Journal of Immunological Metbods, 148 (1992) 57-64

57

© 1992 ElsevierScience Publishers B.V. All rights reserved 0022-1759/92/$05.00

JIM 06226

Evaluation of a simple method for improving the precision of an ELISA detecting antibody in serum R i c h a r d J. W h i t t i n g t o n Microbiology Section, Elizabeth Macarthar Agricultural Institute, New South WalesAgriculture. Camden 2570, Australia

(Received 15 August 1991,revised received 17 October 1991. accepted 15 November 1991)

A method of correcting O D values to improve between-assay precision (PCF4 method) was evaluated using an ELISA detecting anti-Dichelobacter nodosus antibody in ovine serum. Four control sera with pre-defined target O D were included on each plate and used to calculate a correction factor. The mean between-plate CV for eight reference sera tested over a period of 6 months was reduced from 15.2% to 9.4% in a polyclonal conjugate assay (114 plates) and from 13.3% to 9.8% in a monoclonal conjugate assay (128 plates) using this method. The PCF4 method was compared with an existing method that uses only one control serum and was found to be superior. Time course plots of PCF4 values were useful for laboratory quality control. The mean within-plate coefficients of variation for four reference sera tested in paired wells were 8.2% and 5.1% in tile polyclonal and monoclonal conjugate assays respectively. Marked changes in O D due to experimental manipulation of antigen dilution, serum dilution and incubation time, conjugate dilution and chromogen-substrate incubation time were corrected by the method. Key words: ELISA quality control; Precision; Antibody capture ELISA; DichekJbacter nodosus

Introduction An aim of any laboratory assay is to provide meaningful results that are highly repeatable. Since its first application in solid phase assays in the early 1970s, ELISA has become widely used. However, relatively little has been published on methods for quality control. Multi-step ELISA protocols provide many opportunities for errors

Correspondence to: R.J. Whittington, Elizabeth Macarthur Agricultural Institute, New South Wales Agriculture, Private Bag 8, Camden 2570, Australia. Abbreviations: OD, optical density; CV, coefficientof variation; PBS, phosphate-buffered saline; PCF, plate correction factor; Ig, immunoglobulin.

and the complexities of antigen-antibody and enzyme-substrate reactions contribute to variability in the results. This variability may be random, where results for individual samples on a microtitre plate fluctuate about their true values, or systematic, where results for samples are either all higher or all lower than their true values. The use of ELISA in large scale cpidcmiological studies necessitates adequate assay precision. This permits meaningful determination of positivenegative cut-off values and the study of humoral responses in large populations over time. In these situations it is not possible to evaluate all samples of interest in a single assay and between-assay variation must be minimised. Sources of between-assay variation in ELISA have been discussed by MeLaren et al. (1981) and

Wright et al. (1985) and include the differential adsorptive characteristics of polystyrene microtitre plates, fluctuations in pH and temperature of buffers, errors in reagent preparation, reagent instability, variability in washing procedures, timing of reactions and volumetric pipetting errors particularly in the dilution of serum. Several methods for reducing between-assay variation in ELISA have been reported: stopping the enzyme-substrate reaction when the OD of a reference serum reaches a target value (McLaren et al., 1978) or a predicted endpoint (Wright et al., 1985), expressing results as the ratio between the OD of test sera and the OD of reference sera (Tijssen, 1985), expressing results as units read from a standard curve constructed from control samples tested on each plate (Malvano et al., 1982) and transforming OD data based on the OD achieved by a control sample with a pre-defined target OD (Voller et al., 1979). Of these methods, only that of Malvano et al. (1982) appears to have been reported in detail (Calamel and Lambert, 1988), although all may be in general use. The aim of this study was to evaluate a simple method for reducing between-assay variation. Voller et al. (1979) recommended the use of a correction factor for each plate (PCF) calculated as the ratio of a single control serum's target OD to its actual OD (henceforth referred to as the PCF1 method). The actual OD of test sera are then multiplied by the PCF. The PCF1 method was adapted by using four contro| sera in the calculation of the PCF (henceforth referred to as the PCF4 method). The PCF4 method was evaluated in a diagnostic laboratory using an indirect ELISA detecting anti-Dichelobacter nodosus antibody in sheep serum. In addition, the ability of PCF to correct OD values for systematic errors caused by major changes in reagent concentrations and incubation times was examined.

