Академический Документы
Профессиональный Документы
Культура Документы
In the RA-1000, a random-access discrete analyzer, an inert a relative (to water) density of 1.9, it settles to the bottom of
fluorocarbon fluid is used to prevent interaction and car- the cuvette. The analytical reaction that takes place in the
ryover. Production-model instruments were evaluated in two cuvette can be monitored by colorimetric or ultraviolet
laboratories with respect to determination of glucose, creati- detection, with the use of direct end-point or initial or
nine, total protein, inorganic phosphorus, cholesterol, alka- kinetic rate reactions. The probes are raised and moved
line phosphatase, lactate dehydrogenase, creatine kinase, y- horizontally while the reagent and sample trays rotate to a
glutamyltransferase, and aspartate and alanine aminotrans- position appropriate for the next analysis. Back-and-forth
ferases. Within-run, among-run, and day-to-day (for 15 days) and stop/start rotation of the reaction tray mixes the sam-
precision was assessed, and results were correlated with ples with the reagents.
With the RA-1000, random access analysis is possible at
those obtained by the methods routinely in use in our
an analytical rate of 240 assays per hour. The operator’s role
departments. Precision was excellent, correlation accept-
is merely to place samples in the cups, ensure that reagent
able. containers are in place, and then request analysis, identify-
ing the specimen through the computer keyboard. Calibra-
In 1982, Technicon Instruments Corporation introduced a
tion, test-result analysis, detection of machine problems,
new generation of automated chemical analyzers (1). This
and calculation of results are all done by the computer.
instrument, the “RA-lOOO,” permits rapid “random-access”
Our evaluation of this instrument, reported here, includ-
clinical analysis without the limitation of cross-contamina- ed studies of the precision of the methods, linearity of the
tion (carryover) that exists at both sampling and reagent
procedures, and correlation of the values generated by the
probe in all previously described continuous-flow and many
RA-1000 with those obtained in our respective clinical
discrete clinical analyzers. In the RA-1000, an immiscible,
laboratories by our routinely used methods.
nonreactive fluid is used as a positive barrier between
sample and reagent and the interior and exterior surfaces of Materials and Methods
the respective probes. This fluorocarbon material provides
an inert, deformable surface that prevents carryover and The production-model HA-bOO and reagents we used in
ensures accurate delivery. In operation, the adjacent aque- these studies were provided by the manufacturer. The tests
ous liquids have contact only with the fluorocarbon and not we evaluated included alkaline phosphatase (EC 3.1.3.1) (2),
with the solid surface of the tubing or other fittings on the alanine and aspartate aminotransferases (EC 2.5.1.2 and
instrument. 2.6.1.1) (3), cholesterol (4), creatine kinase (EC 2.7.3.2) (5),
The instrument consists of sample probes and reagent creatinine (6), glucose (7), lactate dehydrogenase (EC
probes, which operate through Hamilton-type syringes; a 1.1.1.27) (8), total protein (9), inorganic phosphorus (10), and
circular 30-position sample tray; a reagent tray (maximum y.glutamyltransferase (EC 2.3.2.2) (11). We selected these
of 14 reagent containers); a 100-position reaction tray; and a tests for assessment because they represent the several
computer, which controls the operation. types of assays that can be accommodated in the RA-1000:
those depending on direct and end-point reactions (total
In operation, the sample- and reagent-probes are initially protein, inorganic phosphorus, glucose, cholesterol), kinetic
positioned over their respective trays with the syringes in rate reactions (y-glutamyltransferase, alkaline phospha-
the discharged position but with the reagent lines ifiled with tase, creatine kinase, lactate dehydrogenase, and the amino-
the fluorocarbon. A small air bubble is aspirated into each transferases), and initial rate reaction (creatinine).
probe, and the probes are then lowered into the sample cup The instruments used in the comparative studies (at the
and reagent container. Sample and reagent are aspirated respective laboratories) were the szsc (Technicon Instru-
into the tubing by the syringe pumps. The computer controls ments Corp.)’; the Cobas-Bio centrifugal analyzer (Roche
aspiration of a specified volume of sample (2.5-30 pL) and Analytical Instruments, Nutley, NJ 07110)1; the Astra 8
reagent (325-375 pL) through the syringes and into the (Beckman Instruments, Inc., Fullerton, CA 92834)3.2; the
tubing. The probes are then raised, moved horizontally, and aca discrete analyzer (Du Pont Instruments, Wilmington,
positioned over the reaction cell cuvette, on the reaction DE 19898)2; the Multistat analyzer (Instrumentation Labo-
tray. The probes are lowered into this cuvette, and sample ratory, Lexington, MA 02173)2; and the AutoAnalyzer LI
and reagent are discharged by the syringe pumps. A small (Technicon)2.
