Transport CommitteeWritten evidence from Re: Liability (Oxford) Ltd (WL 58)

Re: Liability (Oxford) Ltd is an independent provider of evidence-based information and guidance to UK liability insurers and other liability risk managers. The service covers a wide range of technical issues and has been established for 11 years.

The key medico-legal problem to address is the conversion of honest testimony1 and medical opinion into legal fact concerning diagnosis and prognosis. It is my view that systematic error in this process is largely responsible for the permitting the disruptive effect of whiplash claims on the compensation system. Essentially, medical standards have been used in place of medico-legal standards. Industrial scale exploitation opportunities followed from this.

The consultation addresses diagnosis by asking who should be making the diagnosis, but without yet addressing exactly how and by what rules it should be done. So, this promises “more of the same”, but by different people, or by the same people as now, but with a different name.

Prognosis is the key determinant of general damages once a diagnosis has been made, but this is not mentioned in the consultation.

I propose that diagnosis and prognosis should both be approached from the point of view of legal fact finding and that medical opinion is just a step in that process.

Q. How do you convert honest testimony and medical opinion into legal facts?

The key to this is a familiar-sounding rule:

In the absence of objective evidence of exception, and, but for the negligent act, the condition of the claimant will be assumed to have been, or ought still to be, the same as that of his peer group—ie those not recently exposed to a similar negligent act.

There are those who assert that the above rule is already applied in legal fact finding and I agree that there are examples of this.2

When this rule is applied to research evidence:

Based on data presented in some recent leading scientific reports, my estimate is that 45% of people given a whiplash diagnosis would not be probably different from their peer group. In the absence of exceptional evidence they would not be medico-legally diagnosed. There would be no claim.

More than 70% of injured claimants would have a medico-legal prognosis of below three months.3 So if there is a medico-legal diagnosis the compensation would be less than half the current standard.

Around 16% to 20% of claimants, all else being equal, would have serious long term problems and should be compensated at a higher rate depending on causation4 evidence.

Using the proposed rule:

The “bar would be raised”, simply by employing the basic principles of the law and legal fact-finding.

The consultation objectives would be achieved (by a different route to that proposed).

The reform would be sustainable.

Accurate predictable damages tools could be developed.

The proposal, if finessed and adopted, would have a wider effect on the disruptive potential of other largely subjective injuries such as mild traumatic brain injury.

Question 1: Do you agree that, in future, medical reports for whiplash injury claims should be supplied by independent medical panels, using a standard report form, and should be available equally to claimants, insurers, and (for contested claims) the courts?

I agree that the family GP is not always able to apply his skill purely on the basis of objective evidence in such cases. In this setting, the lack of: objective facts, audit and other feedback opportunities, helps perpetuate a mythology which may not be accurate and is probably not self-correcting.

The lack of objective facts, audit and feedback also allows the reports produced as a result of the MRO system to become systematically distorted by market forces.

Independent medical panels would sustainably solve this distorted mythology problem if: facts were established on the basis of legal principles, a system of audit was employed as part of its make-up and, the feedback from audit was accurate. Inaccurate fact-finding and feedback would create new problems.

A standard report form would only be of benefit if it reported the right details in the right way and interpreted them according to the right system of thought.

So, an answer to Q1 begins with the following conclusions:

Independent panels should increase the probability of objectivity but need to employ accurate legal fact finding and feedback if their findings are not to become just a new problem.

Engaging a recognised authoritative body to set the standards to be used by this panel or these panels would make it more difficult to challenge the official interpretations. But what if they get it wrong? Internationally, the track record of such bodies is not good.

Standardisation of forms is a matter of efficiency not policy; the policy issue is whether or not the contents and interpretations are accurate. Doing the wrong thing more efficiently is just “wronger”.5

Transparency is a policy issue; it is hard to see any justification for secrecy.

Overall: independence, standardisation and transparency are nice things, but there is no doubt that each of the above could be circumvented by an imaginative, resourceful, profit-making industry.


The main unsolved problem is the current use of medical decision-making in place of common law decision-making. I propose that the medico-legal problem to address now is the conversion of honest testimony and medical opinion into legal fact. Here’s how…

In place of a medical approach, the following proposed rule is consistent with the workings of the common law:

In the absence of objective evidence of exception, and, but for the negligent act, the condition of the claimant will be assumed to have been, or ought still to be, the same as that of his peer group—ie those not recently exposed to a similar negligent act.

Normally, claimants will not have objective evidence of exception. The proposed rule is effectively a preferred choice of baseline fact.6

The choice of baseline acts principally through the medium of diagnosis but through this: causation, harm assessment, foreseeable benefit from rehabilitation and, prognosis are also modified.