Materials and methods

ELISA protocol The microtitre plate assay to detect antibodies in ovine serum against D. nodosus (Whittington et al., 1990) consisted of the following steps: (1)

antigen (1/1600 in borate buffer), 100 p.I, overnight, 4°C; (2) wash 5 x (0.05% v/v Tween 20 in distilled water); (3) serum (10 ttl plus 1990 p.! 0.1% W / V ovalbumir, in PBS), 100/tl, 90 min, room temperature (RT). (4) wash as above; (5) antibody-enzyme conjugate (in 0.1% ovalbumin v/v in PBS), 100 pl, 90 min, RT; (6) wash as above; (7) chromogen-substrate A B T S / H 2 0 2, I00 /zl, 20 rain, RT; (8) stop reaction (sodium azide), 50 tLl; (9) read at 405 nm. Two lg-horseradish peroxidase conjugates were used in separate assays: a monoclonal anti-ruminant Ig at 1/1000 (Bi2, Australian Monoclonal Developments) (AMD-ELISA) and a polyeional rabbit anti-sheep Ig at 1/2000 (Kirkegard Perry Laboratories) (KPL-ELISA). A Model 120 plate washer (Flow Laboratories) and a Multiskan mierotitre plate reader (Flow Laboratories) were used for all tests. Sera and diluents (up to 5 ml) were dispensed using continuously variable pipettors (Gilson), large volumes of diluents were quantitated in measuring cylinders, while reagents common to all wells on a plate were dispensed using 8-place multichannel pipettors (Flow Laboratories). Buffers and diluted reagents were prepared freshly each day.

PCF4 method Four control sera (a negativ:.~ - 4907, and a low - 1333, medium - 1385 and high - 1921 positive) were chosen to span the range of optical densities generally recorded in the D. nodosus ELISA, that is approximately 0-2.5 GD units. These sera were diluted 1/10 in 50% glycerol in 25 mM Tris-HC! pH 7.4, 0.15 M NaCI buffer and stored at -20°C. Each control serum was tested in duplicate at a final dilution of 1/200 on each plate, using wells Ell-H12. For the KPL-ELISA, a target OD was assigned to each control serum based on results obtained in several titration experiments. For the AMD-ELISA, target OD were defined as the mean OD obtained under standard conditions during the previous 6 months. A correction factor was calculated for each of the control sera for each plate as target OD/actual OD. A PCF4 value was calculated for each plate as the mean of these four values. All OD values recorded from the plate were then multiplied by the PCF4 value. These calculations were

programmed into a computer interfaced to the plate reader, using a commercial software product ( A U T O M A T E , Flow Laboratories). A report detailing O D for individual wells, mean O D for paired wells, CV for paired wells, target O D for control sera, correction factors for individual control sera, PCF4 and corrected O D for paired wells was printed. Quality control charts were maintained by plotting PCF4 values for each plate over time.

Evaluation of the PCF4 method in routine ELISA testing Four quality control sera (A, B, C and D) were included in wells A1 I-D12 on each plate in order to monitor the effectiveness of the PCF4 method in reducing variability of results in test sera. These controls were prepared by diluting serum from a sheep that had been immunised with a commercial D. nodosus vaccine with control negative ovine serum, to achieve a range of serum antibody titres. Aliquots of 10/~l of each prepared serum were stored at - 2 0 ° C and diluted for use each day in parallel with test sera. Uncorrected and corrected O D values for each of the quality control and PCF4 control sera were recorded for a period of 6 months during routine use of the D. nodosus ELISA tests in a diagnostic laboratory.. The mean, standard deviation and CV were calculated for each control serum.

Comparison of PCF4 and PCFI methods Corrected O D values were also calculated using the PCF1 method (see Voller et al., 1979). Serum 1385 was used as the control serum for calculation of the PCFI value for each plate.

