De minimis – a practical step
In a nut shell
I propose that the explicit recognition of a two stage test of de minimis may lead to clarity of fact finding in difficult cases such as marginal noise induced hearing loss, and minor neck sprain. Further, by focusing on the first stage test, the significance of marginal exposures to risk can be resolved objectively.
The conventional two-stage test
The concept of de minimis is fairly straightforward to assess in cases of broken limbs, burns and other ‘impact’ events. Indeed it is so straightforward that the assessor and opposing parties may not make explicit that they are using a two stage test.
In general when assessing injury:
- The first stage is to assess whether the state of the claimant after the event was/is ‘probably different’ to their state immediately prior to the event. Has anything changed? ‘Probably different’, is properly assessed on the balance of probabilities.
- If yes, the second stage is to assess whether the difference is sufficient to merit the term ‘damage’.
The broken leg
While it may often seem unnecessary, medical experts could meaningfully assist with the likelihood that the leg was broken before the impact event, even if no competent medical exam had taken place at that point in time. Typically, evidence of limping, expressions of pain, impaired activities of daily living, the adoption of coping mechanisms and witness accounts etc. would be referred to. More usually, the assertion by the credible claimant that there was no broken leg just prior to the event is sufficient. Their expertise is deemed sufficient because the objective change from unbroken to broken has inevitable consequences which a reasonable person would correctly interpret. That interpretation is based on his experiences and on basic medical knowledge appropriate to his peer group. If the leg had been in a state of complete fracture before the event the claimant would usually have known with a high degree of certainty. Closer examination can be reserved for the more unusual circumstances[1].
The situation would be quite different if the claimant’s peer group had a very high rate of recurrent broken legs, perhaps due to a medical condition. In this case the question ‘is there a probable difference’ would require a more complex approach which includes assessing whether or not the claimant could be expected to know the pre-event state and, to what degree of certainty. The peer group is relevant to the question of pre-existing fracture and how likely it is that the claimant would know. In this situation the claimant on his own is a less certain source of fact.
The point here is that in a stage one test, reference is always made to the relevant peer group and in some cases the reference to peer group must be explicit if it is to be informative.
A more difficult example – noise-induced-hearing-loss
In case of noise-induced-hearing-loss (NIHL) the unmeasured state of the claimant’s hearing prior to negligent exposure to noise (beginning thirty years ago) is not the most pressing question. Instead, the question is whether or not the current hearing is typical of that of the unexposed peer group, matched for age, sex, and other relevant attributes. When there is no sudden change in hearing level evidenced by years of individual claimant audiometry the peer group is the only sensible comparator.
To be consistent with the tradition of the common law, the relevant test is ‘is his hearing ability probably different from that of his peer group’? This is a question which can be precisely answered by expert witnesses. If ‘no’, then there is nothing further to consider. Stage one of the de minimis test has done its job.
In marginal cases it is axiomatic that the proposed degree of hearing loss is close to the uncertainty in the measurement of the individual and to the uncertainty in the predicted value for the relevant peer group. For example, if the predicted hearing threshold[2] was 45 dB(A) ± 5.0 dB(A) and the measured hearing threshold was 48 dB(A) ± 6.0 dB(A) the probability that the there is any difference at all is 8.6%. The claimant is, on the balance of probabilities, no different from his relevant reference peer group and is therefore not injured. Indeed if the uncertainties used here really are typical of the science of audiometry then in this example, the hearing level of the claimant is not ‘probably different from that of his relevant reference peer group’ until 54 dB(A). That is, a measured difference of 9 dB(A). A table of minimum differences can readily be generated but it is just as easy to make the calculation based on the facts of a given case.
It has recently been proposed that for convenience, a fixed deficit of between 3 and 5 dB(A) be adopted for NIHL stage one de minimis assessment. This proposal is inconsistent with the common law and like all such wanderings from the straight path, should be vigorously resisted by the courts. While it may seem attractive to adopt a very simple test, the common law compatible test illustrated above uses the same agreed facts but is specific to the individual case. (it took me less than 1 minute to perform the calculation[3]). An accurate and relevant stage one test of de minimis is available for individual claims without any recourse to an arbitrary test.
A flat rate test is inconsistent with the common law because each relevant peer group has its own associated measurement uncertainties. Is there merit in applying a 3 to 5 dB(A) test of difference when the uncertainty in the peer group figure is 25 dB(A) or even 1 dB(A)? The flat rate method would very often be unjust and therefore be a cause of uncertainty. Far from reducing costs, a flat rate invites detailed appeal.
The common law compatible test is whether or not the claimant’s hearing threshold is probably different from the relevant predicted value. So, in the above example a measured difference of 5 dB(A) has a probability of difference from prediction of just 13%. Accepting a probability of difference of just 13% as proof of deficit would require some other measure of justice than that provided by the common law.
Measurement and prediction are the core business of NIHL experts. By focussing on stage one of the two stage test they will remain within their areas of expertise and will assist the court. Other matters should be left to the decider of fact[4].
Whiplash
Neck pain is remarkably prevalent in the adult population. Yet, for most people it goes unnoticed until asked about. A person just copes with it or ignores it and this becomes habitual. When asked, a self-reported pain level would typically be 1.5 ± 1.0 on a VAS scale for the general adult population. For a new pain to be probably different from the reference peer group the pain score should be at least 3.1 on the day it is assessed by the medico-legal expert. Any lower than this and there is no reason to find that there is a new pain. But pain is subjective. A claimant can assert that there was absolutely no pain before the event, and may even believe it! A medical examiner is no position to assess it unless provided with other information.
