SA’s listeria outbreak and the Swiss Cheese model of disaster
“Having spent the last 20 years of my life invested in food safety management systems in the South African food industry, I have to say the Listeriosis outbreak has caused some self-doubt. What have we done wrong? What have we missed?” These are some very big questions asked by Linda Jackson of compliance website and consultancy, FoodFocus.
Although nothing we can do now will change the tragic outcome of this outbreak, it is imperative that in the root cause analysis of the incident, collectively as an industry, we apply the learnings.
It is my opinion that multiple failures have led to this incident. In addition to the company in question, other authors have commented on the potential role of the supply chain, the regulators, the certification bodies and even the auditors.
Given the prominence of the brand, it is likely that many of us have been involved in some capacity over the years. Perhaps we have all contributed in some way to this incident, too. It is clear from the update below from the NICD that we still face a number of challenges. LmST 6 does seem to be the tip of a very very big iceberg:
“Whole-genome sequencing analysis has been performed on 521 clinical isolates to date. Of these, 85% (443/521) were identified as sequence type (ST) 6. The remaining isolates (15%, 83/521) represented 19 sequence types including, ST1, ST54, ST876, ST2, ST5, ST204, ST219. ST224, ST71, ST101, ST121, ST155, ST3, ST403, ST515, ST7, ST8 and ST88.
Whole genome sequencing has been performed on 595 food and environmental isolates. Of these, 13% (79/595) were identified as ST6. The remaining isolates (87%, 516/595) represented 26 sequence types, including ST20, ST1, ST121, ST5. ST321, ST9, ST155, ST2, ST3, ST87, ST120, ST378, ST101, ST108, ST2288, ST31, ST7, ST11, ST122, ST14, ST37, ST4, ST54, ST76 and ST88.”
What controls should be in place?
Although we cannot generalise in all sectors of the food industry, as a manufacturer you should have the following levels of protection in place:
1. Minimum compliance
A valid certificate of acceptability issued by the local municipality in terms of Regulation 962 of the Foodstuffs, Cosmetics & Disinfectants Act.
Its general hygiene requirement is as low as you can go. This certificate should ensure your facility is adequately designed and constructed to handle food.
The waste systems should conform to municipal requirements and the National Building Regulations.
This regulation does place emphasis on the training and behaviour of food handlers – often a weak link in any food safety chain. It also places the full legal liability on the person in charge.
There are many other regulations under the FCDA that relate to the composition of your product.
There is the letter of the law and then there is the spirit of the law. Although the approach of our National Department of Health is reactive, the intention is that you, as a responsible manufacturer, should be proactive.
2. Pre-requisite programmes
Call them PRPs, GMPs or whatever you like, but make sure you have the basics in place.
The focus of many retail and hygiene audits are the basic building blocks of cleaning and sanitation, preventive maintenance, supplier controls, storage and preservation of your product and personnel hygiene practices.
The new draft regulation R364 which will soon replace R962, will include more of these requirements as a legal foundation.
And with these requirements comes the need for documentation to defend your systems and the effective daily implementation of the right practices.
Once the basics are covered, we should be engaging in a formal risk assessment of the product and processes.
What is it about YOUR product and process that could go wrong and in so doing harm the consumer? What makes ready-to-eat products such a high risk in comparison to handling ingredients like dry rice?
What are YOUR specific hazards in the process and how can you be sure you are controlling these? And control means you can reduce, eliminate or prevent them – not manage them as best you can!
We have seen this as a voluntary requirement in most sectors of the food chain. It’s time to revisit that thinking.
Ensuring you as a manufacturer fully understand and control hazards to consumers health is not a nice to have – it’s your legal and moral obligation.
Along with this goes the processes of validating and verifying those processes – can you trust them, and can you prove they work?
Taking the next step to comply with customer requirements can be seen as a grudge purchase. Surviving the plethora of audits has not been seen to add value in many companies. Having all the right certificates does not guarantee your systems, as we have seen.
Ensuring robust internal audits are in place that fully interrogate your food safety activities should be the focus. Competent external auditors can then verify implementation, but you should be validating your science.
Relying on external auditors during a brief announced visit will not be an adequate assessment of your system’s health.
5. The next step
If you have it all in place, where do you go to from here?
The next step is open and transparent sharing of information in the food chain to ensure we do have food safety from farm-to-fork.
Confidentiality agreements have hampered the progress in identifying the source of the outbreak. Is it time to revisit this thinking?
Do we need a forum where we can share results without fear of litigation in order to improve our response times to the next outbreak? We need to build a bridge with the regulators not higher walls in order to regain consumer confidence.
But we have all that in place? What went wrong?
Given that we have so many layers of protection in place, how could an outbreak of this magnitude occur? Given what is in place, it would seem that company, auditors, labs and retailers all missed something.
Seems we may have to consult the rocket scientists on this one.
The Swiss cheese model of accident causation is a model used in risk analysis and risk management, including aviation safety, engineering and healthcare. It is often used as the principle behind layered security such as cybersecurity systems.
In the Swiss cheese model, an organisation’s defenses against failure are modelled as a series of barriers, represented as slices of cheese.
As we have seen in food safety, we rely on a series of barriers previously discussed. The holes in the slices represent weaknesses in individual parts of the system and are continually varying in size and position across the slices.
When an incident occurs, it is a result of a gap or failure in a barrier. A gap or a failure produces a weakness in our food safety management system.
If we do not address this gap, with time it can, along with other gaps, cause a crisis. This would be when a hole in each slice (barrier) momentarily aligns, permitting (in Reason’s words) “a trajectory of accident opportunity”, so that a hazard passes through holes in all of the slices, leading to a catastrophic failure.
Each failure on its own would have been minor, but the cumulative act effect can have far reaching consequences.
The model includes both active and latent failures. Active failures encompass the unsafe acts that can be directly linked to an incident. Latent failures include contributory factors that may lie dormant for days, weeks, or months until they contribute to the accident.
Latent failures can be organisational influences, supervision, preconditions, and specific acts or omissions.
In the application of HACCP and FSSC 22000, we are required to analyse our hazards and the consequences. Do we do this in relation to the series of controls we apply and the potential simultaneous failure of one of more of these control measures? Is it time we also reviewed the effectiveness of organizational influences and supervision in the same way as we calibrate temperature probes?
Do we truly understand our processes and are we reviewing the hazards and how they can change and adapt with the right level of scrutiny?
I would respectfully suggest that we need to take a good look at our cheese.