Trends

There is a tradition at year end to review and commemorate the passing of notables. The New York Times Magazine has its "The Lives They Lived" issue at the end of every year. In this spirit, and with more than a bit of mordant humor, we offer this obituary.

CLSI EP22: In Memorium

December 2010
James O. Westgard, PhD and Sten Westgard, MS

CLSI EP22 died a sudden death October 4, 2010.  The offspring of a union between Advamed and CMS, EP22 spent its life as a foster child of CLSI living in a group home governed by industry, government, and professional caretakers.  Born in 2004 in difficult circumstances, EP22 was raised with the hope of providing the solution to CMS’s early mistake with electronic QC, a mistake further complicated by CMS’s conception of “Equivalent QC” (EQC).  EP22 faced a difficult life from the beginning.  It was fast-tracked by CLSI, but proved unable to withstand a consensus process where there was no previous scientific evidence and no existing laboratory risk analysis practices upon which to build a “best practice.” Early on it was abandoned by FDA, which inflicted a systemic defect that could not be overcome.  It struggled to find its place first as a guideline for determining the frequency of QC and later as a guideline for manufacturer’s presentation of risk information.  But in the end, abandoned by industry, lacking commitment from laboratory professionals, the EP22 committee took its own life by a vote of 6 to 3.

Survivors include CLSI EP18, an older step-child of different parentage and heritage, and CLSI EP23, a younger sibling now faced with its own struggle to become a relevant guideline for medical laboratories.   Other family members include ISO 14971, which spawned the medical industry’s interest in risk analysis and fostered Advamed’s early recommendation to utilize risk analysis for EQC “Option 4”; ISO 15198, which provides guidance for the use of risk analysis for validating manufacturer’s QC instructions; and CLSI’s C24, which provides guidance for planning Statistical QC procedures.

The passing of EP22 leaves a void in the guidance for use of risk analysis to develop laboratory QC Plans.  While manufacturers are required to perform risk analysis as part of Good Manufacturing Practice and while that information on failure modes and related control mechanisms provides a sound basis for developing laboratory QC Plans to monitor residual risks, manufacturers do not need to disclose their risk analysis information to laboratories.  The importance of EP22 was to encourage manufacturers to provide that information in a standard format that would enable laboratories to customize their QC Plans.

Although EP23 survives and is expected to reach the age of maturity as an approved guideline by mid 2011, its useful life is now challenged.   Laboratories will need to request risk analysis information from the manufacturer or perform their own risk analysis to develop QC plans as alternative forms for laboratory QC.  CLSI has indicated there will be new siblings in the form of educational and other supportive products to assist laboratories with implementation of EP23.  An additional challenge for accrediting organizations is how to review and evaluate these customized QC procedures.

The survival of EQC is now much more likely, in spite of the lack of scientifically reliable validation protocols.  CMS is expected to update their interpretive guidance in the State Operations Manual in 2011, but has not made any declaration on the future of EQC, whether it will (a) continue, (b) be replaced by the EP23 guidance, or (c) include both EQC and EP23 as options for “alternative QC.”  The difficulties of performing risk analysis or obtaining risk analysis information from manufacturers conspire to make EQC the easiest option for reducing QC.

The passing of EP22 should also remind us of the importance of defining appropriate QC procedures, the dangers of abandoning tried and true SQC procedures before alternatives have been proven safe, and the limitations of the consensus process when there are conflicting interests between industry, government, and laboratories.   Medical laboratories must always exercise their own responsibility for “verifying the attainment of the intended quality of results” (as stated in ISO 15189) to provide quality test results for our patients.

Editors' Note: We don't want to diminish the effort and dedication of the individuals who spent years trying to guide EP22 to publication. A lot of people worked very hard and expended substantial effort to try and get this guideline to come to fruition. We do not mean to mock them. In the end, however, the heroism of individuals was not enough to overcome the resistance of the group. We wish the participants of EP22 well and look forward to the emergence of a more successful standard in the future.