amount of the fluorocarbon is also introduced into the In our evaluation, we determined precision according to
cuvette by a peristaltic pump. Because the fluorocarbon has NCCLS guidelines (12). For use in the precision studies, six
reference materials with analyte concentrations extending
through the analytical range of the assays were provided by
‘Department of Clinical Chemistry, Memorial Sloan-Kettering Techmcon [Beta SMAC calibrator, RA-1000 calibrator, TQC
Cancer Center, New York, NY 10021. Hi, TQC Lo (linearity controls), and Alert 1 and Alert 2
2Department of Laboratory Medicine, Boston University Hospi-
tal Medical Center, Boston, MA 02118. (assayed controls)]. One group2 also used control specimens
3Technicon Instruments Corp., Tarrytown, NY 10591. with two analyte concentrations, provided by the Quality
Received July 18, 1983; accepted November 14, 1983. Control Survey Program of the Massachusetts Pathology
Table 1. Continued
Total precisIon: mean analyte value (and CV, %)
Beta Alert 1 Alert 2 TQC HI RA-1000 cal TOC Lo In-house 1 #{149} In-house 2
2240 (1.4) 720 (1.3) 3050 (1.3) 4130(1.7) 2390(1.1) 460 (1.6) 2690 (1.9) 990 (1.4)
49(1.1) 15(2.8) 89(1.0) 182 (2.5) 90(1.2) 9 (4.7) 51 (1.9) 10 (5.3)
49 (1.5) 15 (2.0) 89 (1.6) 182 (2.5) 91(1.3) 9 (4.9) 16 (1.8) 52 (1.8)
57(0.7) 68 (2.3) 55 (0.8) 86(0.8) 60(1.1) 26 (1.6) 81 (0.7) 61 (1.1)
57(0.5) 70(1.7) 55 (0.8) 85(0.7) 60(0.8) 26 (1.6) 71 (0.7) 51 (1.2)
56(1.1) 64(1.0) 58(1.2) 82(1.5) 55(0.9) 10 (4.3) 71 (1.9) 37 (1.4)
58 (0.5) 63(1.1) 59 (0.7) 68(1.0) 58(0.9) 13 (2.3) 41 (1.0) 78 (2.2)
1840 (2.6) 1900(2.7) 1200 (3.3) 2400(1.5) 1840 (1.4) 460 (2.5) 2560 (1.4) 1540 (2.6)
1890 (1.0) 1930 (2.4) 1230 (1.7) 2450(1.4) 1870 (1.1) 460 (2.3) 2490 (1.2) 1300 (3.7)
78 (1.4) 48 (8.2) 182 (2.7) 480 (1.3) 104 (1.9) 13 (5.3) 235 (1.6) 83 (6.4)
83(1.6) 57(7.1) 194 (1.6) 510 (1.2) 111 (1.6) 14 (5.7) 75 (3.5) 183 (2.1)
152 (1.9) 37(8.4) 185 (2.4) 86 (1.8) 131 (2.3) 11(12.6) 78 (2.6) 13 (10.1)
164 (0.9) 36 (6.6) 199 (1.6) 307 (1.3) 140 (1.3) 13 (6.2) 36 (3.3) 100 (2.2)
118 (2.1) 18 (8.8) 145 (2.2) 150 (1.6) 76(2.6) 8(13.9) 81 (2.8) 18 (8.3)
128(1.3) 18 (7.6) 154 (1.3) 160(1.5) 80(1.9) 11(10.0) 31 (10.8) 118 (1.8)
598 (1.1) 148 (6.0) 516 (5.5) 713 (1.8) 508 (2.5) 88 (3.3) 291 (7.7) 155 (6.6)
632 (1.7) 148 (5.3) 522 (5.1) 737 (2.1) 506 (2.6) 91 (4.1) 125 (4.8) 428 (8.5)
275 (2.5) 120 (2.5) 397 (1.5) 370 (1.6) 115 (2.1) 85 (2.2) 314 (1.7) 140 (1.9)
294 (1.0) 129 (2.2) 430 (1.2) 392 (1.3) 120 (1.4) 85 (4.1) 174 (1.6) 409 (1.7)
164 (1.9) 28 (8.4) 115 (3.1) 219 (2.2) 135 (2.2) 31 (3.4) 134 (2.2) 25 (3.0)
176 (1.2) 27 (8.4) 124 (1.9) 237 (1.5) 145 (1.5) 33 (2.1) 59 (1.7) 137 (2.0)
Table 3. Correlation of RA-1000 Procedures with Techniques Routinely Used by MSKCC* and BUMC**
Mean values
Correlation
