Control sera

The sol-gel-derived membrane ISFETs based on bis(12-crown-4) and valinomycin were applied for Na+ and K+ assay in blood sera [27]. For reproducibility checking, the metalion assay was carried out ten times for each sample, using two different control sera. The ISFETs gave reliable data both in Na+ and K+ assay. The coefficient of variation ranges between 0.45 and 0.78%, thus indicating excellent reproducibility in the serum cation assay with the ISFETs of neutral-carrier-type sol-gel-derived membranes. The results for real blood sera for normal human for the cation assay showed relative errors within 2% for the Na assay and within 4% for the assay. 

Place a negative control sera slide immediately following one receiving a concentrated antibody ... 

Specific Antibody Determination. Serum samples were prepared from each bleed by centrifugation to remove clotted material. 100 ul of the sera was incubated for 30 minutes with sufficient H-STXOL to provide a ca. 20 fold excess of hapten to the anticipated quantity of specific binding sites. The radioactivity of the protein pellet was determined after ammonium sulfate precipitation. After correction for a small amount of non-specific adsorption of label by control sera proteins the mg/ml of specific antibody in the sample was calculated. 

Reagents of the Linco rat insulin RIA are used except the standards and control sera. 

During the 4 week period used to obtain control sera, determine the purity of a commercial avidin preparation using SDS-acrylamide gel electrophoresis as described in Chapter 5. Use only preparations exhibiting a single electrophoretic species in these experiments. 

Quality control with usual control sera Yes Yes Yes Not known No Yes Not known... 

In the case of reagents bound to carriers, permanent internal and periodically external quality control will partly result in problems. The reasons for some of these difficulties may be ascribed to the matrix of the control material employed. Since additions of synthetic or animal substances are incorporated in almost all control areas, these are not directly comparable with human material. This is the crux of the problem in dry chemistry. Excellent agreement is usually found to wet chemistry when comparing patients samples, whereas control sera can yield great differences. The use of whole blood (Reflotron) as control material is particularly problematic. This involves problems of stability, so that in such cases only so-called tertiary standards can be used, thus essentially limiting the accuracy. 

In practice, clinical laboratories are able to purchase materials from one of several companies that manufacture control sera or control products. These are generally sup-phed as liquid or lyophilized materials that are reconstituted by adding water or a specific diluent solution. Also available are materials having matrices representing urine, spinal fluid, and whole blood. Liquid control materials are also available and have the potential advantage of eliminating errors caused by reconstitution. However, the matrices of these liquid materials contain other materials that may be a potential source of error with some analytical methods and instruments. 

Rowan RM, Laker ME, Albert GMM. The implications of assaying external quality control sera under special conditions. Ann Clin Biochem 1984 21 ... 

Section 3 discusses quality control techniques, together with criteria for selecting satisfactory standards for the evaluation of a method and the subsequent use of standard materials as part of a control program. The tendency to use the word standard in several different contexts is potentially confusing, and has given rise to the misconception that control sera can be used at one and the same time both as standard materials and as an objective means of checking the quality of results. Because of the dangers inseparable from such potential confusion, it has been considered advisable to discuss some of the differences between standards and controls in this section on definitions. 

The control materials used in clinical chemistry include simple aqueous solutions of single compounds or mixtures of compounds (e.g., W13), and a wide variety of commercial protein-containing preparations (B6). For analyses of plasma or serum, protein-containing controls should be included among the specimens analyzed and these samples will be referred to as control sera. Under certain conditions, control sera can be used instead as standard preparations, and for some purposes there is no satisfactory alternative to standard sera for this (Sections 3.1.1.1 and 3.1.1.3). 

Even with control sera in which some of the constituents have been removed by dialysis or ultrafiltration and then returned as weighed amounts of pure material, it is essential to check the composition of the final product by the appropriate methods of chemical analysis, unless the use to which the control serum is to be put is distinctly circumscribed (e.g.. Section 3.1.1.2). Laboratories that prepare their owi) control sera establish the true analytical values, using the methods they employ, by performing a suflScient number (usually about 20) of replicate determinations under conditions that meet the definition of repeatability. 

