IVD Technology Magazine | IVDT Article Index
Originally published January, 1997
Commentary
Quality control: What system is right for IVDs?
Judith J. Smith
Ever since the Clinical Laboratory Improvement Amendments of 1988 (CLIA) were implemented, the Centers for Disease Control and Prevention (CDC) have been wrestling with the issue of quality control regulation for in vitro diagnostic products (IVDs).
CDC has been obtaining input from many groups as part of its ongoing program to address this subject. The effort was expanded last September, when CDC sponsored a workshop to gather information from industry and user representatives regarding current quality control systems and practices. Specifically, CDC requested comments on the following concerns:
- The ability of current test systems to assess the potential for error in the total testing process.
- The types of processes necessary to monitor the quality of clinical laboratory results over time.
- The types of processes necessary to assess appropriate test performance by operators.
In response to this call for comments, the Association of Medical Diagnostics Manufacturers prepared a presentation on the first and third of the topics under discussion at the workshop. Founded in 1973, AMDM is a trade association made up of in vitro diagnostic manufacturers representing a wide range of clinical laboratory products, technologies, methodologies, and analytes. Its presentation at the workshop was based on a recognition of the complexity of the IVDs and QC procedures represented among its membership.
The current requirements for QC testing are spelled out in several sets of regulations and standards. Unfortunately, these documents sometimes appear vague or contradictory. For instance, the CLIA regulation requires that laboratories "perform and document control procedures using at least two levels of control materials each day of testing" (42 CFR 493.1202( c)(4)). The Health Care Financing Administration's section on "Survey Procedures and Interpretive Guidelines for Laboratories and Laboratory Services" states that "instrument or procedural control checks (at least two levels) may be used to meet this requirement" (State Operations Manual, app. C, p. 108). The College of American Pathologists Commission on Laboratory Accreditation's inspection checklist for diagnostic immunology and syphilis serology says that "for certain reagent systems with built-in controls, external controls may not be required. Such systems must have been classified as moderate complexity or waived tests under CLIA '88 and the internal control must measure reactivity, not process." These somewhat contradictory statements have created confusion and uncertainty for diagnostic manufacturers and field inspectors alike.
One of the reasons such standards and regulations can be confusing is that the world of IVDs encompasses a tremendous range of complexity. This fact is well known among those in industry, and has been highlighted by the struggle to categorize IVD tests according to their level of complexity, as required by CLIA.
Current IVD products range from manual tests containing several operator-dependent steps, to manual tests unitized with minimal operator-dependent steps, to automated tests containing several operator-dependent steps, and finally to fully automated systems requiring little operator intervention. To make matters more complicated, IVDs tend to evolve over the course of their effective lifetimes, beginning as manual tests and becoming increasingly automated as they undergo the constant process of product improvement. Thus, IVDs are not only complex in their initial format but are also dynamic in their individual complexity over time.
Overlying these wide differences in the complexity of their technologies, IVDs also display a wide range in the quality control procedures developed by their manufacturers. IVD quality control systems can range from none at all, to monitors of all internal systems, to systems that monitor everything short of adding the analyte to a sample cup. Manufacturers of manual assays with several operator-dependent steps sometimes provide or recommend external quality controls to monitor all reagents and steps. Manufacturers of manual assays with minimal operator-dependent steps may incorporate into an assay controls that monitor its vital reagents but are not needed for operator monitoring. Manufacturers of automated assays with several operator-dependent steps may recommend some type of quality control procedure to monitor the operator steps, or may include separate built-in controls to monitor instrument function. And finally, manufacturers of automated assays that require little operator intervention may build in controls that monitor instrument and reagent function and that, once again, don't require operator monitoring.
The results of all these varieties of quality control procedures and technologies may also provide a wide range of data. Some results may be strictly qualitative, while others are either semi- or fully quantitative. The implications of these variations on the need for operator interpretation are obvious.
Although CDC's effort to develop a single, clear-cut requirement for IVD quality control is a worthy one, the wide variations in current technologies, monitoring systems, and results are likely to make such an effort extremely frustrating. Given this complexity, the overriding question posed by CDC's September workshop--what type of quality control system should be recommended for all IVDs?--is difficult to answer in a single recommendation. The answer, it seems, is "it depends." A number of factors are involved, including the following:
- What needs to be monitored based on reagent composition.
- How sensitive the test's reactions are to environmental changes.
- How many and what type of operator-dependent steps are used.
- What shelf-life studies have been performed by the manufacturer in support of product performance over time.
For this reason, AMDM believes that no single system can be established for all IVDs. At the CDC workshop, the association presented an alternative proposal that would take into account some of the many variables that characterize the world of IVDs. AMDM recommended that a task team representing FDA, industry, and user groups be appointed to establish technology groupings of IVDs and to recommend general QC procedures for each. Once the groupings had been established, the team would meet periodically to review changes in technology and QC systems, and to adapt the groupings and recommended QC procedures as appropriate.
This system would provide manufacturers with specified levels of quality control toward which they could direct their product development efforts. By meeting the QC requirements specified for a particular grouping, manufacturers would be able to reduce their burden of proof at the end of the product development process. Moreover, the task team would be able to identify quality control issues relevant to particular groupings so that manufacturers could develop QC systems specifically to address them.
If the program were implemented appropriately, new technology groupings could be added within the same regulation, thus providing flexibility without creating the need to revise the regulation. Until the QC groupings were established, users would follow the manufacturers' QC recommendations as described in each product's package insert.
Reaction to the AMDM recommendation was mixed. Many participants supported the association's approach, agreeing that the tremendous diversity of IVD products precludes the use of a one-size-fits-all quality control method. They also commented that the slow progress of change in regulations should be taken into account, and that whatever regulation is developed now must be flexible enough to keep up with rapidly evolving technologies in the future.
Others at the CDC workshop disagreed, arguing that products already undergoing FDA 510(k) review and CDC complexity categorization should not have to undergo a third product-specific level of review. This criticism permitted the association to make it clear that its proposal was not intended to create another product-specific system, but to address broad technology categories of IVDs.
In the opening remarks of its September workshop, CDC presented the key elements of the CLIA regulation as a three-legged stool--quality control, personnel, and proficiency testing. If all three of these legs are in balance, then quality assurance is achieved. CDC's difficult task is to revise QC requirements while maintaining that balance. To do so, it plans to review all the comments it has received from various forums and to develop a proposal by early this year. IVD manufacturers and users look forward to reviewing the proposal and providing additional input.
