A comparison of EF-18 agar and modified brilliant green agar with lutensit for isolation of Salmonella from poultry samples

Two selective plating media were compared to determine their performance in a three-step protocol for isolation of Salmonella from faecal, litter, and fluff samples. The encichment protocol consisted of: 1) Preenrichment in buffered peptone water (BPW), 2) Selective enrichment in Rappaport-Vassiliadis broth, and 3) Plating onto EF-18 agar and BGA/L simultaneously. From a total of 1101 samples, Salmonella was isolated from 158, 157 of which were faecal samples. Thirtyone of these isolates were recovered on one medium only, 18 could not be found on BGA/L and 13 could not be found on EF-18 agar. The relative specificity and sensitivity of each plating agar was determined by enumeration of false-positive and false-negative reactions. EF-18 agar compared favourably with BGA/L, displaying a sensitivity of 0.92 as opposed to 0.89 for BGA/L, calculated for the ''fecal samples'' group only. The calculated specificities for each group of samples were likewise considerably higher for EF-18 agar (0.75-0.91) than for BGA/L (0.35-0.55). Though EF-18 agar is slightly more expensive than BGA/L, the routine use of the former may result in a considerable reduction in overall laboratory costs due to its superior selectivity On the other hand the combination of the 2 media clearly would reduce the number of false negative results, with little extra cost.