate are significantly
lower than those in urine (P ,0.005 for each As species)
because of the concentrating capacity of the kidney. The
analysis of total As and As species on each of the four
exchanges of CAPD dialysate showed no significant dif-
ferences in concentrations on 1 day for the same patient.
Considering a mean volume of 8680 6 912 mL of drained
dialysate per day, the mean amount of total As is 35.1 6
45.7 mg per patient/24 h. Fig. 1 (bottom) compares the
concentrations of total As and the As species in the three
different fluids (serum, dialysate, and urine).
Neither inorganic As species of As(III) and As(V) nor
methylmalonic acid and AsC could be detected in serum,
urine, and dialysate of the CAPD patients, with one
exception. Patient 5 showed measurable concentrations of
As(III) and As(V) in urine, but not in her 24-h dialysate.
The possibility that the urine was contaminated with
inorganic As species during collection cannot be ex-
cluded, although a very clean collector was used for the
urine, and the patient was instructed about the collection.
As this result was only observed in one specimen, more
patients and urine samples are needed to confirm this
finding.
By dividing the subjects into the groups according to
the serum creatinine concentrations, we found that the
group with significantly higher serum creatinine concen-
tration has a significantly higher As concentration
(P ,0.05). The creatinine concentrations for the groups of
healthy subjects and CAPD and HD patients were respec-
tively 80 6 25, 646 6 244, and 914 6 173 mmol/L, whereas
the As concentrations in the three groups were respec-
tively 0.96 6 1.52, 4.67 6 5.41, and 6.47 6 4.28 mg/L. A
similar conclusion was obtained previously when we
studied the accumulation of As in the serum of uremic
patients on conservative treatment and of patients on
hemodialysis treatment [6, 9].
We thank M.C. Lambert for the assistance on the collec-
tion of the samples.
References
1. De Kimpe J, Cornelis R, Mees L, Van Lierde S, Vanholder R. More than
tenfold increase of As in serum and packed cells of chronical hemodialysis
patients. Am J Nephrol 1993;13:429 –34.
2. Astrug A, Kuleva V, Kiriakov Z, Tomov A, Djingova R. Trace elements in blood
and plasma of patients with chronic renal failure treated with maintenance
haemodialysis. Trace Elem Med 1984;1:65–70.
3. Van Renterghem D, Cornelis R, Vanholder R. Behaviour of 12 trace elements
in serum of uraemic patients on hemodiafiltration, J Trace Elem Electrolytes
Health Dis 1992;6:169 –74.
4. Wallaeys B, Cornelis R, Mees L, Lameire N. Trace elements in serum,
packed cells, and dialysate of CAPD patients. Kidney Int 1986;30:599 –
604.
5. Cornelis R, Ringoir S, Mees L, Lameire N, Wallaeys B, Hoste J. Trace
element patterns in blood of patients with renal failure. In: McHowell J,
Gawthorne JM, White CL, eds. Trace element metabolism in man and
animals. Canberra: Australian Academy of Science, 1981:530 –3.
6. Zhang X, Cornelis R, De Kimpe J, Mees L, Vanderbiesen V, De Cubber A,
Vanholder R. Accumulation of arsenic species in serum of patients with
chronic renal disease. Clin Chem 1996;42:1231–7.
7. Jaffe M. Ueber den Niederschlag welchen Pikrinsäure in normalen Harn
erzeugt und ueber eine neue Reaktion des Kreatinins. Z Physiol Chem
1886;10:391– 400.
8. Versieck J, Vanballenberghe L. Determination of arsenic and cadmium in
human blood serum and packed cells. Trace Elem Man Anim 1985;5:
650 –5.
9. Zhang X, Cornelis R, De Kimpe J, Mees L, Vanderbiesen V, Vanholder R.
Total arsenic determination in serum and packed cells of patients with
chronic renal insufficiency. Fresenius J Anal Chem 1995;353:143–7.