Evaluation of the ability of PCF to correct for systematic variation in OD due to experimental manipulation o f the ELISA protocol Thirteen experiments were undertaken in which reagent concentrations or incubation times were deliberately altered from those in the standard protocol. In each experiment, one of the steps in the protocol was varied, while all other steps were unchanged. Antigen was tested at dilutions of 1/100, 400, 800, 1600 and 3200. Serum was tested at dilutions of 1/100, 200, 300 and 400 and at incubation times of 30, 60, 90, 120, 150

and 180 min. KPL conjugate was tested at dilutions of 1/1500, 2000, 2500 and 300{). Substrate incubation time was tested at 20, 25, 30, 40, 50, 60 and 70 rain. Finally, the concentration of H 2 0 2 in the chromogen-substrate solution was tested at 1/10, × 1 and × 10. In each experiment between 2 and 12 ~ r a wzre tested and the mean O D of duplicate wells was determined. Mean O D from wells that corresponded to standard reagent concentrations or incubation times were defined as the target OD, while mean O D from wells in which the protocol had been varied were defined as the actual OD. A correction factor (target O D / a c t u a l O D ) was calculated for each serum sample for each reagent concentration or incubation time. Two or four sera were selected as c~atrols and the mean correction factor value from these controls was multiplied by actual O D to yield a corrected O D for each serum at each reagent concentration or incubation time. Both an uncorrected ratio (actual O D / t a r g e t OD) and corrected ratio (corrected O D / t a r g e t OD) were calculated for each serum for each reagent concentration or incubation time. The hypothesis that the population mean for each of these ratios equalled 1 was tested using Student's t test.

Evaluation of within-plate variation Within-plate variation in O D was estimated by testing a single serum across all wells of a plate, using standard test conditions. PCF4 control sera 4907, 1333, 1385 and 1921 were tested with the KPL-ELISA while quality control sera A, B, C and D were tested with the AMD-ELISA. Mean ODs for adjacent paired wells were calculated. The mean, standard deviation and CV were then calculated for individual and paired well data. Data was examined visually for non-random variation such as edge effects.

Results

Evaluation of the PCF4 method in routine ELISA testing The PCF4 method was evaluated following its use over a period of 6 mouths. Over this period

six b a t c h e s o f a n t i g e n , t h r e e b a t c h e s o f K P L - c o n jugate and three batches of chromogen-substrate w e r e u s e d . All a s s a y s w e r e p e r f o r m e d by t h e s a m e p e r s o n . A total o f 114 plates were t e s t e d with t h e K P L - E L I S A while 128 were t e s t e d with the AMD-ELISA. The PCF4 method improved between-plate precision in O D r e s u l t s in b o t h A M D - a n d K P L E L I S A tests ( T a b l e !). T h e m e a n ( r a n g e ) bet w e e n - p l a t e C V for t h e eight control sera in t h e K P L - E L I S A w a s 15.2% ( 1 1 . 7 - 1 8 . 6 % ) b e f o r e correction, c o m p a r e d with 9.4% ( 6 . 7 - 1 3 . 9 % ) a f t e r correction. C o r r e s p o n d i n g f i g u r e s for t h e A M D E L I S A were 13.3% ( 6 . 6 - 1 7 . 7 % ) before correction a n d 9.8% ( 6 . 9 - 1 2 . 8 % ) after correction. O v e r t h e 6 m o n t h period, t h e m e a n ( r a n g e ) P C F 4 v a l u e for t h e K P L - E L I S A w a s 1.12 ( 0 . 8 4 -

1.521 while t h a t for t h e A M D - E L I S A was 1.02 (0.76-1.321. P C F 4 v a l u e s a p p e a r e d to be approxim a t e l y n o r m a l l y d i s t r i b u t e d for t h e A M D - E L I S A , b u t b i m o d a i l y d i s t r i b u t e d for t h e K P L - E L I S A (Fig. 11. N o long t e r m t r e n d s in P C F 4 v a l u e s w e r e o b v i o u s by e x a m i n a t i o n o f quality control c h a r t s (Fig. 2), a l t h o u g h t h e b i m o d a l f o r m o f t h e K P L E L I S A P C F 4 f r e q u e n c y distribution w a s s e e n to be d u e to a n a v e r a g e r e d u c t i o n in P C F 4 v a l u e s in t h e latter p a r t o f t h e study, probably d u e to t h e u s e o f a d i f f e r e n t stock o f c o n j u g a t e . In general, t h e C V o f seropositive s e r a (1333, 1385, 1921, B, C a n d D) w e r e greatly r e d u c e d by u s e o f t h e P C F 4 m e t h o d , while t h e C V o f r e s u l t s for s e r o n e g a t i v e s e r a 4907 a n d A w e r e r e d u c e d by a s m a l l e r d e g r e e . Seropositive s e r u m D in t h e A M D E L I S A w a s a n e x c e p t i o n in w h i c h C V

