The Internationa] Mussel Watch 



The need for quality control and intercalibration of analyses for chemical contaminants in 

 environmental samples has been documented numerous times during the past two decades (e.g., 

 UNEP, 1990). Some advantages of inter-comparison exercises include: 



• create a frame of reference so that data from multiple labs can be used in a more 

 comprehensive, regional assessments. 



• introduce and evaluate advanced analytical methods 



• permit self-evaluation by participating laboratories and assist with training new staff 



• impose an external incentive to maintain internal quality control programs 



identify variation between laboratories and common sources of error which lead to this 

 variation. 



A goal of inter-comparison exercises is to reduce the inter-laboratory variation in analytical 

 results. Such exercises are a mutual learning experience and are not a "test" to determine how 

 close any particular analyst comes to the "correct" answer. A step-wise intercalibration exercise 

 should sequentially include: a) analysis of standard solutions, b) check of participants ability to 

 prepare quantitative standards mixtures, c) analysis of cleaned extracts, d) analysis of whole 

 extracts (participant clean-up), and finally e) analysis of environmental samples. 



In addition to analysis by the two central laboratories, Host-Country scientists participating 

 in the Initial Implementation Phase will be encouraged to analyze a subsample of the organisms 

 collected at their site and to participate in an inter-laboratory comparison exercise. Standard 

 Reference Materials (SRMs) and a homogenized bivalve tissue sample will be distributed to 

 participating analytical laboratories. This material should be used to establish the precision of the 

 laboratory by performing an initial analysis of three aliquots of the reference material. Levels of 

 the designated chlorinated hydrocarbons should be tabulated on both a dry weight and a lipid 

 weight basis. Means and standard deviations are then computed. These data will establish the 

 expected precision of replicate analysis for each compound analyzed. The laboratory is requested 

 to submit to the Project Secretariat results of the initial replicate analysis, and the field-collected 

 tissue as well as their quality control charts for inter-comparison with other participating and 

 reference laboratories. 



Quality control charts are constructed by reanalyzing further aliquots of a reference material 



on a regular basis. If a subsequent analysis deviates by more than one standard deviation of the 



expected mean value, then the analyst should reassess the precision of the analytical procedure and 



i 

 correct errors. Plot the results of the repeated analyses of SRMs on a simple chart, which contains 



106 