Lipemia Interference with a Rate-Blanked Creatinine
Method, Glen L. Hortin* and Katherine Goolsby (Dept. of
Pathol., Univ. of Alabama at Birmingham, 618 S. 18th St.,
Birmingham, AL 35233; *author for correspondence: fax
205-975-4468, e-mail hortin@wp.path.uab.edu)
The reaction of picric acid (2,4,6-trinitrophenol) with
creatinine under strongly alkaline conditions, originally
described by Jaffe .100 years ago [1], remains the most
common means of measuring creatinine in clinical sam-
ples. However, picric acid reacts with many compounds
besides creatinine to yield colored products [1–5]. Ke-
tones, glucose, proteins, cephalosporins, and other com-
pounds react to yield positive interferences; also, de-
creases in the color of bilirubin and hemoglobin under the
reaction conditions yield negative interferences [2–7]. To
counter these interferences, analysts have tried many
variations of the Jaffe reaction, including modifications of
sample preparation, addition of oxidizers to remove in-
terferents, solvent extraction of interferents, pH adjust-
ment of the reaction, continuous-flow dialysis, and kinetic
measurement of the reaction [2–5]. In kinetic methods, the
most common approach to decreasing interferences, the
timing of the absorbance readings can be selected to
minimize the effects of components that react faster or
more slowly than creatinine. For most samples, these
modifications of the alkaline–picric acid methods have
improved the specificity of measurement. Results have
become closer to the “true” creatinine values determined
with reference methods [2–5]—improvements in mea-
surement that have led to a downward adjustment of
creatinine reference intervals by ;3 mg/L [8, 9] (to ex-
press creatinine in mmol/L, multiply mg/L by 8.84).
Despite improvements, substantial problems of inter-
ference with creatinine measurements remain [2–5]. Re-
cently, a rate-blanking method has been introduced that
compensates for absorbance changes of bilirubin and
hemoglobin under the strongly alkaline reaction condi-
tions [6, 7]. Although this modification decreases the
interference from bilirubin and hemoglobin, we find that
it introduces a new interference from lipemic samples.
Introduction of a new rate-blanked compensated creat-
inine method from Boehringer-Mannheim Corp. (India-
napolis, IN) was considered advantageous in our labora-
tory because the method decreases negative interference
from chromogens such as bilirubin and hemoglobin [6, 7].
However, we noted unexpectedly low creatinine values
for several patients after introduction of the method. In
our clinic laboratory, occasional samples—all lipemic—
 Clinical Chemistry 43, No. 2, 1997 409

yielded negative values, and other samples yielded creat- range of 5–15 mg/L, where highly accurate measure-
inine values substantially below what was clinically ex- ments of serum creatinine are needed for calculating
pected from the patient’s previous values, body mass, and creatinine clearance, monitoring renal transplant patients,
clinical status. Use of some of these measured serum and detecting early stages of kidney dysfunction from
creatinine values to calculate creatinine clearance yielded diabetes and other etiologies [2, 10].
clearance values of 200 mL/min per 1.73 m2 body surface Differences between the two Jaffe methods for lipemic
area or more, which are inconsistent with usual reference samples were not related to differences in calibration. The
values [2, 10]. Nonetheless, the manufacturer claims the same calibration material was used, and method compar-
method is not subject to significant interference from ison studies showed close agreement for nonlipemic sam-
lipemia, as tested by adding a synthetic lipid emulsion to ples; analysis of 17 samples with triglyceride ,4 g/L
samples to give a final triglyceride concentration of as (lipemic index range 5– 63, mean 20) during the same
much as 10 g/L [7]. At this concentration, lipid emulsions week as analysis of lipemic samples showed no difference
showed a small negative interference (;1 mg/L decrease .1 mg/L for comparative analyses by the two methods.
in creatinine) for creatinine concentrations near the upper The interference of lipemic samples in the rate-blanked
limit of normal (;13 mg/L). Any effect ,10% was con- compensated method appeared to relate to nonlinear
sidered not significant in the evaluation study [7]. changes in measured absorbance by lipoproteins during
To investigate the problem of interference by lipemic the analysis. After addition of the first reagent containing
samples, we compared creatinine values obtained with NaOH, there was a curvilinear increase of absorbance—
three different methods—the rate-blanked compensated rapid at first during the period used to rate-blank the
method from Boehringer-Mannheim, a compensated ki- absorbance changes in the sample, but slowing before
netic method from Boehringer-Mannheim, and an enzy- addition of the second reagent (containing picric acid).
