Scientific Publications by FDA Staff

J AOAC Int 2016 Mar;99(2):444-50

Matrix extension study: validation of the compact dry CF method for enumeration of total coliform bacteria in selected foods.

Mizuochi S, Nelson M, Baylis C, Green B, Jewell K, Monadjemi F, Chen Y, Salfinger Y, Fernandez MC

The Compact Dry "Nissui" CF method, Performance Tested Method(SM) 110401, was originally certified for enumeration of coliform bacteria by the AOAC Research Institute Performance Tested Methods(SM) program for raw meat products. Compact Dry CF is a ready-to-use dry media sheet, containing a cold-soluble gelling agent, a chromogenic medium, and selective agents, which are rehydrated by adding 1 mL of diluted sample. Coliform bacteria produce blue/blue-green colonies on the Compact Dry CF, allowing a coliform colony count to be determined in the sample after 24 +/- 2 h incubation. A validation study was organized by Campden BRI (formerly Campden and Chorleywood Food Research Association Technology, Ltd), Chipping Campden, United Kingdom, to extend the method's claim to include cooked chicken, fresh bagged prewashed shredded iceberg lettuce, frozen fish, milk powder, and pasteurized 2% milk. Campden BRI collected single-laboratory data for cooked chicken, lettuce, frozen fish, and milk powder, whereas a multilaboratory study was conducted on pasteurized milk. Thirteen laboratories participated in the interlaboratory study. The Compact Dry CF method was compared to ISO 4832:2006 "Microbiology of food and animal feeding stuffs-Horizontal method for the enumeration of coliforms-Colony-count technique," the current version at the time this study was conducted. Each matrix was evaluated at either four or five contamination levels of coliform bacteria (including an uncontaminated level). After logarithmic transformation of counts at each level, the data for pasteurized whole milk were analyzed for sr, sR, RSDr, and RSDR. Regression analysis was also performed and r(2) was reported. Mean difference between methods with 95% confidence interval (CI) was calculated. A log10 range of -0.5 to 0.5 for the CI was used as the acceptance criterion to establish significant statistical difference between methods. In the single-laboratory evaluation (for cooked chicken, lettuce, frozen fish, and milk powder), sr and RSDr values were analyzed and r(2) was reported. Statistical differences were indicated between the Compact Dry CF and ISO 4832 methods in two of five contamination levels tested for lettuce, and in the low contamination levels for cooked chicken, frozen fish, and dry milk powder. For the low levels of cooked chicken, frozen fish, and milk powder, only a few colonies were recovered for each method, and thus not a true indication of the methods' performance. In most cases, mean differences between the Compact Dry CF and ISO 4832 methods were small (<0.5 log10), with CIs within the acceptance criterion. The sr and RSDr values were similar for both methods, and r(2) values were >0.94 for all matrixes. In the multilaboratory study, no statistical differences were indicated between methods. The sr, RSDr, sR, and RSDr values were similar for each method and even slightly smaller for the Compact Dry CF. The r(2) value was 0.99. The Compact Dry CF method offers comparable results to ISO 4832 in a space saving, easy-to-use format.