Having stressed the importance of control sera as distinct from simple aqueous control preparations, it is strongly recommended that the staff responsible for the day-to-day performance of analyses on patients should in addition be provided with their own reliable means of assessing whether all stages of their techniques are probably working satisfactorily. For this purpose, each analyst should have available to him a pool of sera which may be called the bench standard. Although the true composition of this need not be determined, the analyst responsible for the par-... 

Consideration of the analyses performed on the 8-channel and both models of the 12-channel Technicon equipment in relation to the earlier discussion means that these latest developments in the field of Auto-Analyzer instrumentation demand standardized sera for calibration purposes. In fact, the successful operation of the SMA-12 instrument is entirely dependent upon the careful analysis and subsequent stability of the standardizing serum, since this is used both for the initial calibration of the various analytical channels and for the subsequent monitoring for drift and application of any correction needed as a result of instrumental drift. Apart from this large demand for standardizing serum (about 70 ml in an 8-hour day), the performance of the SMA-12 should in addition be checked by means of control sera, as for any other method or combination of methods in clinical chemistry. The expense of the standardizing (and to a lesser extent the control) sera used in SMA-12 operation constitutes an important but nevertheless essential fraction of the operating costs of these instruments. 

The use of anonymous control sera will be referred to as fictitious patient controls, since the staff are given no indication that they differ in any way from other samples in a batch. Even with AutoAnalyzers it is possible to include bias in the handling of samples known to be particularly important (e.g., care in the correct identification and reading of the relevant peak on a chart record), and this danger of bias influencing the results for control sera applies much more to manually performed methods. Fictitious patient controls overcome this tendency to bias, but it would be much more difficult to introduce them with equal certainty of anonymity into a laboratory with a small workload. 

The assessment of the precision of a method before its introduction into routine use requires a measurement of its repeatability, but the reproducibility cannot be determined until the method has itself been in routine operation for a sufBcient period of time (see Section 2 for definitions). Most published descriptions of techniques provide data about their repeatability when referring to precision, but there are some published figures specifically for reproducibility (e.g.. Cl, S7, T2). Having decided that a method is sufficiently reliable (on the basis of the SD for repeatability) to be introduced into routine practice, the SD for its reproducibility should be obtained as soon as possible, for instance by calculating the SD for the results of analyses on control sera included in the first 20 batches of determinations this SD reflects the between-batch variation. Subsequent monitoring of the method should then be related to this first value for reproducibility, if it compares favorably with the SD for repeatability. 

When duplicate control sera are analyzed in a batch, it may be found advantageous (especially early in the use of a technique) to plot an additional control chart showing the difference between the duplicate samples (G5), plotting means on one and differences on the other. This chart will display the within-batch error more clearly than the standard chart that plots the mean of the duplicates and is of greater use for monitoring between-batch error. 

In addition to these charts for control sera, charts can be used to display the results obtained with standards and with blanks (H7a) and, by using duplicate standards or controls, charts can be plotted showing the differences between duplicates (L3). These difference charts have been found valuable for detecting whether the variation between pairs of analyses performed on different days is significantly greater than the variations between pairs analyzed as part of the same batch. 

Blaivas (B4) described the use of an IBM-1710 process-control system, on-line to 20 AutoAnalyzers, in which the computer program included instructions for the acceptance or rejection of standard curves, and a series of checks for monitoring the acceptability of values obtained for control sera. Any results that failed to meet the requisite criteria were indicated within a few seconds of the charting of a peak by a message typed on an on-line typewriter. This system has the ability to include daily statistical assessments of the quality of results, but these possible additional control features were not listed in the description (B4). 

The routine performance of all determinations in duplicate is not advocated, particularly in the many laboratories where the pressure of repetitive work is already a major factor contributing to unsatisfactory quality of results. A limited number of duplicate determinations, however, is valuable, and some of these can depend on the repeated analysis of samples from individual patients. For the investigation of within-batch variation, patients samples can serve equally well as control sera and do not involve additional expense. 

Control sera prepared and analyzed in the laboratory. Various dilutions of these can be prepared to reduce the possibility of error of bias. 

Commercially prepared control sera that have had their analytical composition determined by selected reference laboratories. These can also be prepared in different dilutions. 

---