matic method on the Vitros 700 analyzer from Johnson Thus, the blanking for baseline absorbance changes was
and Johnson (Rochester, NY)—for 17 lipemic samples excessive. For the rate-blanking to be accurate, the base-
received in our clinic laboratory during 1 week. The two line absorbance changes (without picric acid) would need
Jaffe reactions were performed simultaneously with the to be linear throughout the entire reaction. Identifying the
same lot of reagents on different channels of the same problem as involving the rate-blanking step explains why
Hitachi 747 analyzer. Lipemic samples were identified by a simpler compensated method using the same reagents
the lipemic index reported by the instrument (an index of does not have similar lipemia interference. Ways to avoid
changes in absorbance measurements attributable to light the lipemia interference include using Jaffe methods with-
scattering by lipoprotein particles of a specimen diluted in out rate blanking, using enzymatic methods, or storing
isotonic saline). Because the lipemic index depends on samples to allow separation of lipoproteins.
lipoprotein size as well as quantity, it does not correlate Interferences from lipemia are difficult to study, given
exactly with triglyceride concentrations. We also deter-
mined triglycerides in some samples as another measure
of lipemia, using the Hitachi 747 and a coupled enzymatic Table 1. Comparison of creatinine concentrations in lipemic
method from Boehringer-Mannheim without blanking for samples measured by three methods.
glycerol. As listed in Table 1, the rate-blanked compen- Concn., mg/L
sated method (Jaffe 1) provided values that averaged 5 Lipemic Triglycerides,
Jaffe 1 Jaffe 2 Enzymatic indexa g/L
mg/L lower than another compensated method (Jaffe 2)
24 5 6 634 44.3
and 4 mg/L lower than the enzymatic method. For some
3 11 12 400 .17.0
samples, the difference between the rate-blanked method
11 16 16 215
and the other comparative methods was as much as 10
11 17 16 213 16.2
mg/L. However, there was good agreement between the
8 12 11 139 7.8
compensated method and the enzymatic method for the
14 15 15 139 4.9
lipemic samples.
10 15 13 131
The magnitude of interference by lipemia in the rate-
2 14 10 120
blanked creatinine method did not correlate closely with
13 18 16 111 6.7
the lipemic index or triglyceride concentration. One of the
9 11 11 110
problems with this interference is that it is difficult to
13 17 16 104 5.5
identify a cutoff point below which interference from
6 11 9 90
lipemia does not occur. We routinely analyze by an
9 14 11 84
alternative method samples with a lipemic index .65;
11 15 13 81
;0.5–1% of clinic samples have a lipemic index above this
18 22 22 67 3.1
value, so several samples per day require alternative
20 20 20 63 2.0
analysis. The frequency of lipemic samples from our clinic
5 11 10 63 14.4
is likely to be high relative to that for inpatient settings,
Mean 9.3 14.4 13.4
because many of our samples are from nonfasting pa-
a
tients. This interference has a large potential clinical Lipemic samples were identified by a lipemic index measured with the
Hitachi 747 analyzer.
impact on the measurement of creatinine in the critical
410 Technical Briefs

the lack of a readily available source of lipoproteins for
adding to samples. Effects of lipemia are often estimated
by using synthetic lipid emulsions, which may or may not
accurately replicate the physicochemical properties of
actual lipoproteins during analyses. With regard to inter-
ference from lipoproteins in a creatinine assay, we found
that the effects of lipemia disappeared when samples
were stored overnight, even when samples were vigor-
ously vortex-mixed to try to disperse the lipoproteins in
the samples. Use of stored samples, as commonly done in
many method-evaluation studies, would not detect the
interference found in freshly drawn samples. This exam-
ple illustrates the need to use actual lipemic samples
freshly drawn from human subjects when investigating
interference from lipemia.
References
1. Jaffe M. Über den Niederschlag welchen Pikrinsäure in normalen Harn
erzeugt und uber eine neue Reaction des Kreatinins. Z Physiol Chem
1886;10:391– 400.
2. Narayanan S, Appleton HD. Creatinine: a review. Clin Chem 1980;26:1119 –
26.
3. Spencer K. Analytical reviews in clinical biochemistry: the estimation of
creatinine. Ann Clin Biochem 1986;23:1–25.
