For many years, Environmental Standards, Inc. (Environmental Standards) has conducted approximately 100 commercial laboratory audits each year for its clients. These are at-the-bench audits – we cover both general laboratory operations, but also spend considerable time with Analysts to go through their procedures step by step. By far, the majority of these audits are performed at domestic United States (US) analytical facilities. However, there is a growing interest from our multi-national clients to assess the quality of their analytical service providers around the globe. Within the past month, I had the opportunity to perform two laboratory audits in Brazil, while my colleague provided audits in Indonesia and Malaysia. Let’s take a look at the environmental laboratory approach and quality system that is used in Brazil and compare that with the US approach.
The majority of commercial analytical laboratories that are accredited in the US perform under the consensus standards and accreditation system currently operated through The NELAC Institute (TNI). Though this is not universally true, as some US States have developed their own accreditation system and standards, the TNI program (formerly known as NELAP) is the most widely followed environmental accreditation program within the US. In Brazil, the National Institute of Metrology, Quality and Technology (Inmetro/cgre) is the laboratory accreditation agency, and like the TNI system, Inmetro’s is based upon the ISO 17025 standard. Note that Inmetro also accredits Brazilian laboratories for non-environmental services using other ISO standards and the Good Laboratory Practices.
We often hear the term “data of known and acceptable quality” as the standard for a quality system with environmental laboratories. The “known” refers to having procedures, following those procedures, and proving that they are followed and this is the primary emphasis of ISO 17025. “Acceptable quality” implies that metrics have been established so that the data can be evaluated again for general or project-specific objectives. The TNI standards are written at a level of detail beyond that found in the ISO standards, providing method-level metrics for this assessment. As Chemists, it is these instrument, method, and data analysis minutiae where we focus our time. To do so in an audit, we compare the quantitative and qualitative method requirements to what is being performed at the laboratory on the bench. In Brazil, the environmental laboratories cite a published method, such as published by the US EPA, Standard Methods (SM), or ASTM. However, the during my auditing in Brazil, it was clear that laboratories were not required to follow the published methods to the letter; it was up to Inmetro and its Auditors to decide what method steps were required to meet their accreditation.
For example, US EPA Method 8260B includes very specific steps for the initial calibration, including the use of a relative response factor and linearity evaluation for all target analytes and a different metric for the calibration check compounds. The method also includes, with minimum response factor, objectives for a set select of system performance check compounds. Brazilian laboratories use linear regression as the default approach, unlike Method 8260B where relative standard deviation is first evaluated. Likewise their calibration check compounds may include reasonable recovery limits to span the full set of analytes, with a subset analyzed after every 10 samples. This use of a check standard within a set of samples is more common for gas chromatography methods and provides added control for evaluation.
For methods that do not explicitly describe common interferences, the laboratories in Brazil may or may not have this information at hand, and may not take precautions to account for interferences. For example, the cyanide methods in SM 4500 CN, cite interferences, but their discussion is intertwined among the various SM method versions. Brazilian laboratories that don’t have an extensive understanding of the method complexities as they relate to interferences may not incorporate the spot-checks for chlorine interference.
The use of performance test (PT) samples is part of the Inmetro environmental laboratory accreditation program. However, it is likely the PT sample analysis frequency at a Brazilian laboratory is significantly less than that in the US. This may be due to the significant cost of reference materials and PT samples in Brazil. Inmetro has a very limited set of Material de Referência Certificado, so oftentimes reference materials/PTs must be purchased from companies outside of Brazil.
It is refreshing to know that, as a general rule within the Brazilian quality system, laboratory Analysts are prevented from having access to client and sample information. Sample containers at the bench have no client or source information, as a means of ensuring impartiality. They do not have any clue that a sample may be a field blank, or duplicate. This reduces the chance for treating samples unequally, and thus, provides a more realistic evaluation of laboratory operations if blind QC samples are used by the client.
It was my experience that even though ISO 17025 is the normative reference for the Brazilian environmental laboratory accreditation, they do go beyond just a “documented quality” approach. However, the rigor of the “acceptable quality” varied with method and laboratory. Ultimately, an Auditor must have the technical background to look at specific operations and calculations to understand the extent of the quality program and controls, along with a top-down perspective to evaluate the Brazilian laboratory and system relative to project or program quality needs. Direct comparison of their operation to the US system, such as TNI, is insufficient and biased.