Processed meat and colorectal cancer claims –
Yes, No, How Big?
Summary
It has been proposed that processed meat consumption causes colorectal cancer. Since there have been no UK insurance claims for this, and the proposal has not been put to proof at common law, it is described as a potential liability ENID[1], or, emerging liability risk. The background is briefly described below.
As part of the necessary[2] evaluation response, a software model of the potential UK personal injury liability exposure has been developed. The model allows the user to test various technical and legal “what ifs” and to perform sensitivity analyses. A generic causation assessment tool is included in the software and is based on UK legal precedents.
The normal medical response to colorectal cancer includes surgery. Such surgery entails a risk of serious infection and the consequential need for antibiotics. Given the increased rates of antimicrobial resistance (AMR) some bowel surgery cases will die as a result of infection. The processed meat model allows for the AMR effect.
ENID assessment – two classes of questions
For liability insurers the usual questions about emerging risks (ENIDs) come in two general classes. These are 1) the “binary” questions which express the logic of exposure e.g. is there a duty of care, do we cover that risk etc., and 2) the “proportionate” questions which explore the potential scale. If the binary questions lead to the view that the risk is covered, the proportionate risk management response is guided by the assessment of scale.
Processed meat as a cause of colorectal cancer?
If yes or likely to be yes, there would be a potential liability exposure for food manufacturing, preservative manufacturers[3] and food retail and their professional advisers. Strict liability applies to harm arising from the normal use of food products. Product liability, PI and general liability policies are in the frame.
Is processed meat carcinogenic?
In 2015 the International Agency for Research on Cancer (IARC) classified[4] processed meat as carcinogenic to humans (Group 1). The strongest association was with colorectal cancer, but breast cancer was also discussed.
Processed meat was described as:
Meat that has been transformed through salting, curing, fermentation, smoking, or other processes to enhance flavour or improve preservation. Examples of processed meat include hot dogs (frankfurters), ham, sausages, corned beef, and biltong or beef jerky as well as canned meat and meat-based preparations and sauces.
The justification was presented in Monograph 114 – which has recently been released[5] into the public domain.
Of particular concern are nitrite additives (nitrate additives spontaneously convert into nitrites in situ) and the reaction between nitrites and naturally occurring amines in the food. The amine reaction products include well-known carcinogens e.g. N-nitrosodimethylamine (NDMA) (CAS number: 62-75-9). NDMA is at its highest levels in fried bacon, frankfurters, liver sausage and salami. The production of carcinogens in this way has been known since the 1960s. IARC did not say that NDMA production was the actual cause of the association with colorectal and went to some lengths to include processed meats which do not contain NDMA in their analysis. For the court to determine: does the inclusion of known carcinogens cause the product to be defective, should there have been warning labels on processed meat products?
Yes/No obstacles to successful common law claims
Public interest
Nitrites and nitrates are deliberately added, with regulatory approval. The stated aim is to help control the microbiological quality of the food product and to improve the colour. Maximum and minimum levels are decided by the European Food Safety Authority in Europe and the Food and Drug Administration in the USA for example. Those same approved levels are known to generate carcinogens.
The courts might question the continued inclusion of carcinogens in processed meat given that microbiological safety, storage time and artificial colour can be maintained in other ways.
However, retrospective assessment of decisions made by a public authority may be ruled out. Properly appointed politicians must surely have the freedom to regulate and to adopt whatever evidence suits the perceived political purposes of the day. If creating carcinogens in meat was the preferred choice at a given time then this decision is perhaps by definition in the public interest.
Generic causation
Would a common law judge agree with IARC? Food risk science is notoriously poorly designed, poorly executed and poorly interpreted. Studies of risk associated with processed meat consumption are no exception[6]. Different studies show or might show different things. A weight of evidence approach is needed[7].
Given the lack of 100% agreement between studies a persuasive case on causation would require a great deal of education of the judge, sufficient for him to form his own view of the science. If the judge is not provided with these skills he must[8] find that the case has not been made.
Having read the evidence in Monograph 114, it is my view that the epidemiology case in favour of generic causation is slightly better than the case against, and has been so since 2010. This could turn out to be the date of knowledge.
Laboratory evidence of potentially relevant steps in the disease process is plentiful for both red meat and processed meat but such sensitive evidence is not always informative of specific fact, ‘could be’ is not the same as ‘is’. Perhaps more significant is the clear evidence of excess pre-neoplastic lesions found in rats after a lengthy period of regularly eating processed meat. Such an excess is a common sign of carcinogenicity.
In my view, a common law judge, armed with the data in Monograph 114 and the thorough education presented by both parties could reasonably decide either way on generic causation. The cost and risk to the claimant of making the argument might therefore appear to be daunting.
Since 2015 more studies have appeared. By a wide margin, the best of these so far is a study[9] set in Germany. In my view, this study tips the balance in favour of generic causation for processed meat[10].