4. Bowers LD, Wong ET. Kinetic serum creatinine assays. II. A critical evalua-
tion and review: the role of various factors in determining specificity. Clin
Chem 1980;26:555– 61.
5. Weber JA, van Zanten AP. Interferences in current methods for measurement
of creatinine. Clin Chem 1991;37:695–700.
6. Hoffman RJ, Feld RD. A modified procedure for creatinine (CRT) interference
which eliminates bilirubin (BILT) interference on the Hitachi 737 [Abstract].
Clin Chem 1988;34:1313.
7. Foster-Swanson A, Swartzentruber M, Nolen PA, Crawford K, Sine HE.
Multicenter evaluation of the Boehringer-Mannheim compensated, rate-
blanked/Jaffe application on BM/Hitachi systems. Adv Clin Diagnostics
(Boehringer-Mannheim Corp.) 1993;3–12.
8. Foster-Swanson A, Swartzentruber M, Roberts P, Feld R, Johnson M, Wong
S, et al. Reference interval studies of the rate-blanked creatinine/Jaffe
method on BM/Hitachi systems in six US laboratories [Abstract]. Clin Chem
1994;40:1057.
9. Painter PC, Cope JY, Smith JL. Reference intervals. In: Burtis CA, Ashwood
ER, eds. Tietz textbook of clinical chemistry, 2nd ed. Philadelphia: WB
Saunders, 1994:2184 –5.
10. Levey AS, Perrone RD, Madias NE. Serum creatinine and renal function.
Annu Rev Med 1988;39:465–90.
Editor’s Note: Kenneth A. Slickers and Harrison E. Sine
(Boehringer-Mannheim Corp., Lab Diagnostics, 9115
Hague Rd., P.O. Box 50446, Indianapolis, IN 46250-0446)
comment:
Hortin and Goolsby have demonstrated not only that the
Jaffe creatinine assay remains imperfect (despite all efforts
to confer specificity) but also that the interferograph
model for assessing interferences has yet another short-
coming. We have previously recognized that the age of
the hemolysate can influence hemoglobin interference
and that the type of bilirubin can influence the icterus
interference. Now, apparently, in addition to the concen-
tration and type of lipids in the sample, even the age of
the lipids that give rise to lipemia may also be important
in predicting interferences based on turbidity measure-
ments. Clearly, as valuable as the serum indexes are, they
are still only indicators of possible interferences, and not
quantitative estimates.
Studies are currently under development to determine
what, if any, improvement can be made in the current
Jaffe creatinine assay that will preserve the improved
performance vis-à-vis bilirubin that the rate-blanking in-
troduced, while minimizing this newly recognized
influence of endogenous lipemia. We would like to deter-
mine more specifically which components of the sample
lipids are responsible for the behavior. But we recognize
that it is also possible, based on the poor correlation of the
bias to the sample lipemia, that the actual interference is
due to some other sample constituent, and that the
lipemia is simply a marker for this interfering substance.
Upon completion of these studies, operating parameters
and (or) assay limitation statements will be appropriately
amended.

Correction
In the Beckman Conference paper by J.B. Silkworth and J.F. Brown, Jr., entitled “Evaluating the impact of exposure
to environmental contaminants on human health,” 1996;42:1345–9, insert the underlined words:
p. 1345, line 8 of the abstract: “ . . . are also produced, often to greater extents, by naturally occurring constituents . . . .”
p. 1345, line 7 of the article: “ . . . (DDT), might have on wildlife and even on human health. The press then
popularized the issue and may have been more effective at promoting fear than understanding.”
p. 1346, first column, line 9: “ . . . specific cytoplasmic receptor protein known as the aromatic hydrocarbon receptor.”
p. 1348, first column, line 11: “ . . . effects produced in humans by the synthetic chemicals of this type are no different
from those produced— generally to a much greater extent— by long-accepted natural components of the human
environment.”
On page 1345, second column, end of first paragraph, replace “ . . . lowest observable effect concentration.” with
“ . . . lowest observable effect level.”
The final sentence on page 1345 should read: “Some of these congeners are of environmental concern because they
tend to bioaccumulate and to produce toxicological effects in heavily dosed test animals [7,9 ], including immunotoxicity,
birth defects . . . .”

View publication stats