It is important for the industry to continually monitor the presence of E. coli O157 in the beef supply. The prevalence of E. coli O157 detected in cattle and food products is greatly influenced by the sensitivity of the test used to detect the organism. Generally, two types of tests, rapid and traditional, are used to determine if E. coli O157 is present. Traditional methods result in culturing of the actual pathogen at the end of the process. Rapid methods are usually based on detection of a substance (chemical, antigen, etc.) or DNA to indicate that the pathogen was present in the sample (no viable organism is present at the end of the test). Traditional methods are considered the gold standard but usually require more labor and time to complete. Because of the labor and time restrictions of traditional methods, rapid methods are gaining in popularity. The primary drawback of rapid tests is the false negative and/or false positive results that they may yield. This occurs because background flora may interfere with the rapid test or that they may detect dead organisms.
Recently, rapid ELISA-based methods have gained popularity in detecting E. coli O157 in ground beef samples. While there are several that have been approved for use in ground beef there are some drawbacks to these types of tests. There appears to be a false positive rate that increases as the age of the meat increases. This is likely due to the fact that background flora in the meat that may cause a positive reaction increase during storage.
Despite the recommendation that they not be used on animal samples, some processors are using these ELISA tests to detect E. coli O157 in feces and on hides of incoming cattle, and are using this to monitor the E. coli O157-status of suppliers. Using these rapid tests is likely producing inaccurate and misleading estimates of prevalence. The accepted method to determine prevalence in fecal or hide samples is an immunomagnetic separation method that actually cultures the pathogen from the sample. This method is very sensitive and presumably, is more accurate than ELISA tests because the actual pathogen is grown from the sample.
One processor, using rapid ELISA tests, reported a 100% prevalence of E. coli O157 on inoming beef hides. In order to verify the data, these samples were tested side by side using the IMS method and the ELISA method. To date, 360 samples have been tested using both testing methods. Of these, 45.53% were positive using the ELISA test and only 0.27% were positive using the IMS method. Several bacteria were recovered from the hide samples that resulted in positive ELISA tests, but were determined by DNA sequencing not to be E. coli O157. Thus, the false positive rate on hide samples when using the ELISA test is excessive and produces biased (and misleading) data.
Another processor was conducting an internal verification of the HACCP plan by holding ground beef samples and testing those using rapid ELISA tests. As the samples were held over time, more samples became positive. Samples were sent for outside testing and several species of bacteria were identified that were yielding false positive results. It appears that the ELISA testswere accurate on fresh ground beef samples, but over time those samples became positive because of the background flora in the meat.
Another issue facing the beef industry is the amount of E. coli O157 present when a positive sample is found. The researchers have developed an MPN (most probable number) method to enumerate E. coli O157 in fecal and hide samples. The method is very sensitive, specific, and accurate on inoculated samples. MPN counts were similar to the counts determined by direct plating. Because MPN is a statistical representation, the counts were not exactly the same, but very close. Researchers were able to recover the pathogens from the fecal material in inoculated samples because strains containing antibiotic resistant markers were used. Of course this is not possible when enumeration is needed in naturally infected animals.
Researchers have been successful in enumerating the pathogen in some naturally infected samples, but more samples need to be collected and subjected to enumeration in order to fully validate the protocol. Enumeration of E. coli O157 is very important because a lower load of pathogens going into the plant reduces the amount that ultimately ends up in the food supply.
The objectives of this study were as follows:
The researchers obtained fecal and hide samples from commercial beef harvest plants. A total of 157 fecal (naturally infected) and hide (naturally infected) samples were obtained during May through October, 2005.
Because of the low prevalence of E. coli O157 in ground beef and on carcasses, a low level inoculation of 1 g/10 g of ground beef was used for testing. E. coli O157 originally isolated from cattle was used for the inoculation and a cocktail mixture of 4 separate strains of the pathogen was used. Fresh ground beef was obtained from the Texas Tech Meat Laboratory. Ground beef was inoculated with the pathogen and subjected to mixing in a pilot plant grinder in thepathogen-processing laboratory. Non-inoculated controls were prepared to determine if the background flora contained pathogenic microorganisms and if there were false positive tests from the non-treated samples. For hide, fecal, and ground-beef samples, three separate AOAC approved rapid ELISA test kits were used to determine if E. coli O157 is present in the samples. For each test kit, the protocol detailed by the manufacturer was followed. IMS was also used to isolate the pathogen. When false positives were present, individual isolated colonies were obtained and sent to an outside laboratory for DNA sequencing to determine the identity of the organisms giving the false positives.
All fecal and hide samples that test either IMS- or ELISA-positive for E. coli O157 were subjected to an enumeration protocol to determine the quantity of the pathogen in the sample and the limits of detection for both the ELISA and IMS sampling methods.
Cattle feces samples were assayed using IMS methodology and three commercially available rapid ELISA test kits, VIP, ECOLINOW, and REVEAL. Results of the testing indicated that for fecal samples there was poor (practically no) agreement between the IMS test results and the ELISA results. All of the rapid ELISA tests gave very high false positive results compared to the IMS methods in feces. The results also showed that there was poor agreement between the three rapid ELISA tests in detecting E. coli O157 in fecal samples.
Similar results were observed for hide samples. Although somewhat better, results of the testing indicated that for hide samples there was poor agreement between the IMS test results and the ELISA results. The trend was that the ELISA test kits resulted in high false positive results. The results again showed that there was poor agreement between the three rapid ELISA tests in detecting E. coli O157 in hide samples.
Due to this observation in both feces and hide samples, colonies that gave positive results on the ELISA test kits were isolated and sent for identification using DNA sequencing. Several isolates that were not E. coli O157 gave positive test results on the ELISA kits. Among the isolates that tested positive using this test kit were Enterobacter cowanii, Aeromonas media, Aeromonas hydrophi, Escherichia coli C3, and Aeromonas hydrophila.
Implications
The results of this study indicate that the use of the rapid kits for ground beef is acceptable and has an excellent agreement with the data provided by detecting E. coli O157 using IMS methods (traditional – culture method). However, using these kits for detection of E. coli O157 on hides and in the feces of cattle is not acceptable and has a poor agreement with the data obtained using traditional culture methods (IMS).