Questions of Quality

Quality continues to be an issue for many laboratories. In some countries, laboratories are only just developing quality management systems, processes, and procedures, not to mention practical implementation tools and techniques; in other countries, laboratories are so busy, and analysts have such heavy workloads, that finding time for quality has become more and more difficult. Often, the business strategy in laboratories is to “do more with less” – which means less time for activities such as performance evaluation, quality control, and quality improvement.

Improvements in QC practices won’t lead to better quality management if laboratory analysts lack the time to learn new techniques and properly implement them, though. An example is the new US CLIA option to use Individualized QC Plans (IQCP) that use risk assessment to identify potential failure modes and then select appropriate control mechanisms to detect those failures. The approach sounds good in principle, but in practice it lacks the basics: standardized risk assessment methodology, documentation of the performance of control mechanisms (other than statistical QC), objective assessment of the acceptability of residual risk, and mechanisms for review by inspectors in accreditation programs. These shortcomings exist in spite of a new CLSI standard: EP23A “Laboratory Quality Control Based on Risk Management” and a new CDC/CMS “Step-by-Step Guide” for developing an IQCP.

So, although the existence of guidelines has influenced manufacturers to implement risk management, and despite manufacturers then persuading organizations like CMS and CLSI to adopt a risk management approach for quality control, the fact remains that laboratories generally have no formal training and little experience in risk assessment. They have to learn risk management on the job – which is time-consuming and often problematic.

Not all laboratories are the same, however. For example, those situated in developing countries usually show more interest in improving the quality of their testing services and often invest more time and effort in it. By contrast, laboratories in developed countries are sometimes less interested in QC and quality improvement because of the increased pressures for productivity and efficiency that they face.

There are also differences in practices depending on how laboratories are inspected and accredited, and this too can vary geographically. For example, the US follows the “CLIA Rules” for inspection and accreditation, whereas many other countries are adopting the ISO 15189 standard for quality and competence in medical laboratories. That ISO standard adheres to certain metrological concepts, which in turn requires adoption of new quality practices. In particular, laboratories have traditionally embraced the TAE (Total Analytic Error) model for understanding, measuring, and managing quality, but ISO 15189 instead recommends the use of the metrology “uncertainty model” (Measurement Uncertainty, MU). MU has been advocated for many years now, especially in Europe, but has yet to find broad acceptance and practical use in service laboratories, and is mainly a requirement for accreditation by ISO 15189. So the field is still not adequately harmonized.

This complexity means that many organizations find that they need outside support to develop, understand or meet quality requirements; but the different organizations may themselves bring different challenges. In the US, I have found the government agencies particularly taxing, as it seems that they mostly listen to trade organizations, then to large manufacturers and then professional organizations. Laboratory practitioners are the last to get heard, and this can lead to problems. The change of QC practices from EQC (Equivalent QC) to IQCP is a good example. In the original proposal from industry, manufacturers were to provide laboratories with risk information pertaining to their analytic systems, so that the laboratory could develop risk-based QC plans accordingly. Two CLSI documents were developed – one for guidance to manufacturers on how to present their risk assessment information to laboratory users, and another for guidance to laboratories on how to use that risk assessment information to develop QC plans. CLSI submitted both of these documents for review, but only the one intended for laboratory guidance was finally approved; the document for manufacturers was canceled, because manufacturers dissolved the committee! Thus, it remains the case that laboratories lack sufficient information from manufacturers about the performance and expected failure modes of manufacturers’ controls. Nonetheless, CMS adopted IQCPs even though laboratories lack the risk assessment information to properly plan and develop an IQCP. So we’re getting there, but slowly; perhaps the quality of patience is an essential attribute in QC!