ABSTRACT
In preliminary studies, several commercial polyvalent fluorescent antibody (FA) preparations were evaluated for specificity and crossreactivity and an FA method was developed for the screening of Salmonella in products. Approximately 4000 product samples were tested by the FA method and the results were compared to those from the official final action AOAC method, 46.013-46.026. Only 4 FA false-negatives were found for a total of 619 confirmed positive Salmonella samples. The FA false-positive rate was 7%. The method was then subjected to a 2-phase collaborative study. In Phase I, 22 analysts tested 5 inoculated and 5 uninoculated samples of dried milk. In Phase II, 5 naturally contaminated and 5 presumably uncontaminated foods were analyzed. The study was designed to compare results from 11 analysts experienced in FA methodology with those from 11 analysts with little or no experience. Selenite cystine (SC) and tetrathionate (TT) broths were used for enrichment and both were inoculated into SC for post-enrichment. All 4 combinations (SC, TT, SC-SC, and TT-SC) were used with the FA method to determine the best technique. Results were compared to the analysis with TT and SC by the AOAC culture method. In all studies, FA analysis with SC-SC gave the highest correlation with the AOAC method. In a total of 200 samples, the experienced group found 125 AOAC positives and 127 FA positives; no FA false-negatives and only 2 false-positives were reported. The inexperienced group reported 9 FA false-negatives and 5 FA false-positives. All false-negatives occurred in only 3 of the inexperienced laboratories. These studies showed that enrichment and post-enrichment in SC gave the best FA results and that training in FA methodology is required for correlation with existing AOAC methodology. The FA method for the detection of Salmonella has been adopted as official first action.
