Ochratoxin A in wheat: a second intercomparison of procedures

The European Commission, Measurements and Testing Programme (BCR) has undertaken a project to improve methodology and to prepare certified reference materials for ochratoxin A determination. The first phase of this project, an intercomparison of procedures for the determination of ochratoxin A in wheat, at a content of approximately 13 micrograms/kg, has already been reported. The second intercomparison study, described in this paper, involved 26 European laboratories, from 11 countries, which analysed wheat naturally contaminated at a level of approximately 7 micrograms/kg, and a 'blank' wheat sample (ochratoxin A content < 0.2 microgram/kg). The participants used a variety of procedures which involved different extraction solvents and clean-up procedures. All laboratories used HPLC as the determinative step. Some laboratories also used immunoaffinity column clean-up in comparison with their normal method. Recoveries of the normal methods used by laboratories ranged from 58 to 114%; only three laboratories obtained recoveries outside the accepted range of 70 to 110%. Recoveries of the immunoaffinity column methods, using two sources of column, ranged from 58 to 114% for one and from 4 to 86% for the other. The between-laboratory reproducibility coefficient of variation for all results was 34% for the normal methods, and 34 and 42% for the two types of immunoaffinity columns. It was noted that, after the results were corrected for spike recovery, some laboratories became outliers owing to low spike recoveries. Further investigations of the spiking protocols used by each laboratory showed that the time left for evaporation of the spiking solvent was crucial to the recovery obtained.