CorrelatIon No. of RoutIne Regression .quatlon coeffIcIent,
ConstItuent (ref.) method samples method RA-1000 (y = PA-bOO) r S,.1
Glucose (20) Astra 156 1170 1230 y = 1.06x - 7.1 0.998 37
Totalprotein(9) 132 63 63 y = 1.02x - 1.2 0.996 1.5
Totalprotein(21) *Cobas 71 65 64 y = 1.OOx - 1.9 0.996 1.7
Total protein (21) **iL Multistat 49 62 61 y = O.97x + 5.7 0.996 1.2
Creatinine (22) SMAC 134 21 20 y = O.94x - 0.1 0.998 1.3
Creatinine (6) *Astra 8 70 21 22 y = 1.04x + 0.1 0.998 1.4
Creatinine (6) #{176}#{176}Astra
8 50 48 48 y = I .03x - 0.7 0.996 5.6
Phosphorus (10) SMAC 131 37 38 y = 1.03x + 0.2 0.985 2.6
Phosphorus (10) Cobas 40 38 39 y = 0.93x + 3.8 0.981 3.6
Phosphorus (23) #{176}AutoAnal
Ii 44 40 42 y = 1.06x - 2.7 0.999 1.0
Cholesterol (4) SMAC 132 1790 1830 y = 1.02x - 1.2 0.996 1.7
Cholesterol(4) *Cobas 71 1850 1850 y = 1.OOx - 1.9 0.996 1.5
Cholesterol(24) **aca 49 2490 2460 y = O.97x + 1.7 0.996 1.2
Alk. phosphatase (25) #{176}swc 37 175 173 y = 0.96x + 4.9 0.992 9.8
Alk. phosphatase (26) **lL Multistat 59 119 172 y = 1.49x - 5.2 0.992 9.5
Asp. aminotransferase (27) *SMAC 70 41 38 y = 0.91x - 0.89 0.999 5.9
Asp. aminotransferase (28) *Cobas 125 47 44 y = O.99x - 2.50 1.000 3.4
Asp. aminotransferase (28) “IL Multistat 86 37 54 y = 1.62x - 8.44 0.999 7.8
Ala. aminotransferase (27) *5AC 116 35 42 y = 1.17x + 1.55 0.998 3.4
Ala. aminotransferase (28) *Cobas 65 32 36 y = 1.19x - 2.22 0.999 5.9
Ala. aminotransferaso (28) “IL Multistat 65 58 105 y = 1.88x - 3.60 0.999 6.3
Creatine kinase (29) *SMAC 128 73 90 y = 1.07x + 12.0 0.996 11.9
Creatine kinase (30) *Cobas 68 65 91 y = 1.18x + 14.2 0.998 8.3
Creatine kinase (30) **lL Multistat 60 158 330 y = 1.99x + 1.17 0.996 14.1
Lactate dehyd. (31) *SMAC 120 241 193 y = O.78x + 6.40 0.995 9.8
Lactate dehyd. (32) *Cobas 65 263 215 y = 0.79x + 7.07 0.998 9.9
Lactate dehyd. (32) **lL Multistat 60 195 320 y = 1.58x + 8.9 0.997 15.8
-Glutamyltrans. (32) *SMAC 118 102 92 y = 0.87x - 4.01 0.998 0.14
y-Glutamyltrans. (34) *Cobas 65 90 90 y = 0.85x + 0.16 0.998 0.13
-Glutamyltrans. (34) 58 149 100 y = 0.64x + 5.47 0.998 10.2
A11Multistatassayswere performed at 30 #{176}C.
Unitsfor glucose, creatinine,phosphorus,and cholesterol,mglL;totalprotein,g/L; all enzymes,tilL.
results obtained for the same specimen with the RA-1000 temperature results in a wide difference in slope, but the
and by the routine procedures mentioned above. We empha- correlation coefficients (r) were excellent.
size that all the routine comparison enzyme measurements As part of the evaluation, all of the correlation assays
at one institution2 were done at 30#{176}C,
whereas all RA-1000 were done in duplicate, by both the RA-1000 and the
assays were carried out at 37 #{176}C.
This difference in assay comparative methods. Table 4 lists the standard deviations