TABLE I BETWEEN-PLATE VARIABILITY IN THE OD OF CONTROL SERA TESTED IN THE KPL-ELISA The PCF4 method was compared with the PCFI method of Voller et al. (1979). The CV of the OD of each serum before (uncorrected OD-UOD) and after (corrected OD-COD) correction were compared. The CV af control sera were substantially reduced by the PCF4 method, n = 114 plates. Serum

Value

Mean

Stdev

Min

Max

CV

4907

UOD COD-PCFI COD-PCF4

0.16 0.18 O.18

0.030 0.035 0.025

0.11 0.12 0.13

0.32 0.38 0.32

18.64 19.37 14.05

i 333

UOD COD-PCF 1 COD-PCF4

0.68 0.76 0.75

0.109 0.065 0.053

0.45 0.60 0.59

0.94 0.92 0.91

16.10 8.62 7.12

1385

UOD COD-PCFI COD-PCF4

0.90 1.00 0.99

0.152 0.000 0.066

0.63 1.00 0.83

1.35 1.00 1.22

16.95 0.00 6.72

1921

UOD COD-PCFI COD- PCF4

2.05 2.31 2.27

0.251 0.245 0.163

1.26 1.68 1.81

2.55 2.87 2.60

12.24 10.62 7.19

A

UOD COD-PCFI COD-PCF4

0.29 0.33 0.33

0.049 0.053 0.045

0.16 0.19 0.20

0.42 0.46 0.44

16.75 15.99 13.70

B

UOD COD-PCFI COD-PCF4

1.05 1.18 1.16

0.160 0.143 0.118

0.69 0.82 0.85

1.51 1.60 1.45

15.29 12.1 I 10.16

C

UOD COD-PCFI COD-PCF4

1.72 1.93 1.90

0.233 0.19~ 0.148

1.18 1.48 1.38

2.37 2.43 2.25

13.57 10.I 1 7.81

D

UOD COD-PCFI COD-PCF4

1.92 2.16 2.12

0.225 0.244 0.171

1.34 1.51 1.70

2.43 2.84 2.74

I 1.72 I 1.26 8.04

61 increased from 6.6% b e f o r e correction to 9.2% after correction.

ing significantly from ! ( p < 0.05) in each case w h e r e the s t a n d a r d protocol w a s altered. Corrected ratios w e r e not, however, significantly different from i, indicating that actual O D w e r e converted to values a p p r o a c h i n g target O D w h e n a correction factor calculated from two o r four sera w a s used.

Comparison with tire PCFI correction method C o r r e c t e d O D calculated using the P C F I m e t h o d w e r e m o r e variable t h a n those calculated using the PCF4 m e t h o d (Table I). T h e C V actually increased for 6 o u t of 8 sera w h e n the P C F I m e t h o d was used in the A M D - E L I S A .

Within-plate variation N o obvious n o n - r a n d o m variation such as an edge effect was n o t e d o n examining the O D results, in the K P L - E L I S A , the m e a n (range) C V for paired well data was 8.2% 16.0-11.1%) while the m e a n (range) C V for individual well data was 8.8% (6.6-11.6%). T h e c o r r e s p o n d i n g results for the A M D - E L I S A , w e r e 5.1% ( 3 . 7 - 5 . 9 % ) for paired well data and 6.1% ( 4 . 7 - 6 . 6 % ) for individ-

Correction of variation due to experimental manipulation of ELISA protocol Alteration o f r e a g e n t c o n c e n t r a t i o n s o r incubation times caused substantial increases or decreases in the O D o f sera relative to the O D o b t a i n e d using the s t a n d a r d protocol. This was evident statistically by u n c o r r e c t e d ratios differTABLE il

BETWEEN-PLATE VARIABILITY 1N THE OD OF CONTROL SERA TESTED IN THE AMD-ELISA The PCF4 method was compared with the PCFI method of Voller et al. 119791. The CV of the OD of each ~rum before (uncorrected OD-UOD) and after (corrected OD-COD) correction were compared. The CV of control ~ra were substantially reduced by the PCF4 method, n = i28 plates. Serum

