Trends

A $100,000,000 fine? What's up? The 1999 Abbott-FDA Consent Decree resulted in a such a hugh penalty because of a long standing failure to comply with FDA's Good Manufacturing Practices or Quality System Regulation. How does this affect you? Dr. Westgard explains how the Consent Decree should alert laboratories and manufacturers of the importance of an independent laboratory quality system that assures the quality of routine test results.

• Just the facts - the background of the Abbott-FDA consent decree
• A sentinel event calling for independent laboratory quality management
  • Analytical quality is not a given!
  • Responsibility for quality can't be outsourced!
  • Compliance is not enough!
  • Statistical QC can't be eliminated!
  • Minimum personnel standards and skills are not sufficient!
• Help is available - you can even get some of it for free
• Independent laboratory QC is still dependent on manufacturers
• Conclusion

On November 2nd, 1999, Abbott Laboratories and the Food and Drug Administration entered into a legal agreement - a consent decree - to settle a dispute about manufacturing practices. This agreement called for Abbott to stop manufacturing diagnostic test kits by December 6, 1999 and to make corrective changes in the manufacturing processes in its Lake County IL facilities. On November 19, 1999, the Northern District Court of Illinois extended the deadline till January 10, 2000.

Just the facts - the background of the Abbott-FDA Consent Decree

In the FDA "Dear Colleague Letter", the consent decree was described as a permanent injunction that requires Abbott to discontinue sales of certain tests within thirty days because of long standing failure to comply with FDA's Good Manufacturing Practices (GMP) or Quality System Regulation (QSR) and its failure to fulfill commitments to correct deficiencies in its manufacturing operations.

In a press release, Abbott pointed out that the consent decree does not require the recall of any diagnostic products and that certain products will continue to be available.

"The decree allows for the continued manufacture and distribution of medically necessary diagnostic products made in Lake County, Ill., such as assays for hepatitis, retrovirus, cardiovascular disease, cancer, thyroid disorders, fertility, drug monitoring, and congenital and respiratory conditions. However, Abbott is prohibited from manufacturing or distributing certain diagnostic products until Abbott ensure the processes in its Lake County,Ill., diagnostics manufacturing operations conform with the current Quality System Regulation."

The November 3, 1999, Washington Post brought the issue to the public's attention with it's article on "Medical Test Maker Fined $100 Million for Poor Quality Control," in which it stated:

"Abbott Laboratories, the nation's largest maker of medical laboratory tests, agreed yesterday to pay a $100 million fine for failing to correct defects in its manufacturing processes despite six years of warnings. The penalty was the largest ever levied by the Food and Drug Administration."

The LA Times also drew public attention with its article "Abbott Labs to Pay $100m to FDA."

"The issues have to do with manufacturing processes and not with known instances of patient harm,' said FDA medical devices chief Dr. David Feigal. Under the consent decree, Abbott is permitted to continue selling only certain tests the FDA deems medically necessary… The FDA offered an important warning for laboratories, which have certain quality-control to catch glitches with products. Labs that continue to use Abbott diagnostic products should 'increase their vigilance in these areas, to make sure these tests are performing,' Feigal said."

Additional information for the laboratory community is available from professional organizations. AACC delivered three MANAGEMENT ALERT bulletins via fax broadcasts in the days following the consent decree and established a webpage with sources of information. In a fax broadcast on November 9, CLMA announced it was petitioning the FDA to extend the notice before discontinuing sales of the tests. If granted, the petition would extend the thirty day period by an additional fourteen days. CLMA also established a special page on its website (http://www.clma.org/) and organized a special audioconference November 18, 1999 to focus on practical strategies for transitioning products affected by the consent decree.

A sentinel event calling for independent laboratory quality management

The FDA-Abbott consent decree sounds a warning that laboratories need to independently assure the quality of the test results if they are to maintain the services needed by their customers. Virtually all laboratories are affected because of Abbott's dominant role in supplying materials for diagnostic testing. These laboratories are faced with the immediate need to validate new methods and provide better quality control of routine testing. For many laboratories, the personnel resources and skills needed to accomplish this are now very scarce due to changes in laboratory management practices and priorities during the last decade.

