On January 29, CDC published a very comprehensive report, which details the various facets of the investigation into the peanut butter outbreak. The latest figures show 550 cases of infections in 43 states; the most recent case developed illness on January 17.
I wish CDC had provided a timeline of the outbreak investigation as they did in the spinach-associated E. coli O157:H7 outbreak in 2006. To that end, I tried to construct a timeline of events from the data provided in the MMWR report. I invite informed readers to bring any mistakes and omissions in the chronology to my attention in the comments section.
Click on image for a larger version.
After reading the report, I came up with a list of the following questions and concerns:
- The PulseNet network appears to have detected the outbreak at an early stage: when the case count was 13 and it appeared to be a 12-state cluster on November 10. It is interesting that all patients in the disease cluster had been infected with a previously not seen strain of Salmonella Typhimurium that had a unique DNA “fingerprint.” This is much more meaningful than a cluster of 13 case patients infected with a frequently encountered strain of S. Typhimurium.
- On November 24, PulseNet reported a second cluster of 27 patients who had been infected with a strain of S. Typhimurium that was closely related to the strain from the first cluster.
- By November 25, the number of cases in the first cluster had increased to 35. This is when the CDC OutbreakNet team began an epidemiological investigation with its partners in state and local health departments. My question: Were any patient interviews conducted between November 10 and 25? Were any food exposure questionnaires administered to the cases in the cluster between November 10 and November 25? Could the lapse of time have contributed to problems with accurate recall of food exposures?
- By December 2, the cases in the second cluster had increased from 27 to 41. Additional laboratory characterization of the salmonella isolates from the first and second clusters and epidemiological data from the two clusters indicated that they were part of the same outbreak.
- I am struck by the time it took (six to eight weeks, November 10 to December 28, when King Nut came under suspicion, and January 12, when the company’s involvement was confirmed) to identify the source of the outbreak. I wonder if it was because the peanut butter was distributed in bulk packages to institutions such as nursing homes and schools and was NOT sold in retail outlets. Also, it is likely that the epidemiological investigations were confounded because the contaminated peanut butter was used as an ingredient in many different food products that were sold through retail outlets under several different brand names. In any case, it is apparent that the early warning signal provided by the PulseNet network did not prevent a large buildup of cases associated with the outbreak.
- This is the third peanut butter-associated outbreak of Salmonella infections that has been recognized and investigated. These include the Australian outbreak of 1996 and the U.S. outbreak of 2006. Interestingly, the serotypes of Salmonella associated with all three outbreaks (Salmonella Mbandaka in 1996, Salmonella Tennessee in 2006 and Salmonella Typhimurium currently) have been isolated from the product implicated in today’s epidemic (more than 2,700 serotypes of Salmonella are known. Typhimuirum is a very commonly encountered serotype; Tennessee and Mbandaka are infrequently isolated). In fact, the exact same strain of S. Tennessee that caused the 2006-2007 outbreak was isolated from a product implicated in the current one.
- The low numeric value of DNA “fingerprint” pattern JPXX01.0459 that was isolated from the ill humans and from the implicated product in Connecticut, Michigan and other states indicates that this pattern previously existed in the PulseNet database (pattern numbers are assigned sequentially by CDC PulseNet Database Team in the order in which they are deposited in the database). Since the MMWR report did not elaborate on the significance of JPXX01.0459, I am assuming that this was probably a sporadic case isolate without much clinical/epidemiological information.
Table of salmonella strains associated with outbreak. Click image for larger version (on CDC site).
We will learn a lot more about the outbreak and the investigation of it in the next weeks and months. The finding of multiple, closely related Salmonella Typhimurium strains in ill persons and the implicated product suggests that the parent strain of Salmonella Typhimurium may have been resident in the peanut butter processing facility for some time. If that was the case, why did it not cause illnesses before? What triggered this outbreak? What is the significance of the finding of the salmonella strain from the 2006-2007 outbreak in King Nut peanut butter? Is there a connection between the two outbreaks? I am sure CDC is in hot pursuit of answers to these questions.