Intra-laboratory variation

G Sarwar, DA Christensen, AJ Finlayson, M Friedman, LR Hackler, SL Mackenzie, PL Pellett, R Tkachuk. Inter- and intra-laboratory variation in amino acid analysis of food proteins. J Food Science 48 526-531, 1983. 

Heavy oils and residua can be separated into a variety of fractions using a myriad of different techniques that have been used since the beginning of petroleum science (Speight, 1999). However, the evolution of these techniques has been accompanied by subtle inter-laboratory (and even intra-laboratory) variations to an extent that many of the nuclear fractionation procedures appear to bear very little relationship to one another. 

Buikema Jr., A.L. (1983) Inter- and Intra-laboratory Variation in Conducting Static Acute Toxicity Tests with Daphnia magna Exposed to Effluents and Reference Toxicants, API Publications 4362. American Petroleum Institute, Washington, DC. 

Consequently, it was proposed to define (Burns et al. [2005]) Robustness of an analytical procedure is the property that indicates insensitivity against changes of known operational parameters on the results of the method and hence its suitability for its defined purpose and Ruggedness of an analytical procedure is the property that indicates insensitivity against changes of known operational variables and in addition any variations (not discovered in intra-laboratory experiments) which may be revealed by inter-laboratory studies (Burns et al. [2005]). 

Property of an analytical procedure that indicates insensitivity against changes of known operational variables and in addition any variations (not discovered in intra-laboratory experiments) which may be revealed by inter-laboratory studies. 

The glucose values are worthy of attention because of the large number of determinations that have been made in many laboratories. There is evidence of substantial intra-individual variation in glucose values which is in part responsible for the range of the observed values.33,34 In spite of this normal variance in values in the same individual, however, there is evidence that inter-individual differences exist among well people. Some, on repeated tests, tend to have low values, some intermediate, and some high. 

In experiments carried on in the author s laboratory several years ago, the B vitamin content of milks from individual cows and from individual human mothers was determined.63 So far as this study is concerned, involving as it did only B vitamins, the intra-individual variation (that is, from day to day) appeared to be greater than interindividual differences. The number of human cases studied was not sufficient to be the basis of any sound conclusion on this point. 

In analytical chemistry, validation of the analytical methods is of utmost importance [4,5]. One of the aspects of this validation is the robustness of analytical methods against variations in experimental circumstances. The term experimental circumstances is very broad it might even include inter-laboratory variation. In this book, only intra-laboratory experimental conditions are considered. No explicit attention is given to inter-laboratory variations, although some of the presented methodology might be useful in that area. 

In a protocol about collaborative studies [10] it is also considered what is called preliminary estimates of precision. Among these the protocol defines the total within-laboratory standard deviation . This includes both the within-run or intra-assay variation (= repeatability) and the between-run or inter-assay variation. The latter means that one has measured on different days and preferably has used different calibration curves. It can be considered as a within-laboratory reproducibility. These estimates can be determined prior to an interlaboratory method performance study. The total within-laboratory standard deviation may be estimated fi-om ruggedness trials [10]. 

Intra-laboratory CVs range from 9.9% for ethylmalonate (at 102 pmol/1) to 40.7% for suberylglycine (at 48.6 pmol/1) and inter-laboratory CVs from 42.5% for ethylmalonate (at 102 pmol/1) to 757.4% for tiglyglycine at 83.5 pmol/1. This wide variation is also accompanied by marked variability in the reference ranges used by different laboratories an example is shown (Fig. 1.2.1) for a single return from 18 respondents who quantitated ethylmalonate in a single sample (sample 109) and reported both the result and the upper limit of normal used by their laboratory. Clearly the clinical significance of this apparently extreme variability depends upon the clinical context... 

The mean recovery of 100% (range 63% for orotidine to 124% for 2,8 dihydroxy-adenine) is probably acceptable similarly, the mean intra-laboratory imprecision ( V =11.9%, range 6.0 for pseudouridine to 21.9% for succinyl adenosine) is likely to be adequate for most clinical applications. However, the interlaboratory imprecision is somewhat disturbing mean CV = 126% (range 16.8% for pseudouridine to 295% for orotidine). This variation indicates the need to harmonise standardisation. 

Current inter- and intra-laboratory evaluations indicate the Equilibrium Jar s precision is 8% and between laboratory variation is about 10%. 

Precision defines the scatter of repeated analysis, or the coefficient of variation of analytical results. Both intra-assay and inter-assay precision must be investigated. Intermediate precision describes the influence of different analysts, equipment, days and other intra-laboratory variabihty. Inter-laboratory comparison is also of interest in establishing the precision of the method. AU testing on accuracy and precision must be carried out by replicate analyses of a statistically relevant number of samples. Depending on the use of the method, it may be necessary to estabhsh both parameters over the measurable range, or in the case of content determination simply in the range of 80-120% of the nominal value. 

The accuracy and reproducibility of the analysis performed in such a way is high, resulting in inter-laboratory coefficients of variation (CV) of 3% to 5%, and intra-laboratory CV of less than 1%, both with adequate sensitivity (LOD around 1 mg/dl) and high sjjedficity 0ones Schuberth, 1989 Jones, et al., 1992 Fenton, 1985). 