The need for independent laboratory quality control is fundamental! It begins by having control materials that are independent of the calibration materials. It is also good practice to have control materials that are made by a different manufacturer than the test system, as confirmed in this situation by FDA's recommenation for "the use of quality control materials made by other companies to increase the assurance that these devices are performing successfully". Furthermore, laboratories should establish their own means and standard deviations when implementing statistical QC procedures, rather than use bottle values provided by a manufacturer. And laboratories should select appropriate control rules and numbers of control measurements on the basis of the quality required by the patients and physicians they serve and the imprecision and inaccuracy observed for their methods during routine operation in the laboratory [1].

It's important to recognize that the problem here isn't Abbott. The current predicament could have occurred with any manufacturer. The real problem is the poor quality management practices that have evolved in laboratories during the last decade. Laboratories aren't ready or able to exercise an independent responsibility for the quality of their own test results. Laboratories have their own quality system problem.

The consent decree should be a sentinel event to alert laboratories to the serious limitations of their current quality management practices. Government agencies should reflect carefully on the state of laboratory quality management and the susceptibility of laboratory testing on a national scale to a problem from a single manufacturer. Accreditation and professional organizations also need to examine their priorities to be sure that the basic requirement for correct test results hasn't been down-played in their efforts to focus laboratory attention on utilization and outcome.

Analytical quality is not a given!

The consent decree clearly exposes the fallacy of the current myth that "analytical quality is a given." I've been worried about the complacency that has been developing in the 90s - complacency in laboratories, complacency in professional organizations, complacency in regulatory agencies. This issue has been addressed on this website with the essay The Myths of Quality, which also is the introductory lesson in our course on basic method validation [2]. In fact, our educational efforts for the past two years have been focused on "back to basics", first dealing with Basic QC Practices and then Basic Method Validation, then Basic Quality Planning. Together, we hope this series will provide the fundamental background needed for laboratories to manage analytical quality in an independent way.

Responsibility for quality can't be outsourced!

In today's busy laboratories, as managers and analysts, we are quite happy when they can find ways to reduce their activities, such as bundling quality management activities into contracts for purchasing analytical systems. For example, when a laboratory buys a new instrument, we expect the manufacturer to not only to install the system, but to perform the method validation studies needed to satisfy CLIA validation requirements, provide the QC instructions, control materials, method manual procedures, and inservice training. Manufacturers provide these services to satisfy the "wants" of the laboratory customers.

In effect, laboratories are outsourcing many quality management activities by purchasing complete systems from a manufacturer. That practice can be dangerous. The responsibility for the quality of the test results can never be outsourced, even if some of the activities are performed by others. Quality can be guaranteed only when the laboratory accepts its responsibility and properly manages its own testing processes.

Compliance is not enough!

"Compliance" is the new management focus as we approach the next millenium. Professional organizations tout compliance training programs. Scientific meetings feature compliance workshops. Laboratories are required to develop compliance policies and programs and may even have compliance officers. What does compliance mean? Do this, and no more, and we comply with the law. We avoid being charged with fraud and we stay out of jail. Does anyone think this inspires us to improve or even maintain quality?

Compliance management actually got its start from CLIA. It isn't mandated by the CLIA regulations, but compliance has certainly been the main response to the regulations. Its evident by the management response that the minimums stated in regulations are all that needs to be done, therefore these minimums become maximums in laboratory practice. For example, if the regulations say it's necessary to run two controls, then if we seek only the bottom-line, we never run more than two. We seem to accept that the government has become very smart in the area of quality control, even though we don't think the government has any clue when it comes to about anything else.

The CLIA rules might provide good guidance for quality management if the regulations were ever completely implemented. It would be nice to have the "final-final" CLIA document published. It would be even better if the FDA were to implement a process for QC clearance. The continual postponement of the regulations dealing with validation of manufacturer's QC is a serious problem and compromises the overall effectiveness of the regulations.

Statistical QC can't be eliminated!

Statistical QC has been a fundamental quality management technique in industry since the 1930s and in laboratories since the 1960s because it provides an independent way of monitoring the quality of a production process. In spite of this, there have been efforts to do away with statistical QC for laboratory testing. These efforts have been initially pushed by industry to expand point-of-care markets. However, statistical QC still provides the most efficient technique for monitoring many steps in a testing process, as well as providing a quantitative measure of operator proficiency, thus it is still the most essential technique for analytical quality management [3]. Expanded quality management systems should be built on statistical QC, rather than attempting to eliminate statistical QC.