(1)Diagnosis is the art of identifying difference. In the absence of evidence of exception, the diagnosis is made by comparison to the appropriate peer group. If a person is not probably different from that peer group, then he is probably not injured. To diagnose in such a case would require exceptional justification.

(2)If injured, then the degree of harm is measured relative to the state of the peer group. Again subject to exception.

(3)The foreseeable benefit of rehabilitation eg physiotherapy, is in returning the injured person more quickly to or with greater probability to the population-normal.

(4)Prognosis will be a prediction of when he will return to the norm for his peer group.

The mechanics of the problem

It can be shown that in cases where diagnosis involves the assessment of common subjective symptoms and common signs, medical diagnostic thresholds are by design, well below those that would satisfy a balance of probabilities test.


Medical research, which rightly informs medical practice, is based on the maximisation of the area of the ROC curve.7 The effect of this for whiplash injury, a high proportion of which is not all that different from the normal peer group, is that medically preferred diagnostic thresholds are well below those that would satisfy a test based on the balance of probabilities.8

When the injury is extreme, this threshold shift effect is unimportant (a broken leg is accurately diagnosed under both schemes), but when the injury is not much different from the normal presentation, this fundamental incompatibility of approaches has significant effects.

It can be shown mathematically that in this latter case, the probability of difference is ~ 25%, when the ROC curve area is maximised. The medical approach diagnoses as injured, people who are not probably different from normal.

Whiplash research, which is based on a medical approach, will also systematically describe people as injured who are in reality not probably-injured.

This has had a strong effect on causation, and prognosis research.


When viewed under the lens of the proposed rule,9 much of what is now regarded as self-evident becomes non-factual. For instance, there is probably no relationship between: head restraint position, head position, delta v, direction of impact and, injury frequency. Those who have been educated with a conventional physiological model of whiplash injury mechanism will be surprised by this, but these findings are in my opinion compatible with the proposed rule when combined with a balance of probabilities test.


When viewed under the lens of the proposed rule, those who are correctly diagnosed mostly return to population-normal by three months and, but for a special subgroup, prognosis beyond six months is extremely unlikely.

The special sub-group is remarkable in one key respect—pain hypersensitivity. This can be objectively measured—but simpler tests of it are very hard to fake. This group has a poor prognosis and comprises roughly 16% of people who experience a whiplash event. Causation is still mysterious and some people will have been pain hypertensive before the accident. Physiotherapy provides no benefits. Compensation should reflect their plight and true cause.

Curiously, the foreseeability of benefit from physiotherapy has not been altered by looking at it from the proposed rule’s “point of view”. It is perfectly clear, even with medical standards, that physiotherapy does not have a foreseeable effect on indemnity. It simply doesn’t do any measurable good, no matter how you measure it.10 Why it is paid for under an indemnity policy is a result of precedent set when injury and treatment were much more clear-cut. Lack of foreseeable contribution to indemnity should give rise to a re-evaluation of its provision under an indemnity policy.


Medical opinion, if informed by medical research will not automatically be compatible with common law decision-making. If the proposed rule is accurate then the broad body of research could be re-analysed from a rule-compatible point of view and examiners re-educated in the light of these findings.

As a pilot project I have made this re-analysis to a very limited extent. I now believe it is worth doing systematically, but would suggest this be done/overseen by an expert body.

Does this meet the aims of the consultation?

Given the preliminary results of viewing the science from a proposed-rule point of view, it seems evident that the needs for independent diagnosis, prognosis and audit can be met. In the process, the number and cost of whiplash claims would be reduced.

Medico-legal diagnosis protocol

In a representative peer group, there is pain, limited neck movement, disability related to neck problems, soreness, dizziness, headache and other symptoms/signs. These are perfectly normal and usually unexplained. Often they are unnoticed until asked about. In the absence of objective evidence of exception the claimant must be assumed to have been and ought still to be roughly the same as his peer group. That baseline is the fact, unless there is objective evidence to the contrary.

Each of the above symptoms/signs has been measured in representative normal populations.11 Average presentation and expected range of presentation are now well known. Measurement precision is well known. It is therefore possible to define diagnostic thresholds which meet the test of “probably different from normal”.

These thresholds can be used as the basis for medico-legal whiplash diagnosis.

At the moment, the neck disability index is probably the best tool for this purpose. It is widely used in whiplash research and can be directly related to prognosis using the data collected by researchers.

