QC - THE MATERIALS
|
QC - THE MATERIALS |
|
PLEASE NOTE: An updated version of this lesson is available in Basic QC Practices, Second Edition
The purpose of a statistical quality control procedure is to monitor the analytical quality of the measurement during stable operation, detect changes from the stable operation, and eliminate reporting of results with medically important errors [1]. From the technologist’s standpoint, the objectives of the control procedure are simply to "alert me when the method has a problem" and "don't alert me when the method is working okay." These correspond to "true alarm" and "false alarm" situations, which are characteristics of the QC procedure. In short, technologists want to know about real problems, but can't afford to waste time when the method's working okay. Any additional information that can aid in the troubleshooting the method is an "extra bonus."
Much has been written about how to perform the statistical calculations needed for quality control procedures, how to choose control rules, how to apply the rules, how to construct control charts, and how to interpret results from the control procedures, all of which assume we have found appropriate stable control materials. We need to pay careful attention to the selection of the control materials. Important attributes are the stability, vial to vial variability, assayed versus unassayed, appropriate analyte levels, and pretreatment procedures. The very success of the control procedure depends upon these attributes!
The International Federation of Clinical Chemistry defines a control solution or control material as a "specimen or solution which is analyzed solely for quality control purposes, not for calibration" [2]. We use the term control material or control product to refer to a control solution that is available, usually commercially, in liquid, frozen, or lyophylized form, packaged in small bottles suitable for use on a daily basis. Such control materials are widely available today for most laboratory tests. They can be obtained from manufacturers who specialize in the production of control materials, and also are often provided by the same companies who are selling you the reagents, methods, and instrument systems. It is common today to purchase complete testing packages that include the necessary control materials.
Matrix refers to the substance or base from which the control material is prepared in addition to all the additives such as spiking materials, preservatives, etc. added to make the product desirable to the user. The American Society for Testing and Materials has defined matrix as " the principal element or elements in a sample" and a matrix interference as "an effect due to the presence of a constituent or characteristic."[3]
Ideally, control materials should have the same matrix as the specimens being tested so that they will behave the same as the real specimen. For example, choose a whole blood control for point-of-care blood glucose analyzers and for blood gas and whole blood electrolyte analyzers to maintain a similar matrix; use a serum/protein based control for analyzers that perform tests on serum or plasma. Control materials are also available having matrices of urine and spinal fluid. In general, materials prepared from human sources have been preferred in the past, however because of the potential biohazard risk today, bovine based control products have become more popular.
Control materials, even when selected to have the appropriate matrix, undergo a substantial number of manipulations during their production which can alter the properties of the matrix. These alterations include human and non-human additives for achieving specific concentrations and/or stability as well as physical changes to the material such as freeze drying or lyophilization. These alterations may in turn cause interferences in the testing process which may not be present in fresh human samples.
Certain test methodologies may also influence the selection of control materials. For example, a bovine based control material will usually assay low by a bromcresol purple albumin method, which has been optimized for human albumin. In contrast, bovine controls are acceptable for use with the less specific bromcresol green albumin method. For some assays, such as those used in lipoprotein testing, fresh or frozen human pools may be the most appropriate control materials [4]. Careful consideration to the control material matrix is an important consideration in the QC planning process.
When possible, it is desirable to purchase at least a one year supply of the same lot or batch number. Many products are now available with expiration dates of more than two years. The desired expiration date of the control product should be included in the specifications listed at the time of purchase. This planning step will pay off in the ability to provide a continuous monitor of the analytical process through many method and instrument changes, while reducing costs by minimizing the crossover testing necessary during the checking out period of new control lot numbers. It usually isn't necessary to purchase and store the entire lot for the duration of the expected time of use because most vendors are willing to sequester specified lot numbers for the desired period and set up automatic shipment and billing schedules at monthly, every other month, or quarterly intervals. This strategy also has the advantage of not requiring payment or inventory of the control materials until they are needed for use.
