Message from the President
Referee appointment
Associate Professor Stephen J Clarke of the Audubon Sugar Institute as the new Referee for Subject 10, Stephen advises that he has many ideas he will be discussing with his Associate Referees, as well as making a start on some collaborative testing which is needed for completing methods for the Methods Book.
We announce also the appointment of Mr Clive Shelton as General Referee Subject G4 Molasses. Clive is a Public Analyst at Central Scientific Laboratories, London. We are especially grateful that Clive is able to undertake the duties of this Refereeship in addition to the Treasurership. He brings to the job much experience as an analyst and also a considerable involvement in writing up analytical methods for the UK Association of Public Analysts.
Methods Book
The main input from Referees concerning the contents of our Methods Book has been received. A list of methods and proposed authors will be circulated shortly.
Polarisation of sugar beets approved
For the first time, the Commission has had one of its Method Recommendations approved by postal ballot. The General Referee for Subject G6 Beet sought the Official adoption of the determination of polarisation in sugar beet using aluminium sulphate clarification. The collaborative test results had previously been reviewed by the Referee for Subject 3 Mrs Godshall assisted by Dr Margaret Nemeth whom delegates will recall attended our Colorado Springs Session. All National Committees which responded voted in favour of adoption of the method.
Laboratory proficiency – Need and assessment
Alan Patey and Roger Wood
(Food Science Laboratory, Ministry of Agriculture, Fisheries and Food, Colney Lane, Norwich, UK, NR4 7UQ
There is much legislation presently being developed within the European Community which has at its core the requirement for accreditation. Industry is becoming increasingly aware of the need for accreditation, generally to the ISO 9,000 Standards (or, in Europe, to the EN 29,000 Standards). Industry laboratories attached to industrial processes that seek accreditation will themselves have to be accredited either as a part of the ISO 9,000 accreditation or as a separate accreditation, normally to Iso Guide 25 (i.e. the EN 45,000 Standards in Europe).
Laboratory quality standards will be prescribed for Community enforcement Laboratories; such standards are likely to be based on accreditation and participation in proficiency schemes.
In addition, it should be appreciated that food-related legislation requires that data be available on any residues and contaminants which might be present in foodstuffs. The effective enforcement of Regulations regarding food composition (e.g. labelling of the carbohydrate content of foods) also needs data to be reliably checked. For the reasons described above it is now being recognised that a lack of independent assessment of the data being produced hampers the work of enforcement authorities and prejudices the recognition of results and certificates which is at the heart of the EC Single Market. It also limits the scope and reliability of food surveillance work on the food supply.
Central to the assessment of data quality is participation in proficiency exercises. This is being recognised by the Codex Committee on Methods of Analysis and Sampling. which is expected to include laboratory proficiency as an agenda item at its next Session. and by the IUPAC/lSO/AOAC Working Group dealing with laboratory quality assurance, and which prepared the lUPAC Protocol on the Harmonisation of Collaborative Studies, which lCUMSA has accepted. That Group has now prepared a draft protocol dealing with laboratory proficiency.
IUPAC/lSO/AOAC have prepared a draft protocol on the harmonisation of proficiency testing schemes. That protocol was first discussed in Geneva this year and will probably be finalised at Delft in May 1992. It will be published, and International Standardising Organisations will then have the opportunity to study and endorse the protocol. Results from proficiency testing schemes which conform to the protocol will be transparent, this being particularly important in view of the increasing need for the mutual recognition of laboratory certificates.
For all the reasons cited above the UK Ministry of Agriculture, Fisheries and Food decided to introduce a proficiency testing scheme for food analysis laboratories. This scheme, known as the Food Analysis Performance Assessment Scheme (FAPAS) helps to provide the independent assessment of laboratory data and conforms to the requirements of the draft IUPAC protocol.
FAPAS design
Assessment is made in eight separate areas of food analysis. Series of samples are prepared to cover each of these areas. The series are described below. For any one series, i.e. to cover a particular area of determination, samples are issued every four months. In that four months participants are allowed the first two months to analyse and report results. In the third/fourth months the FAPAS Secretariat collates and analyses results and distributes a Report. In the fourth month the administrative arrangements for the distribution of the next set of samples for that particular determination are made. Day to day running of the scheme is the responsibility of the FAPAS Secretariat. Preparation of samples is contracted out to commercial laboratories. Participants are able to use the analytical method of their choice.
The statistical analyses employed
Every participating laboratory producing results is given a z-score.
The z-score is calculated from:
z = (x – .rt)/σ
where:
x reported value of analyte concentration in the test material
xt the best estimate of the “true” concentration of the analyte
σ standard error that describes the variation of unbiased values of x, i.e. σ = a + bxt where a is the detection limit of the system and b can be estimated from collaborative trial data.
z-scores can be interpreted as follows:
|z| < 2 Satisfactory Performance
2 < |z| 1 < 3 Questionable Performance
|z| > 3 Unsatisfactory Performance
The 8 Series in operation
These are: Series I -Nutritional Components; Series II -Veterinary Drugs; -Series III -HPLC Procedures; Series IV -Aflatoxins; Series V -QC-Pesticides; Series VII – Trace Elements; Series VIII -GC Procedures and Series IX – OP-Pesticides. 147 Laboratories currently participate in the Scheme.
Sugar analysis
Series III – HPLC Procedures -is now in its second round. In the first Round (February-April 1991) three orange drinks were analysed for a range of sweeteners, a preservative and total sugar content. Participants could use the HPLC procedure of their choice. 44 Laboratories returned results for sugar analysis. Only 41 % of these sets were adjudged to be satisfactory using the z-score system described above. In contrast, 82% of results were satisfactory for the estimation by HPLC of the saccharin levels in the drinks. Sugar analysis by HPLC is known to be an area of analytical difficulty, and this is confirmed by the FAPAS results.
Address of the Editor (changed since the September issue of lCUMSA News): Dr. Robert Pieck, Donystraat 85, B-3300 Tienen, Belgium – Tel.: +32 16 823 096 – Fax: +32 16 820 826 – Tx: 27105