Specific causation
What you eat, is a lifestyle choice. Perhaps the hardest element of a lifestyle choices liability claim is to make out specific causation. Consumers usually have a broad range of sources of exposure to processed meat. Which source should be given the role of defendant? How many consumers have records that would validate any claims of levels of exposure? How likely is it that a consumer was exposed to only one product, every day for a lifetime? Etc.
For non-specific cancer, in the UK courts, the claimant has to show a doubled risk if he is to make out specific causation[11]. Experts will undoubtedly differ as to what constitutes a doubled risk consumption level. The key question though is would the courts always use the doubled risk test to identify the right defendant? It would seem that the answer is now “NO”. Thanks to the Heneghan case[12] the claimant need not identify a doubled risk from any particular defendant. All he needs to show is that there was overall a doubled risk and that this or that defendant was responsible for 10% or 43 % of the risk. Each would pay in proportion, unless perhaps one defendant was shown to have been responsible for more than 50%, in which case they might pay 100% of the indemnity. This remains unclear.
£ Liability ENID model for the UK
A “processed meat causes colorectal cancer” personal injury liability ENID exposure model has been developed for the UK. Users will be able to decide if the problem is big enough to warrant answers to the binary questions and then if needs be, to estimate proportional responses.
There is also increasing interest in a model of breast cancer[13] risk associated with processed meat consumption.
Findings
The processed meat industry in the UK has an annual turnover of £17 × 109 and a value added of £2.9 × 109. Neither of these figures is exceeded by the mean annual loss or the “1 in 200 year” annual loss generated by the model. A date of knowledge of 2010 however would have a cumulative effect and this would be more challenging.
If the answers to the binary questions include the possibility of exposure, insurers would have a guide as to a proportionate response. Explicit cover would be an option.
Very similar models for the EU and for the USA are under development. Jurisdictional effects on the standards of the common law will be key.
As with all product liability scenarios the next problem is to find out how such a loss would impact a given insurance portfolio. Mapping the risk onto the insurance portfolio must be fact based, transparent and consistent with the workings of the common law and insurance terms. For some insurers such a mapping could provide a binary step in exposure assessment. Wording adjustment could have the same effect going forwards.
What about AMR?
Colorectal cancer surgery presents a very high risk of infection and the consequential need for antibiotics. Given the increased rates of antimicrobial resistance (AMR) some bowel surgery cases will die as a result of infection[14]. AMR would lead to a higher death rate among those with a potential processed meat-related claim.
The ENID model finds that AMR would increase the number of deaths attributable to processed meat consumption by 510 (± 190) cases a year in the UK.
andrew@reliabilityoxford.co.uk
[1] Liability emerging risks are often referred to as liability events not in data (ENIDs). This term evolved from actuarial concerns that there is no retrospective data with which to model the potential exposure. It is often interchanged with liability IBNR.
[2] Under Solvency II, Liability insurers are expected to identify and evaluate liability ENIDs.
[3] The Bacardi case suggests that preservatives manufacturers would not be liable. They would not be regarded as having created a defect. Professional advisers may fare differently however. Bacardi-Martini Beverages Limited v Thomas Hardy Packaging (2002) EWCA CIV 549.
[4] https://www.iarc.fr/en/media-centre/pr/2015/pdfs/pr240_E.pdf
[5] http://publications.iarc.fr/_publications/media/download/4810/597702b3422b689b361779aac82862ef811ddd3c.pdf . Released on the 2nd Feb 2018. ISBN 978-92-832-0180-9
[6] For example, food risk studies which fail to distinguish between processed meat, red meat and non-red meat are still being accepted for publication in the food science literature. Precise interpretation is excluded by design.
[7] Weight of evidence cannot be provided by numerical meta-analysis when study designs are heterogeneous in significant ways. The cause of heterogeneity must first be identified and its significance for the causal proposal must be assessed. In some cases the original study can be re-analysed, but often the study should simply be excluded from further consideration.
[8] See for example: Outer House, Court of Session [2005] CSOH69 M McTear against Imperial Tobacco Limited (ITL). The judge rightly decided that the claimant had not educated him well enough to form his own view of the evidence. He had no choice but to find that causation had not been established. At issue: does tobacco smoking cause lung cancer?
[9] Eur J Epidemiol (2017) 32:409–418. PR Carr et al. Associations of red and processed meat intake with major molecular pathological features of colorectal cancer
[10] The attributable risk from red meat was one 70th of the attributable risk from processed meat.
[11] In the USA it may be sufficient to show that reasonable warnings were not given. While this is not a logical proof of causation it can substitute for causal logic in the minds of the jury.
[12] [2014] EWHC 4190 QB Heneghan v Manchester Dry Docks etc. This precedent may enable a number of hereto unsuspected liability ENIDs and, cause a great deal of confusion in current claims.
[13] See for example: European Journal of Cancer (2018) Vol.90 p 73-82. JJ Anderson et al. Red and processed meat consumption and breast cancer: UK Biobank cohort study and meta-analysis.
[14] The effect of AMR on clinical negligence exposure for colorectal cancer surgery can also be modelled.