Rao et al.20 demonstrated a fluorescence polarization immunoassay for evaluating serum concentrations of tricyclic antidepressants (amitriptyline, imipramine, clomipramine, and doxepin) with respect to nonresponse, compliance, therapeutic window, and influences of age, sex, substance abuse, and toxicity. Abbott Laboratories TDx/TDxFLx Toxicology Tricyclic Assay FPIA (fluorescence polarization immunoassay) was used. This assay of 50 /uL samples contained tricyclic antidepressant antibodies raised in rabbits and fluorescein-labeled tricyclic antidepressant as a tracer. The assay was calibrated with imipramine in the range of 75 to 1000 fig/L (268 to 3571 nmol/L). Intra-assay and inter-assay coefficients of variation for internal quality control samples from the manufacturer were 4.2 and 4.7%, respectively. The limits of detection were 72,71,64, and 72 nmol/L for amitriptyline, imipramine, clomipramine, and doxepin, respectively. This high-throughput immunoassay was easy to use although amitriptyline, dosulepine, desipramine, and nortriptyline showed cross-reactivities ranging from 74 to 100%. 

The interlaboratory results obtained from the analysis of defined standard solutions, but also from the analysis of sediment extracts prepared either by the coordinator of the study or by the participants themselves, also provide a measure of the variation between laboratories. The results show that the interlaboratory reproducibility ranges from 6.5% for the defined dioxin sample to 27.9% for the sediment sample extracted by the participants themselves. As was mentioned before, the reproducibility for this last sample is relatively high and most presumably due to the introduction of extra handlings (extraction and cleanup) to the total procedure. In addition, the fact that not all the participants had prior experience with the extraction protocol to be used could have added to the increase in variability of the process. Furthermore, the dilution factor was not dictated. This also introduces a certain degree of variation. For the reproducibility of the DR CALUX bioassay itself and not caused by differences in operating extraction conditions, the maximum variation between laboratories was observed to be 18.0%. The results for the sediment extract samples can also be used to estimate the method variability for extracts, that is, based on samples of unknown composition. Again, given the intra-as well as the interlaboratory variations observed in this study, it appears justified to conclude that the standard deviation of the means provides a reasonable estimate of the method variability, based on the overall aver-... 

Several overall conclusions can be drawn based on the statistical evaluation of the data submitted by the participants of the DR CALUX intra-and interlaboratory validation study. First, differences in expertise between the laboratories are apparent based on the results for the calibration curves (both for the curves as provided by the coordinator and for the curves that were prepared by the participants) and on the differences in individual measurement variability. Second, the average results, over all participants, are very close to the true concentration, expressed in DR CALUX 2,3,7,8-TCDD TEQs for the analytical samples. Furthermore, the interlaboratory variation for the different sample types can be regarded as estimates for the method variability. The analytical method variability is estimated to be 10.5% for analytical samples and 22.0% for sediment extracts. Finally, responses appear dependent on the dilution of the final solution to be measured. This is hypothesized to be due to differences in dose-effect curves for different dioxin responsive element-active substances. For 2,3,7,8-TCDD, this effect is not observed. Overall, based on bioassay characteristics presented here and harmonized quality criteria published elsewhere (Behnisch et al., 2001a), the DR CALUX bioassay is regarded as an accurate and reliable tool for intensive monitoring of coastal sediments. 

Any analysis of biomonitoring data should include an assessment of the importance of sources of variation. Basic sources to be considered are laboratory, intra-individual, interindividual, and variability attributable to groups. 

Examples of quality assurance protocols that are considered standard practice in any data collection scheme include the use of both internal control samples (e.g. use of field blanks and spikes6) and external quality assurance samples (e.g. duplicate samples of known concentrations sent to different laboratories) to determine the extent of intra- and interlaboratory variation. Ensuring that the data have not been compromised or corrupted may also require setting up accessible data archives of original paper or electronic records so that the accuracy of summaries of the data published in documents and articles can be verified. 

There are various approaches that have been adopted for the use of AFs with the available toxicity data. They differ from authority to authority in the species number and type required and in the factors that are applicable to the given toxicity data (acute versus chronic). The various factors are summarized in Table 4.6. These factors reflect uncertainty in intra- and interlaboratory variation, intra- and interspecies variation, extrapolation between short- and long-term toxicity, extrapolation of results from laboratory to field, the possibility of indirect effects such as interspecific reactions (e.g., loss of predators, affecting prey), and the fact that environmental contaminants are often present as complex mixtures. 

The accuracy of exposure assessment is determined by systematic and random errors in the assessment. For quantitative exposure assessments, important sources of error include measurement errors (i.e. from laboratory and field monitoring techniques), as well as variations in exposure over time and space. For qualitative exposure proxies (e.g. self-reported past exposures, occupational histories or expert evaluations), the most important sources of error are recall bias (systematic differences in exposure recall between cases and controls) and random error, expressed in terms of intra- and inter-rater agreement. Although systematic errors can result in serious misinterpretations of the data, especially due to scaling problems, random errors have received more attention in epidemiology because this type of error is pervasive, and its effect is usually to diminish estimates of association between exposure and disease. The magnitude of random errors can be considerable in epidemiological field studies. 

Repeatability (intra-assay or within-day precision) is obtained when the analysis is carried out in one laboratory by one operator, using one piece of equipment over a relatively short time span. It reflects the variation in replicate procedures performed within a short time period, with the same operational conditions. 

The precision of an analytical method is the closeness of a series of individual measurements of an analyte when the analytical procedure is applied repeatedly to multiple aliquots of a single homogeneous volume of biological matrix [16], The precision is calculated as coefficient of variation (C.V.), i.e., relative standard deviation (RSD). The measured RSD can be subdivided into three categories repeatability (intra-day precision), intermediate precision (inter-day precision) and reproducibility (between laboratories precision) [16, 78, 79, 81],... 

---