A more objective whiplash exam would use a set of peer group comparators including VAS pain scale, range of movement (ROM)(in three axes), psychological state and pain gain tests, each assessed on the probability of being different from the relevant reference peer group. When comparing the results of whiplash research to a small set of tests like this, 30% of people assessed by medics as having whiplash are found to be no different from the reference peer group on the day of the exam[5]. To the usual standards of evidence adopted by the common law they were not injured when examined. If they had been assessed earlier, they might have been objectively assessed as being injured[6]. Asking the examiner to assess whether there had once been an injury in a case where there is no injury at examination is to ask him to stray beyond the bounds of his expertise, and he should politely refuse to answer.
The research data, when subject to a common law compatible stage one assessment, also show that 80% of probably injured claimants return to peer group state within 6 months of the index event. A common law compatible prognosis of more than 6 months would be exceptional. Examinations after 6 months would usually return a finding of absent injury.
Without a stage one test based on the relevant reference peer group the main thrust of a medical exam in minor whiplash is a non-expert self-diagnosis.
Mesothelioma
The risk of developing mesothelioma following inhalation of asbestos fibres has been studied among other things by fibre type, age at first exposure and exposure intensity. It is known that >97% of people in the Western world are exposed to asbestos dust yet fortunately very few succumb to this terrible disease.
The question that is often asked is, at what level of exposure in addition to ‘innocent’ background exposure does the risk of mesothelioma become significantly different from the background risk? The first stage test of de minimis is to ask, ‘is the risk associated with a given exposure (plus background), probably different from the background risk?’
Amosite, which is particularly relevant in the industrial history of the UK, was recently studied[7] in great detail. Applying a common law compatible stage one test to the data from this journal article leads to two observations:
- Compared with background, and for asbestos fibres which are longer than 5 micrometres, the smallest detectable increase in the risk of mesothelioma occurs at a lung fibre burden of between 19,000 and 20,000 fibres per gram of dried lung tissue. The risk detection threshold is equivalent to ~ 1.1 f /ml.years.
- For high exposures, e.g. > 700,000 fibres per gram, the probability that further exposure increased the risk is always below 50%. In other words, once the lungs are exposed to a certain level, further exposure is causally innocent (in statistical terms).
The first point is very different from the ‘one-fibre-is-enough‘ assertion that was popular until recently[8]. It is extremely unlikely that transient exposure to dust from disturbed ceiling tiles would be sufficient[9] to establish a risk that was probably different from background.
Surveys suggest that 42% of men born in the UK around 1945 will have fibre burdens of more than 20,000 fibres per gram of dried lung tissue and that 0.86% of these will develop mesothelioma. If as claimants they successfully proposed a 1.1 f /ml.years threshold test, these men would have little problem establishing material exposure to risk. Those with lung fibre-burdens below this would have to persuade the court of material exposure by some other means.
Whether or not it is public policy to deny marginal risk mesothelioma claims is currently shrouded in the mystery of what exactly is the meaning of de minimis and how to measure it. An explicit stage one test would allow the courts to make their public policy transparent. Is it public policy that defendants whose negligent exposures probably do not add to the measured risk should pay 100% of the damages?
In the UK[10], all negligent asbestos exposures in mesothelioma claims are potentially subject to joint and several liability even those which are too low or too late.
Summary
A two stage test of de minimis is the convention but stage one is often implicit. By making it explicit, and recognising the need to include reference to measurement of the relevant peer group in difficult cases, the opportunity presents itself for an objective stage one test to be applied.
Three controversial results are described here:
- A minimum hearing level effect is readily established in all individual NIHL cases; 9dB (A) would be typical.
- 30% of whiplash claims where the claimant has been medically diagnosed would not find support for that diagnosis if common law compatible testing was applied. In those who are properly found to be whiplash injured, prognosis would rarely be over 6 months.
- Causally innocent asbestos exposures can be objectively defined in case of mesothelioma. The same arguments could be offered for all those situations where risk is measurable.
Expert witnesses can help the court with stage one assessment, and at the same time adhere to their areas of expertise.
An objective and explicit stage one test of de minimis is practicable and may help make public policy more explicit.
[1] For example, a person who was unconscious or drugged before the impact event wouldn’t know if their leg was already broken. The circumstances of the impact event would be relied upon to determine this fact. In another example, an incomplete fracture may be misinterpreted as bruises.
[2] In fact hearing thresholds are usually distributed lognormally so a linear standard deviation would need to be used with expert care until familiarity is established.
[3] The software to do this has been supplied in confidence to all Radar subscribers. It could be made available more widely. In the meantime my fee for 1 minute’s work is less than £10, plus VAT.
[4] A judge can decide for himself if the effect of NIHL was significant, the claimant is available for questioning. There is little to be gained by asking medics to ask the same questions.
[5] Radar subscribers have this analysis.
[6] A common law compatible stage one de minimis assessment at 4 to 6 weeks after the index event might be useful.
[7] C Gilham et al. Occup Environ Med (2015);0:1–10. doi:10.1136/oemed-2015-103074
[8] This sort of statement is a good example of where experts were pressed to stray beyond their area of expertise. There is no way that they could know this to be a fact yet some were happy to state that it was. At best it is an extrapolation aided by faith in an untestable model.
[9] [2009] EWCA Civ 1211 Wilmore v Knowsley Metropolitan
[10] Compensation Act (2006).