Minimum personnel standards and skills are not sufficient!

CLIA's minimum requirements for personnel have now been achieved in many laboratories. Again, because of the compliance mentality, the net result has been a lower level of the education for laboratory analysts, a serious reduction in training programs for clinical laboratory scientists, and, of course, some reduction in personnel costs, which justifies everything for bottom-line managers.

It has been hard to fight this "dumbing down" trend; even some laboratory professionals have capitulated. One paper published in 1997 [4] concluded that "QC is costly, and laboratorians frequently do not follow established QC practices, in part, because they are complex. To improve compliance, we believe QC practices must be simplified." In short, if the job is too complex for lower skilled people, the solution is to simplify the job. (Let's keep this little secret to ourselves because I don't want my plumber taking over for my surgeon!)

Help is available - you can even get some for free

All is not lost. We can stop the dumbing-downward cycle whenever we want. Laboratories that want to upgrade their skills in analytical quality management can take advantage of many of the Internet training materials and tools that are available on this website. Here are some of the key materials for getting started:

• Review of basic method validation practices. The lesson MV - Real World Applications provides a complete review of the method validation process, with links to additional materials, such as lessons on the comparison of methods experiment, the detection limit experiment, the method decision chart, and the transference of references intervals.
• Review of statistical data analysis. The lesson MV - The Data Analysis Tool Kit provides guidelines for the use and application of statistics for analyzing data from method validation experiments.
• Calculation and plotting tools for method validation data. The Data Analysis Tool Kit provides a Linear-data Plotter that can be use with data from a linearity or reportable range experiment, an SD calculator for data from a replication experiment, a Paired-data Calculator for method comparison data, and a Method Decision Calculator to help judge method performance.
• Review of basic QC practices. The lesson QC - The Practice provides a complete view of basic QC, including the purpose, planning, implementation, and operation of statistical QC procedures, together with links to more detailed lessons and materials.
• Review of QC planning approaches. The lesson QC - The Planning Process provides an overview of the available approaches for selecting appropriate control rules and an appropriate number of control measurements. There are links to example applications and internet tools, such as a Normalized OPSpecs calculator.
• QC calculation and plotting tools. Internet tools for QC training provides an overview of a suite of available interest tools, which include a QC Calculator, QC Plotter, QC Simulator, QC Trainer, and QC Checker. The last tool - QC Checker - allows you to set up a control chart, select control rules, enter your own QC data, and use the tool to check for violations of the selected control rules.

Independent laboratory QC still dependent on manufacturers

Laboratories are dependent on manufacturers to provide the test systems needed in today's laboratories. There is a shared responsibility to deliver the test results that are needed for patient care. An important part of the laboratory's responsibility is to assure the quality of the final test results before they are released and reported. Manufacturers can help the laboratory carry out it's responsibility by supplying QC materials and decision-support software that will allow the laboratory to easily implement and efficiently operate the appropriate QC procedures. However, the laboratory is still responsible to select and implement QC procedures that will assure the quality needed for the patients it serves.

Conclusion

In the wake of the Abbott-FDA Consent decree, we have two options: to get by or to get prepared. It may be possible to stockpile enough Abbott reagents to get by, then hope that this problem never happens again. Or, we can decide to get prepared, take the steps to improve our quality management knowledge and skills, and implement an independent laboratory QC program that will guarantee the quality of our test results. If the government inspectors appear, we can welcome them knowing we're doing our jobs and managing quality in a responsible way.

References

1. CLSI C24-A3. Statistical Quality Control for Quantitative Measurements: Principles and Definitions; Approved Guidelines - Second Edition. Clinical Laboratory Standards Institute, Wayne PA, February 2003.
2. Westgard JO. Myths of Quality. Basic Method Validation. Westgard QC, Madison, WI 1999, pp 3-8.
3. Westgard JO. Taking care of point-of-care QC. Clin Lab News, August 1997.
4. Howanitz PJ, Tetrault GA, Steindel SJ. Clinical laboratory quality control: a costly process now out of control. Clin Chim Acta 1997;260:163-174.

James O. Westgard, PhD, is a professor of pathology and laboratory medicine at the University of Wisconsin Medical School, Madison. He also is president of Westgard QC, Inc., (Madison, Wis.) which provides tools, technology, and training for laboratory quality management.