Value

Mean

Stdev

Min

Max

CV

4907

UOD COD-PCFI COD-PCF4

0.10 O.I0 0.10

11.016 0.018 0.011

0.07 0.06 0.118

0.18 0.16 0.15

16.36 18.31 11.16

1333

UOD COD-PCFI COD-PCF4

0.57 0.58 0.57

0.084 0.064 0.040

0.36 0.37 11.47

0.83 I).86 0.69

14.67 ! 1.01 6.92

1385

UOD COD-PCFI COD-PCF4

0.84 0.84 11.84

0.148 0.000 0.078

0.48 0.84 11.62

1.31 0.84 I. 14

17.72 0.00 9.28

1921

UOD COD-PCFI COD-PCF4

2.33 2.40 2.36

0.217 0.356 0.190

1.87 1.64 !.92

2.73 3.71 2.96

9.33 14.85 8.114

A

UOD COD-PCFI COD-PCF4

0.18 0.18 0.18

11.1127 0.033 0.023

0.13 O.10 0.13

0.25 0.32 0.25

14.97 17.74 12.76

B

UOD COD-PCFI

0.68 0.70 11.69

0.091 0.120 0.081

I).43 11.33

COD-PCF4

0.44

0.93 !.05 0.89

13.41 17.18 11.73

UOD COD-PCFI

1.47 1.51

0.195 0.211

1.05 0.95

2.02 2.33

13.2 I 14.111

COD-PCF4

1.49

0.135

I. I I

1.84

9.119

UOD COD-PCFI COD-PCF4

2.43 2.51 2.47

0.161 0.417 0.227

1.92 1.611 2.112

2.75 4.34 3.23

6.60 16.59 9.18

C

D

62 ual well data. Two operators conducted these assays.

a. AMD ELISA

Discussion This study indicated that the PCF4 method of correcting O D improved the precision of the ELISA tests and was practical in diagnostic applications. Experimentally, the method was robust, and corrected for major changes in O D caused by deliberate alterations to antigen dilution, serum dilution and incubation time, conjugate dilution and substrate reaction time and activity. The results can be extrapolated; for example, the effect of declining conjugate activity or shorter conjugate incubation time would be similar to incrcas-

121 ..................I........ el, 0



~

40

60

!12

11 1.8

.....................................................

1.4.

................................................

o010

2o

m

120

December

b. KPL ELISA

iii

.....................

a. AMD ELISA

80

Plate No.

July

....... " .........

0

July

20

.... 40

........ 60

ii

......... ......... 80

Plate No.

100

" 120

'

1

December

Fig. 2. Qualitycontrol charts for D. nodosus ELISA tests. The PCF4 value for each microtitre plate is plotted against time. a: AMD-ELISAb: KPL-EUSA.

u

lO

O.7

0.8

0.9

1.0

1.1

1.2

1.3

1.4

1.5

1.4

1.5

PCF4

b. KPL ELISA

0.7

0.8

0.9

1.0

1.1

1.2

1.3

PCF4

Fig. 1. Frequency distribution of PCF4 values for microtitre plates tested over a six month period, a: AMD-ELISA(128 plates); b: KPL-ELISA(I 14 plates).

ing conjugate dilution. Between-plate timing inconsistencies inherent in assays when large numbers of samples are tested should also be corrected by this method. The mean within-plate CV for individual sera was less than 10% in these assays. The cause of this random variation is difficult to determine, is inherent in all ELISA runs and cannot be eliminated. If a PCF is calculated from a single control serum, random variability is translated to the PCF and then applied to samples across the entire plate. If four control sera are used to calculate the PCF the random variability of individual sera will cancel out on average, resulting in a PCF4 value close to 1 and therefore minimal change to the actual O D of test sera. However, the use of 4 control sera still facilitates correction of a systematic positive or negative trend due to a procedural or reagent error. Thus use of the

PCF1 method in this study increased the CV of the results for six of eight control sera in the AMD-ELISA and was less effective than the PCF4 method in reducing the CV for the remaining sera in both ELISA tests. In the PCF4 method, an error in determination of the actual OD of one of the control sera will result in an erroneous PCF. Although control sera were not censored during this study~ it is valid to delete a control serum from the calculation of PCF if it is certain that its OD value is incorrect. Stone et al. (1991) discussed the validity of discarding errant replicate data in laboratory assays; in the present context the correction factors of individual control sera can be considered replicates. The computer software used to generate the ELISA report can be readily programmed to recalculate corrected OD using three of four control sera, for example. The control sera chosen for calculation of PCF in this study spanned the range of OD commonly recorded with the assay. The dose response curve for this ELISA system (determined by serum titration, data not shown) is non-linear, conforming to a sigmoid relationship typical of ELISA. The standard deviation of response (OD) increases with dose (antibody level) (Tables 1 and II) (Hamilton and Adkinson, 1988). As the PCF4 method uses a muitiplicative correction it compensates for the greater variability in response observed at higher doses. Therefore the PCF4 method is not limited by the non-linearity of the dose response relationship. Ideally, the target OD for each control serum should be the mean OD from a large number of assays. Examination of the frequency distribution of PCF4 for the KPL-ELISA indicated that the target OD were about 10% too low, resulting in a mean PCF4 of 1.12. This merely reflected the method by which target OD were calculated for the KPL-ELISA. High mean PCF4 values did not reduce the utility of the method, but they did cause unnecessary elevation of corrected OD values. Amongst the voluminous literature on ELISA there is relatively little information on assay precision and therefore no ready means of assessing the adequacy of newly developed assays. Tijssen (1985) suggested that the between-test CV of