The variation observed when monitoring a method is almost entirely due to measurement imprecision and vial to vial variability of the control materials themselves, which is usually a small part of the total variation observed. Commercial control materials that have been lyophilized or freeze dried must be reconstituted with water or specialized diluent, therefore, it is very important to standardize the reconstitution step. Use Class A volumetric pipettes, deonized Type 1 water, and instructions that specify the mixing time and the reconstitution time to minimize the vial to vial variability due to the preparation process.
Many liquid control products which eliminate the reconstitution process are now available. These products are generally more expensive and sometimes contain additives or preservatives which could introduce sources of error due to matrix problems with certain methods. Depending upon the analytical methods to be monitored, the benefits of the reduced vial to vial variability may outweigh any increased costs. In addition, liquid control products are generally stable for 14 to 30 days after the vial has been opened whereas lyophilized products are usually only stable for less than 48 hours after reconstitution. Liquid controls may therefore, in some cases, be a better buy because of reduced waste due to stability, elimination of vial to vial to variability, and reduction in operator errors due to the reconstitution process.
Control products are available as assayed or unassayed materials.
Assayed control materials generally come with an assay sheet
of expected values for analytes assayed by various methods and
instrument systems. These assay sheets usually list, for
each constituent present, expected mean values as well as
expected ranges. Values may even be available for reference methods
used to measure certain analytes. These ranges are provided only
as guidelines until the laboratory has established its own statistical
limits. Assayed controls are generally more expensive than
unassayed controls due to the cost incurred with the value-assignment
process. They may, however, be valuable for the smaller laboratory,
for meeting CLIA regulations, and for troubleshooting
method problems.
Constituent levels of quality control materials should be chosen at medical decision concentrations and/or at critical method performance limits such as upper and lower linearity limits. Two or three different concentrations are often needed for each analyte. Choosing control materials at critical concentrations (medical and/or performance) will allow the analyst to estimate the random error at critical levels of the method during stable operation and to provide a monitor at the most important performance levels for that analyte. Statland has provided recommendations for decision levels for many tests [5], which are summarized in tabular form on this website.
Vendors manufacture sets of materials to have critical concentration levels and also to monitor the working range of a method. Purchase of these materials will often include the services of a quality control program. Services provided by these quality control programs include statistical calculations of your laboratory data, preparation of Levey Jennings plots using your mean and standard deviation, and a comparison report which compares your monthly and cumulative statistics to those of a peer group of users using same lot number of materials. Costs for these programs are usually included in the cost of the control materials. An alternate option is to select materials from different vendors in order to minimize possible matrix problems that may be inherent in all materials from a single vendor.
Many laboratory tests such as digoxin, hemoglobin A1C, and total iron binding capacity require pretreatment of the specimen prior to determination by the analytical process. These procedures often require manual pipetting and mixing steps which are more problem prone that the analytical determination. It is best, for these procedures, to have control materials that are submitted to the pretreatment steps and are therefore treated exactly as the specimen being tested. If the analytical method is also problem prone, it may be advantageous to select one or two control materials which do not undergo the pretreatment process in addition to the one or two materials which are included in the pretreatment. This strategy will aid the analyst, when errors occur, by isolating the analytical process from the pretreatment process.
The selection of appropriate control materials requires consideration of many factors and should be part of the planning process for the selection and implementation of SQC. The process becomes even more complicated when selecting control materials suitable for a multiconstituent analyzer. Compromises may have to be made in order to limit the number of control materials used. There is no right or wrong way to choose control materials for a given method, just as there is no perfect control which behaves exactly the same as a fresh human sample. Selection of appropriate control materials is a balancing act in which cost, stability, ease of use, performance due to matrix effects, and constituent levels must be considered and weighted for each laboratory’s own application and use.
Control materials being used should be reassessed yearly. Product vendors are continually making changes and enhancements in their products so that these products are more marketable. Presently, it is hard to keep abreast of all the changes in immunoassay control products as vendors are scurrying to provide materials with constituents to fit the test menus of these new analyzers. Other quality control services and enhancements are also being provided by instrument vendors. Some instrument manufacturers will be offering quality control troubleshooting services by using a modem hookup to review QC measurements from the instrument at the laboratory site. Of course, control materials purchased from the instrument vendor must be used to obtain the service. The selection of control materials is an important consideration in the statistical quality control planning process and must be reviewed as our needs change.
