Health and Medical news
— OPINION —
This is a crime.
We are now beyond six years from the date that the NICD announced the association between Tiger Brands’ polony products and the outbreak, and the work done by all parties has only generated more evidence that the NICD’s and Tiger Brands’ conclusions are accurate. There is no evidence to the contrary.
In 2017 and 2018, the world’s largest and deadliest outbreak of listeriosis occurred in multiple provinces of South Africa (Figure 1). The outbreak was caused by contaminated polony, a ready-to-eat, processed meat product. Ready-to-eat, processed meats are a well-known vehicle for listeriosis outbreaks (Thomas et al., 2020). The Minister of Health declared that there was an outbreak of listeriosis on December 5, 2017, and, on March 4, 2018, further identified Tiger Brands’ polony products as the cause of the outbreak. The Minister of Health instructed Tiger Brands to recall all polony products the same day. See generally, Minister of Health September 3, 2018, Media Statement. The Ministry of Health based its conclusions on the investigative findings of the Joint Public Health Emergency Co-ordinating Committee, which was established for the specific purpose of identifying the cause of the outbreak and developing measures to prevent further illnesses and other outbreaks associated with processed meat products generally. Id. The relevant epidemiologic findings are set forth in the paragraphs that follow.
Figure 1. Incidence of Laboratory-Confirmed Listeriosis Cases during the Outbreak Period, According to South African Districts (Thomas et al., 2020).
Cases were reported from across the country, with most cases reported from Gauteng Province (58%). Women accounted for 55% of total cases. The ages of cases ranged from birth to 93 years. Neonates (aged £ 28 days) were the most affected age group, accounting for 43% of cases. This was followed by adults of 15 to 49 years of age, accounting for 32% of cases. The disease outcome was known for 806/1,060 (76%) of cases; 27% (216/806) had the known outcome “died” (Smith et al., 2019).
Listeriosis is a serious foodborne infection with a case-fatality rate (“death rate”) of 20-30% (Thomas et al., 2020). People primarily affected by listeriosis have impaired cell-mediated immunity. This includes those who are pregnant, elderly, or immunocompromised from conditions such as HIV, chronic disease, or immunosuppressive therapy (Thomas et al., 2020). The specific outbreak strain associated with the outbreak was Lm ST6. There are two ways that listeriosis can manifest: febrile gastroenteritis and invasive listeriosis (Coulombier). Invasive listeriosis is characterized by bacteraemia, meningitis, pneumonia, endocarditis, and sepsis (Smith et al., 2019).
A total of 1,060 cases were reported during the period of January 11, 2017, to July 17, 2018 (Figure 2)[1]. The outbreak period was defined as a duration of time during which case numbers exceeded and remained above a weekly threshold of five cases per epidemiological week (Thomas et al., 2020). At the peak of the outbreak (mid-November 2017), 41 listeriosis cases were reported in a single week. Prior to this outbreak, listeriosis was not a reportable disease in South Africa; therefore, information is not available on the prevalence, epidemiology, and description of clusters/outbreaks on listeriosis. Due to the lack of surveillance data, the baseline number of listeriosis cases was estimated from counts of listeriosis cases in 2016.
It is known that in 2015 and 2016, clusters of listeriosis occurred in South Africa. The 2015 listeriosis cluster involved 7 cases total, and the predominant strain was Lm ST6. However, researchers did not have sufficient epidemiologic evidence to connect the 2015 cluster to any specific food product (Shuping et al., 2015). For the 2016 cluster, retrospective analysis of Lm cases from the years 2012-2016 was used to calculate the expected case numbers for years 2013, 2014, 2015, and 2016 in the Gauteng province (Mathebula et al., 2016)[2]. Because there were only 3 cases in the 2016 cluster, researchers needed to estimate the baseline number of cases.
When determining if a cluster of diseases is classified as an outbreak or epidemic, it is essential to know what the baseline number of illnesses is in the population of interest. An epidemic refers to an increase in the number of cases of a disease, above what is normally expected in that population in that area, and an outbreak is defined the same but is often used for a more limited geographic area (Centers for Disease Control and Prevention [CDC], 2012).
Figure 2. Distribution of Laboratory-Confirmed Cases of Listeriosis, According to Outbreak Week and Major Events (January 1, 2017 to August 21, 2018) (Thomas et al., 2020).
Case definitions are used in outbreak investigations to help identify cases who are associated with the outbreak. A case definition includes criteria such as the subject population, implicated location, time, clinical features, and/or laboratory test results if available (CDC, 2012). The initial case definition for the primary listeriosis outbreak included all cases of listeriosis that occurred in South Africa from 2017 to 2018. The initial case definition was appropriate due to the lack of whole-genome sequencing (WGS) data at the beginning of the investigation. Multilocus sequencing typing (MLST) was used later to analyse the WGS results from all viable isolates obtained from case-patients. Researchers discovered that 93% of the clinical isolates collected from cases during the outbreak period were Lm ST6 (Thomas et al., 2020; Gerner-Smidt). The case definition later included sequence typing information to increase the likelihood of identifying a common source (Besser). This finding also showed that the outbreak strain of Listeriosis was definitively the Lm ST6 strain.
Further, we note that investigators thoroughly analysed the theoretical possibility that Lm ST6 was coming from more than just Tiger Brands’ polony products. In short, after conducting environmental investigations at all 157 ready-to-eat meat production facilities in South Africa, there was no Lm ST6 in any other products or facilities except Tiger Brands.
