The Bellman's Fallacy and other biases

By Dr John Wilson

For about 15 years now, the workers’ compensation lawyers in my town have been running an annual whole-day seminar called “Both Sides of the Fence”, where the plaintiff and defendant sides mingle both on the podium and on the floor.  Invitations are always sent out to the occupational physicians, who generally manage to show two or three.  The high point of the day is the lunch, which is well lubricated by some very respectable wines.  Every year, the after-lunch session is a hypothetical case study, that is deliberately structured to keep the audience awake.

A few years ago this hypothetical surrounded a case where the injured worker had been medically treated by a doctor whom, in the eyes of the worker’s union, could have managed the case differently. The facilitator, turned to the panel of four experts assembled on the podium, singled out the one who was from the union, and asked him how he would handle the situation.  “I’d refer the worker for an independent medical examination”; “righto!”, declared the facilitator.” And how would you choose the examiner?”  There was a pause, followed by a longer pause. “I’d refer the worker to a doctor, with whom I’d be confident that I’d get the opinion I needed.”

“Gotchaa!”

For many years I had grappled with the observation that in the circus of the medico-legal industry we were seeing the same old doctors peddling the same old opinions for the same old lawyers.  In legal circles it is known as “adversarial bias”, irrespective of which party is at fault. The New South Wales Law Reform Commission looked at this problem and identified three varieties of adversarial bias.

Imagine that I decided to develop a business model for my practice by marketing my services to a particular law firm. I’d need to establish a unique point of difference; now it may be that I have a self-serve coffee machine in the waiting room, or that all my reports are done on rose-scented paper, but if I was really serious I’d promise to deliver the one thing that really mattered to them.  What I am writing about here is called partisanship, which is where the doctor and lawyer have a too-cosy relationship. This arrangement sees a steady flow of cross-referrals, and a consistent stream of medical reports that invariably lean in the direction that pleases the lawyer.  Here are the Commission’s three types of adversarial bias.

• Deliberate partisanship – an expert deliberately tailors evidence to support his or her client.
• Unconscious partisanship – the expert does not intentionally mislead the court, but is influenced by the situation to give evidence in a way that supports the client.  This is also known as commissioning bias.
• Selection bias – litigants choose as their expert witnesses persons whose views are known to support their case.

There are numerous other biases that creep into medical reports.  Here are a few:

Attribution bias, where the patient perceives a particular exposure to be the basis of the medical condition, and where the doctor accepts that as the causation.

Treating-doctor bias, is not something that has been described (in my searching) in the literature.  I read a lot of medical reports, and observe that most doctors, when providing reports about their own patients tend to take on the advocate role.  There is an obvious bias, but the courts tend to place greater faith in the opinion of treating doctors than any independent assessors.  Maybe the treating doctor represents the norm, and it is the independent assessors who are all biased?

Implicit biases, are the biases that pervade our everyday lives, whether they be racial, cultural, religious or gender-based. The following are a few examples of where they have spilled over into medical reports.  “he was born in Iraq, and paid a people-smuggler for a passage on an overcrowded boat, and was in detention at Christmas Island”, “she had to first remove her hijab to enable me to examine her neck”, and “she has been in a long-term relationship with another woman”. All perfectly true, no doubt, but if irrelevant to the purpose of the report, serve only to declare one’s implicit biases.

Pay-back bias.  It seems that everyone who writes about biases gets to describe one of their own invention; this is mine.  Medical reporting generally includes the requirement to comment on the diagnosis, treatment or opinion of another doctor, with the result that some doctors may take offence because another may disagree. This process of disputation can become fuelled by the requirement of insurers and lawyers for the doctors to produce supplementary reports. In these circumstances the medical report often becomes a vehicle to engage in a slanging match between doctors.  Objectivity becomes the casualty.  This “payback” phenomenon has become increasingly evident in recent years.

The law recognises an even more sinister form of association, professional misconduct. In this situation, what changes hands is considerably more generous than a couple of glasses of beer on a Friday evening, usually involves the doctor venturing outside the usual field of expertise, or deliberately falsifying evidence.

Then there are the situations where one is faced with a (potential) conflict of interest.

Suppose your self-managed superannuation fund had a few shares in The Jolly Trolley, which was a national supermarket company, and you happened to have a patient who was a night-fill worker for the company. The extent to which your patient’s worker’s compensation claim was going to impact on the company share price or dividend would be so teeny-weeny that it would not be an issue.  But suppose that all The Jolly Trolley stores within about 10 km referred all their work injury cases to your clinic?

Some years ago I was treating a woman with a back condition, and when coming to one of her appointments with me, she had tripped on some indoor garden ledging in the foyer of the consulting rooms. The priority was obviously to render assistance, make her comfortable in one of the consulting rooms, and to undertake a quick assessment to exclude serious injuries.

The next issue to address was whether I had a conflict of interest if I continued to treat her. The consulting rooms that I occupied were held by a colleague and his company.  My arrangement with that colleague was that his company provided my room and all administrative support, for which I was billed a proportion of my receipts. The company in turn leased the rooms as part of a two-storey office block occupied by several tenants.  The foyer where the patient had slipped was outside the area that my colleague’s company leased.

In that situation I was comfortable that I did not have a conflict of interest.  I had another patient who was involved in a minor slip in the rooms, but in that case I immediately realised that I had a conflict of interest, and arranged for the worker to continue his treatment by a colleague at another location.  When approached by his solicitor for a report, I agreed to furnish a report as to the facts of the incident, but declared that I would not be an expert witness and offer an opinion regarding that patient’s increased symptoms.

I had a colleague who found herself in a very interesting situation.  She had been treating an adolescent with psychological problems that developed out of school bullying.  Eventually the patient (actually it was the parent who initiated legal proceedings) consulted a solicitor, who wrote to the doctor for a medical report.  In that report the doctor stated that it was her opinion that it was not in the interest of the patient’s long-term well-being to pursue a legal solution.  Many years later, during which time the patient was no longer under her care, the doctor received notice to attend court.  She advised the patient’s solicitor that she believed that she had a conflict of interest.  She did not hear further from the solicitor.

Finally, The Bellman’s Fallacy. The Bellman was the central figure in Lewis Carroll’s nonsense poem “The Hunting of the Snark”.  The Bellman declared

"Just the place for a Snark!" the Bellman cried,
As he landed his crew with care;
Supporting each man on the top of the tide
By a finger entwined in his hair.
"Just the place for a Snark! I have said it twice:
That alone should encourage the crew.
Just the place for a Snark! I have said it thrice:
What I tell you three times is true." [Carroll]

The Bellman’s Fallacy was originally described in relation to medical research and heath policy, where an erroneous conclusion may be quoted by another researcher, and again by a third researcher, after which it becomes an accepted truth.  In the arena of medical reporting it occasionally manifests itself, e.g. “carpal tunnel syndrome in a process worker is always work related”.

A few quick tips.

If your opinion is at odds with that of one of your colleagues, and you know that the other fellow has a too-cosy relationship with the lawyer who is furiously writing notes and handing them to the barrister who is cross-examining you, don’t suggest that the other fellow is biased.  Yes, he may be – but he is not on trial.  What the court is trying to do is to test the facts and opinions.

Your opinion, whether it be in a report or given as evidence in a court, will be challenged, and if it is well founded, you are right to doggedly defend it.  That is not regarded as bias, and is indeed the mark of a truly expert expert-witness.