A rule-compatible medico-legal protocol can be prepared. It would include ROM, pain palpation, questionnaires, pain hypersensitivity tests etc but unlike now, very clear legal diagnostic thresholds will be provided for the examiner to help him form a judgment. There would also be accurate guidance on those causation issues that fall within the competence of a medical examiner.

Use of the protocol could be a pre-requisite for submitting claims. Medico-legal panel membership could be conditional on demonstrating accurate use of the protocol.


Having re-evaluated a small sample of the medical literature:

In 70% of probably injured cases the prognosis should be less than three months from the date of the accident.

There would need to be a good reason for providing a longer prognosis, these reasons can be obtained from the medical literature (once adjusted to be compatible with the proposed rule).

16% to 20% would have evidence of cold pain hypersensitivity and would have a prognosis longer than six months.

There would have to be an exceptional reason for recommending physiotherapy. If this was provided for more than 5% of cases this would need to be explained.

Deviation from these and other expectations would trigger a closer examination of the work of that examiner or panel and could be explained by any unusual case selection effects. There is no need to routinely re-examine closed cases or to peer review existing cases. This in part answer to question 4, concerning peer review.

Further work

The above figures and conclusions are based on a preliminary analysis of recent papers which provided usable data.12 There are many more research projects where the data was obtained but not usefully presented in the final publication. It could be obtained.

The proposal is that an expert body obtain this data and use it, from a proposed-rule point of view, to:

Define legally objective diagnostic thresholds [and their acceptable tolerances].

Define legally objective audit criteria [and their acceptable tolerances].

Develop a medico-legal examination protocol and standard reporting form.

Develop a system to be used for audit.

Where is the evidence?

None of the detailed calculations, case law, mathematical methods, publication references etc, are included in this consultation response. The aim here is to illustrate a usable principle and outline its effects. If there is an appetite for pursuing this line of reasoning then the consultation process and subtending works can be adapted to that end.13

Question 4: Do you consider that an element of peer review should be built into every assessment, or only for a sample of assessments for audit purposes?

Medical reporting organisations, whether they are called independent medical panels or have some other name, should collect statistics on:

the number of cases seen,

the number diagnosed as injured,

the prognosis period offered in those diagnosed,

recommendations for physiotherapy,

the number of cold pain hypersensitive cases identified and the prognosis attached,


Ratios should be in line with audit criteria based on the right reading of scientific research. The MRO/panel will then be responsible for any system-wide deviation.

Peer review of individual case notes could from part of the justification of variance if any is found.

Re-examination of claimants would be a possibility, paid for by the MRO.

System-level peer review would seem to be the best option. The MRO/medical panel should administer data collection and QA.

A separate levy should be paid to an organisation set up to monitor the independent medical panels, review the audit criteria in the light of new research, review the ML exam protocol and report form, and present reports to the enforcement authority.

If there is a licensed medical panel system then there is no good argument for accepting the occasional report made by the claimant’s own GP. It follows that there is no need to set up a separate system of peer review for this offering.


The consultation should develop further to include the exploration of a potential incompatibility between medical decision-making and legal decision-making. This may lead to a more appropriate medico-legal examination, standard report and, enforceable audit.

May 2013

1 In my opinion, fraud detection is not within the competence of a medical examiner. Other parts of the system should deal with this issue.

2 Eg, In noise-induced hearing loss the effect of negligence is measured relative to the effect of natural presbyacusis—which is measured in the relevant peer group not negligently exposed to excessive noise.

3 This is well below the conventional medical prognosis period (over eight months on average).

4 Largely determined using medical notes. There are no crash dynamics that explain chronic whiplash. Roughly 5% to 10% of people have a history of unexplained neck pain problems.

5 To partly quote Russ Ackoff.

6 One of the attractions of adopting the proposed rule is that there is no need to doubt the veracity of the claimant’s statement of prior perfect health. It can be regarded as an honest statement, but should not be regarded as a legal fact unless there is good reason to do so. The burden is on the claimant.

7 The greater the area the fewer false positives and false negatives.

8 This can be shown mathematically.

9 In the absence of objective evidence of exception, and, but for the negligent act, the condition of the claimant must be assumed to have been, or ought still to be, the same as that of his peer group—ie those not recently exposed to a similar negligent act.

10 ie, both medically and, according to the balance of probabilities.

11 That is, people who have not recently experienced a no-fault whiplash accident.

12 Papers usually focus on the results of complex analysis but without providing the raw data in a form which can be used to guide legal fact finding.

13 It might be useful to meet with DoH policy leaders to discuss this idea and proposal before the conclusions to the consultation are set in stone.

Prepared 29th July 2013