Collaborative study of the MPN, Anderson-Baird-Parker direct plating, and hydrophobic grid-membrane filter methods for the enumeration of Escherichia coli biotype I in foods.

Five Health Protection Branch laboratories compared two membrane filter methods (the Anderson-Baird-Parker direct plating, and a hydrophobic grid-membrane filter method) against the most probable number procedure (MPN) for enumerating Escherichia coli biotype I in foods. Results were available in 24 h by both membrane filter methods, compared with 10-14 days by the MPN procedure. For ground beef, Parmesan cheese, and cut green beans, the hydrophobic grid method generally gave the highest recovery, although the two membrane filter methods were not significantly different. Both these methods gave significantly higher recoveries than the MPN procedure, and for most foods, either method would be preferable. Further work is required before either membrane filter method can be recommended for bean and alfalfa sprouts, which may contain very high levels of Klebsiella spp.