You are here:
QUALITY ASSURANCE IN RESEARCH LABORATORIES: RULES AND REASON
Citation:
Rogers, R R. QUALITY ASSURANCE IN RESEARCH LABORATORIES: RULES AND REASON. Presented at 23rd Annual National Conference on Managing Environmental Quality Systems, Tampa, FL, April 13-16, 2004.
Description:
Quality Assurance in Research Laboratories: Rules and Reason
Ron Rogers, Quality Assurance and Records Manager, Environmental Carcinogenesis Division, NHEERL/ORD/US EPA, Research Triangle Park, NC, 27709
To anyone who has actively participated in research, as I have (16 years at bench level plus 15 years in various administrative/management roles), there is a clear and logical distinction between research and testing. "Research" often is hypothesis driven; also often characterized as creative, evolving, dynamic exploration and discovery - often changing in response to observations. The 'omics technologies that are changing the landscape of life science research are good examples. On the other hand, "testing" is generally carried out under strict control, primarily under contract, using well defined and validated procedures that are approved by a sponsor for use in the "project" or "study". A portion of intramural ORD research studies fall very close to that category also. This issue is not unique to EPA or ORD, but will be discussed from that perspective because that is where we directly see the need to ensure that appropriate QA policies and practices are in place for application to research.
The graded approach that EPA permits in QA planning was primarily designed to accommodate "projects" or "studies" that differ in scale/scope, goals (i.e., anticipated data use), progression phase, etc., but it must also be able to accommodate the divergent quality issues that exist between "research" and "testing" studies. Achieving that flexibility while remaining faithful to the Agency's QA mandate is not a trivial matter, but in the interest of the quality of ORD science, we must continue to preserve it. A prime example of where flexibility is well within reason is the required QA Project Plan (or equivalent), which is often difficult to prepare at a useful level of QA detail before research is underway and preliminary data gathered, but if employed appropriately, is always an extremely useful element of the completed study record. Preserving the flexibility to focus on the research record rather than just on predicting the course of research will help maintain the credibility of the ORD QA Managers and Quality Systems overall.
As we face challenges in migrating toward new research technologies and increased electronic data collection and records management, we have both an urgent need and an opportunity to build appropriate QA rules and reason into those systems that will be used. There are valid reasons (most originating from the contractor-EPA sponsor issues) for the rules established in the EPA QA Manual, but not all of those reasons apply to all data collection activities. Research QA Managers in ORD, with Quality Staff concurrence, need to define (and heed) those distinctions and/or exceptions in order to reestablish and preserve the flexibility to permit reason and experience to control interpretations and implementation in research programs. ORD QA staff may recognize and choose to focus QA efforts on a subset of those Manual issues or on other internal issues that are seen to present greater potential benefit to the quality and defensibility of our research data. Failure to preserve reason and flexibility in our actions could result in the risk of erosion of the credibility and trust that ORD QA Managers have gained in the past decade from the scientists whom we support. That trust has been gained in part by expressing our confidence that our scientists, unlike the contractors of the '70s whose misdeeds led to EPA's Agency-wide QA mandate, have no bottom-line incentive to fabricate data or ignore errors - their professional reputations are based on those data. Although my QA reviews occasionally result in my detecting errors, I view my primary QA role not as policing ECD researchers' actions, but as supporting them (and ECD managers) in finding and implementing opportunities for improving the quality and defensibility of their data and other records.
This abstract does not necessarily reflect US EPA policy.