Beef Carcass Contamination by Shiga Toxin-Producing Escherichia coli Strains in an Abattoir in Brazil: Characterization and Resistance to Antimicrobial Drugs

A survey was performed to estimate the frequency of Escherichia coli and Shiga toxin-producing E. coli (STEC) in carcasses obtained from an abattoir in Brazil between February 2006 and June 2007. A total of 216 beef carcasses were sampled at three stages of the slaughter process-preevisceration, postevisceration, and postprocessing-during the rain and dry seasons, respectively. Of the carcasses sampled, 58%, were preevisceration E. coli positive, 38% were postevisceration positive, and 32% postprocessing positive. At the postprocessing stage, the isolation of E. coli was twice as high in the rain season. E. coli was isolated from 85 carcasses of which only 3 (1.4%) were positive for stx-encoding genes. No E. coli O157 serogroup isolates were detected. No antimicrobial resistance was found in nine of the isolates (10% of the total). The most frequent resistances were seen against cephalothin (78%), streptomycin (38%), nalidixic acid (36%), and tetracycline (30%). Multidrug resistance (MDR) to three or more antimicrobial agents was determined in 28 (33%) E. coli isolates. The presence of STEC and MDR strains among the isolates in the beef carcasses emphasizes the importance of proper handling to prevent carcass contamination.

Estimation or simulation? That is the question

The issue of smoothing in kriging has been addressed either by estimation or simulation. The solution via estimation calls for postprocessing kriging estimates in order to correct the smoothing effect. Stochastic simulation provides equiprobable images presenting no smoothing and reproducing the covariance model. Consequently, these images reproduce both the sample histogram and the sample semivariogram. However, there is still a problem, which is the lack of local accuracy of simulated images. In this paper, a postprocessing algorithm for correcting the smoothing effect of ordinary kriging estimates is compared with sequential Gaussian simulation realizations. Based on samples drawn from exhaustive data sets, the postprocessing algorithm is shown to be superior to any individual simulation realization yet, at the expense of providing one deterministic estimate of the random function.