results obtained for reference samples under ideal conditions should be less than 10%. Kurstak (1985) accepted a CV of up to 5% for within-assay precision and 10% for between-assay precision. However, a between-assay CV of 10-15% may be more usual (McLaren et al., 1981). Schuurs and Van Weemen (1977), in reviewing precision data for macro and micro immunoas`say techniques, found that the data of different authors may not be comparable due to different methods of estimating precision. However, the within-assay CV ranged from 1.3 to 33% while between-assay CV ranged from 0 to 20%. Precision data for ELISA conducted in mierotitre plates has been reported by relatively few workers. Stemshorn et al. (1983) and Wright et al. (1985) found a within-plate CV of 4.6-20.7% and a between-plate CV of 28.6% respectively in ELISA tests for anti-Brucella abortus antibody, in the latter study, the between-plate CV was reduced to 6.8% through measuring the OD of the reference serum at 4 min to determine the final substrate reaction time. Other reports of as.say precision include: a within-assay CV of 1.6-3.5% and a between-assay CV of 9.7-12.2% (lg irt porcine serum and whey) (Kelly et al., 1988); a within-assay CV of 3.2-6.3% and between-assay CV of 6.8-8.6% (anti-Pasteurella anatipestifer lg in avian serum) (Hatfield et al., 1988); a between-assay CV of 3.0-15.6% (anti-Trichinella spiralis lg in porcine serum) (Singh et al., 1988) and a between assay CV of 14% (molinate in water) (Harrison et al., 1989). In relation to these few references and the problems of compari,g data from different studies it appears that use of the PCF4 method in the D. nodosus ELISA tests resulted in a between-assay precision equivalent or better than that obtained in similar assays. Methods of quality control for ELISA vary from laboratory to laboratory, but often rely on a quality control chart on which the results for control samples over time are plotted (De'Ath, 1988). In this study a time course plot of PCF4 was found to be a simple means of monitoring trends. A PCF4 value provides summary data for four control sera and a steady rise or fall in PCF4 indicates a drift in assay conditions. With experience of a particular ELISA system, an acceptable range for PCF4 can be specified, with values

o u t s i d e this r a n g e indicating t h e n e e d to investig a t e t h e s y s t e m o r r e p e a t a n individual assay. T h e P C F 4 m e t h o d r e q u i r e s t h e u s e o f only 8 wells for t h e control s e r a o n a 96 well microtitre plate ( a s s u m i n g p a i r e d well samples). T h e d a t a generated are corrected OD, which may then be t r a n s f o r m e d to suit t h e specific n e e d s o f reporting t h e test results. In conclusion, t h e P C F 4 m e t h o d is a simple a n d effective m e a n s o f improving t h e precision o f soiid p h a s e E L I S A .

Acknowledgements M i s s J. F a r r u g i a p r o v i d e d skilled technical assistance d u r i n g this study, w h i c h w a s s u p p o r t e d in p a r t by a g r a n t f r o m t h e A u s t r a l i a n W o o l R e search and Development Trust Fund. Graeme Eamens and John Searson provided valuable comments on a draft of the manuscript. Paul Nicholls is t h a n k e d for his i n t e r e s t in this work. T h e s t u d y w a s u n d e r t a k e n while t h e a u t h o r w a s e n r o l l e d as a p o s t - g r a d u a t e s t u d e n t in t h e D e p a r t m e n t o f A n i m a l H e a l t h , University o f Sydney.

