ICMSF methods study. XVII. An international comparative study of the direct plate and hydrophobic grid-membrane filter methods for enumeration of Escherichia coli in foods

Abstract

Eight laboratories compared counts of Escherichia coli from naturally or artificially contaminated ground beef, other meats and poultry, vegetables, fish and shellfish, cheese, and diverse sources such as swabs, by the Anderson–Baird-Parker direct plate (DP) and a hydrophobic grid-membrane filter (HGMF) method. For five of the eight laboratories overall counts by HGMF were significantly low (51–83%) compared with those by DP. Counts by HGMF tended to be lower for naturally contaminated samples; several possible causes were investigated. In a subsidiary study, analyst variation in counting HGMF ranged from 0.8–7.3%, with little evidence of effects from counting positive versus negative grid cells or from the fullness of growth or staining intensity.