Together with the NCID’s clear statements that Tiger Brands polony products were the source of the outbreak, based on epidemiologic and environmental evidence, the constellation of all evidence conclusively establishes that Tiger Brands’ polony products were in fact the sole source. There is no additional analysis that will materially change these facts.
Based on its investigation findings, the Minister of Health issued a recall of Tiger Brand’s ready-to-eat meat products produced at the Enterprise facility. The Minister of Health also recalled all ready-to-eat meat products produced at Rainbow Foods, but epidemiologic and environmental findings during the investigation showed that this was a precautionary measure only—i.e., the Listeria identified at the Rainbow Foods’ production facility on environmental testing was not Lm ST6, and therefore had no causal association with illness in the outbreak. See Minister of Health March 4, 2017 Statement. Shortly after Tiger Brands’ recall, the outbreak essentially stopped (Figure 3).
Figure 3. Epidemic curve of laboratory-confirmed listeriosis cases by date of clinical specimen collection (n=1 038) and sequence type (ST) (n=564), South Africa, 01 January 2017 to 5 June 2018 (n=1049)
The environmental and epidemiologic investigative findings establish the likely causal nexus between Tiger Brands’ polony products and most all Listeriosis cases that occurred in South Africa before, during, and after the outbreak period. (Coulombier). As stated above, 93% of clinical isolates that underwent sequencing were shown to be Lm ST6, the strain that was both epidemiologically associated with illness and repeatedly found in the production environment at the Enterprise facility, and nowhere else. But not all isolates could be sequenced, due to the lack of laboratory resources and personnel available (Besser). Based on the high percentage of Lm ST6 clinical isolates, it is highly likely that a similar percentage of non-sequenced isolates would have been Lm ST6 if sequencing could have been done (Coulombier). As further support that there was no difference between the distribution of sequence types among the non-sequenced samples and the distribution of sequence types observed in the sequenced samples, the process of selecting case isolates to be sequenced was not biased. Therefore, it is a statistically valid and provable fact that, in the absence of evidence to the contrary—i.e., sequencing that showed that a clinical isolate was not Lm ST6–a listeriosis patient diagnosed during the outbreak period had a > 90% probability of being related to the outbreak even without confirmed laboratory results (Gerner-Smidt).
After detecting the outbreak, the Centre for Enteric Diseases, a part of the National Institute for Communicable Diseases, conducted a nested case-control study that provided evidence that cases with Lm ST6 infections were more likely to have eaten polony than those with non- Lm ST6 infections (Thomas et. al, 2020). In outbreak investigations, case-control analysis is performed to estimate the odds ratio for the association between specific food items and the outbreak-associated illness. For this nested case-control study, case patients were those with Lm ST6 infections, and control patients were people infected by another strain of LM (i.e., not Lm ST6) during the outbreak period. Results from this study show that the odds ratio was 8.55 with a 95% confidence interval of 1.66 – 43.35. An odds ratio is a measure of association between the odds of becoming ill from consuming a specific food item versus the odds of becoming ill without having consumed the specific food item (Coulombier). An odds ratio of 8.55 signifies that the odds of having eaten polony in Lm ST6 cases is 8.55 times greater than the odds of having eaten polony in non-Lm ST6 cases. Based on the calculated confidence interval, this result is statistically significant because the confidence interval does not include the null value of 1. Therefore, 95% of the time, the true odds ratio fell within this interval.
Prior to the study, food history interviews were conducted to generate a hypothesis as to which food item could have been the source of the outbreak (Coulombier). The food histories were conducted with the use of a standardized questionnaire that inquired as to food consumed by a case-patient over the four weeks prior to onset of symptoms. Open-ended questions were posed to Listeriosis cases to understand each case’s food habits, such as where they purchase food, name of restaurants patronized, and use (and name) of informal food vendors. Closed-ended questions were posed to determine each case’s exposure to specific food items associated with outbreaks in the past and locally consumed foods thought to pose a high risk for listeriosis such as processed meats (e.g., biltong), cold meats (e.g., ham, polony), soft cheeses, raw milk, and raw vegetables. Brand preferences were also captured in the form. The combination of open and closed-ended questions was and continues to be standard practice for the conduct of epidemiologic investigations internationally and provided investigators with high value data for consideration alongside other epidemiologic and environmental information.
The food history interviews were completed by November 1, 2017. The epidemiologic methods utilized throughout this investigation, including the case-control, were robust and mirrored those used in high-level investigations throughout the world. The investigators’ multi-disciplinary investigative methods were fully appropriate for the outbreak circumstances (Besser).
On January 13, 2018, febrile gastroenteritis developed in 10 children from a nursery in Gauteng Province. Several stool samples were collected from the children, and one yielded Lm ST6. Sandwiches prepared and eaten at the nursery were the only common food exposure, and polony was the common ingredient. Polony was recovered from the nursery refrigerator, and Lm ST6 was identified in the polony produced at Tiger Brands Enterprise Facility in Polokwane (Thomas et al., 2020).
On February 2, 2018, an environmental investigation took place at the Tiger Brands Enterprise Facility in Polokwane following the discovery at the nursery (Gerner-Smidt). Of 317 environmental samples taken from the Polokwane facility, 47 tested positive for Listeria monocytogenes, and of the 47 that tested positive, 34 were subtyped as the outbreak Lm ST6 strain. Additionally,