References Calamel, M. and Lambert, M. (1988) E.LI.S.A. Standardised Technique, Laboratoire National de Pathologie des Petits Ruminants et des Abeilles, Nice. De'Ath, G. (1988) ELISA as a measurement process: some statistical considerations. In: G.W. Burgess (Ed.), Elisa Technology in Diagnosis and Research. James Cook University of North Queensland, Townsville, p. 120. Hamilton, R.G. and Adkinson, N.F. (1988) Quantitative asI~CtS of solid phase immunoassays. In: D.M. Kemeny and SJ. Challacombe (Eds.), EL1SA and Other Solid Phase lmmunoassays.John Wiley, Chichester, p. 57. Harrison, R.O., Braun, A.L., Gee, S..L, O'Brien, D.J. and Hammock, B.D. (1989) Evaluation of an enzyme-linked immunosorbent assay (EL1SA) for the direct analysis of molinate (Ordram ~) in rice field water. Food Agric. Immunol. 1, 37. Hatfield, R.M., Morris, B,A. and Henry, R.R. (1988) Development of an ELISA for the detection of antibodies to Pasteurella anatipestifer. In: B.A. Morris, M.N. Clifford and R. Jackman (Eds.), lmmunoassays for Veterinary and

Food Analysis-l. Elsevier Applied Science, London, p. 309. Kelly, K.W., Kleiss, A.J. and Brief, S. (1988) Quality control criteria for quantitative enzyme-linked immunosorbent assay of porcine immunoglobulins A and M. Am. J. Vet. Res. 49, 944. Kurstak, E. (1985) Progress in enzyme immunoassays: production of reagents, experimental design, and interpretation. Bull. W.H.O. 63, 793. Malvano, R., Boniolo, A., Doris, M. and Zannino, M. (1982) EL1SA for antibody measurement: aspects related to data expression. J. Immunol. Methods 48, 51. McLaren, M.L., Draper, C.C., Roberts, J.M., Minter-Goedbloer, E., Teesdale, C.H., Amin, M.A., Omer, A.B.S., Bartlett, A. and Voller, A. (1978) Studies on the ELISA test for Schistosoma mm~soni infection. Ann. Trop. Med. Parasitol. 72, 243. McLaren, M.L., Lillywhitc, J.E. and Andrew, C.S.A. (1981) Indirect enzyme linked immunosorbent assay (ELISA): practical aspects of standardisation and quality control. Med. Lab. Sci. 38, 245. Schuurs, A.H.W.M. and Van Weemen, B.K. (1977) Enzyme immunoassay. Clin. Chim. Acta 81, 1. Singh, P., Oliver, D.G., Abad, M., Vail, N., Jang, L., Brock, T. and Allison, D. (1988) Development of a practical ELISA for the detection of trichina in pork. In: B.A. Morris, M.N. Clifford and R. Jackman (Eds.), Immunoassaysfor Veterinary and Food Analysis-l. Elsevier Applied Science, London, p. 315. Stemshorn, B.W., Buckley, DJ., St. Amour, G., Lin, C.S. and Duncan, J.R. (1983) A computer-interfaced photometer and systematic spacing of duplicates to control within-plate enzyme-immunoassayvariation. J. Immunol. Methods 61, 367. Stone, A., Schwartz, J.E., Valdimarsdottir, A., Napoli, A., Neale, J.M. and 9, Cox, D.S. . . (1991) An alternative statistical treatment for summansmg the central tendency of replicate assay data. J. Immunol. Methods 136, 111. Tijssen, P. (1985) Processing of data and reporting of results of enzyme immunoassays. In: Practice and Theory of Immunoassays. Elsevier, Amsterdam, p. 385. Voller, A., Bidwell, D.E. and Bartlett, A. (1979) The Enzyme Linked Immunosorhent Assay (EL1SA). A Guide with Abstracts of Microplat~ Applications. FIowline Press, Guernsey. Whittington, R.J., Marshall, D.J., Walker, R.I. and Turner, M.J. (1990) Serum antibody responses in sheep after natu, ral infection with Bacteroides nodosus. Aust. Vet. J. 67, 98. Wright, P.F., Kelly, W.A. and Gall, D.EJ. (1985) Application of a timing protocol to the reduction of inter-plate variability in the indirect enzyme immunoassay for detection of anti-Brucella antibody. J. lmmunoassay 6, 189.

Evaluation of a simple method for improving the precision of an ELISA detecting antibody in serum.

A method of correcting OD values to improve between-assay precision (PCF4 method) was evaluated using an ELISA detecting anti-Dichelobacter nodosus an...
508KB Sizes 0 Downloads 0 Views