Tuesday, April 15, 2008

Part One: Case Study; Systemic Factors; The Amber Case; As Presented By The Affected Families Group;

Among the most insightful reads that have come out of the Goudge Inquiry are studies of several of the cases studied by the Inquiry - with a view to identifying the systemic issues that they raise.

These case studies have been filed by lawyers Peter Wardle and Julie Kirkpatrick who represent the Affected Families Group;

The first case has been identified by the Inquiry as the "Amber case:;


By way of brief background:

Amber was born in Timmins, Ontario on March 13th, 1987.

She died on July 30th, 1988 at the age of sixteen (16) months at the Hospital for Sick Children in Toronto.

Criminal proceedings were initiated against her twelve (12) year old babysitter, SM, who was charged with manslaughter.

The trial commenced on October 15th, 1989.

It extended over some thirty (30) hearing dates over the next thirteen (13) months and ultimately concluded in November of 1990.

In 1991, Mr. Justice Dunn acquitted SM of the charge which she faced.


"The key systemic issues raised by the case involve confirmation bias, tunnel vision, misleading testimony, lack of objectivity in approaching scientific controversy, and failure to appreciate the lessons of the case after the fact," the first case study begins;
"There are several striking examples of confirmation bias in the case," it continues;

"Dr. Smith’s testimony that the autopsy was a “fishing expedition” about which all of the experts were pessimistic is one obvious example.

It suggests that he approached the autopsy with a conclusion already formed.

This obviously troubled Justice Dunn.

Another example is the forehead or subgaleal bruise.

According to the overview report, on August 12, 1988, following the decision to exhume Amber’s body, the police had a meeting with Dr. Young.

Police notes indicate that Dr. Young told them that “the autopsy should reveal evidence of bruising between the scalp and skull which would tend to support the babysitter’s story or the absence of bruising which would strongly suggest infant shaking syndrome”.

On exhumation and autopsy Dr. Smith did in fact note a “reddish-brown “forehead bruise and “reddish-brown bruise on the frontal scalp” corresponding to the forehead bruise179, which Justice Dunn later referred to as “bruise 1”.

This bruise was discounted by Dr. Smith as being a trivial injury of little significance.

He appears to have assumed (and so advised one of the defence experts) that it predated the collapse, although this does not appear to have been the case.

At least four of the defence experts, on the other hand, opined that from autopsy photographs they saw bruising in the subgaleal skin (the deeper layers of skin underneath the scalp) consistent with SM’s explanation of a fall.

In other words, confirmation bias led Dr. Smith to discount findings which did not fit with his a priori conclusion that the case was a shaking death.

Tunnel vision helps explain some obvious questions about the case, such as: how could a 12 year old girl shake an 18 month old infant to death?

And why would she do so?

Having reached a conclusion that it must be a shaking death, the prosecution experts made the evidence fit that conclusion (the “bolt out of the blue”) and ignored or discounted other evidence pointing towards the explanation of a fall (the bruise, unilateral subdural hematoma, lack of any signs of prior abuse, the explanation provided by SM).

Ms. Reginbal’s unwillingness to critically evaluate the decision of Justice Dunn at the SCAN team meeting is further evidence of tunnel vision.

As seen later in Tyrell’s case, once the fall was excluded early on in the investigation the facts were arranged to fit an alternate hypothesis, which was made to appear compelling.

Dr. Smith’s misleading evidence is well documented in the decision of Justice Dunn and the Overview Report.

His use of anecdotes, his references to sociological factors in shaking deaths, and his evidence regarding the minimum height for a lethal death stand out.

However, Dr. Smith’s comments on the academic literature are one of the most striking feature in Amber’s case.

At the time, Dr. Duhaime had just published what became a seminal article on the biomechanical forces at play in shaking injuries.

That article suggested that death from pure shaking was unlikely without the evidence of other blunt trauma.

It may well have been the view of many clinicians and pathologists working in the field of infant head trauma at the time that short falls could not kill, but there was certainly opinion to the contrary.

As time went on, opinions in this area began to change.

Dr. Huyer testified that by the mid to late 1990’s he was aware that short falls could kill in extraordinary circumstances.

Dr. Whitwell testified that increasingly there has been literature to suggest that occasionally a low level fall can kill.

Her 2002 paper cites a number of articles published in the period 1984-2002.

Over the next 10 years after his testimony in the Amber case, however, Dr. Smith clung to the illusion that the literature was moving in his direction:

May 4, 1992: In his letter to the CPSO he stated “...in the months that have followed since her death, the increasing body of medical literature in the area of child abuse serves to underscore my opinion:”

March, 1994: While testifying in Dustin’s case, he described the medical literature “which was very controversial five years ago, in fact has become less so...”

March 9, 1998: In his letter to the CPSO he stated “In my view, the medical literature does not support the hypothesis that Amber’s death is attributable to a fall down several carpeted steps. At the time I testified in this trial, the literature was strong on this point and, in the intervening years, the literature is even more definitive”.

January, 2000: During the preliminary in Tyrell’s case, he stated “the literature is very clear that this type of head injury cannot occur as a result of an accidental fall that occurs about the home. The only exception to that rule is the [epi] dural haemorrhage” , and “These are some good studies published in the last ten years that I think are helping to clarify that...”

March 2, 2001: In his letter to the CPSO he stated “...the literature now is more complete and serves to reinforce my earlier and honestly held opinion...”

One can usefully contrast Dr. Smith’s approach to being an expert witness to the Guidelines from the American College of Emergency Medicine:
The expert witness should not provide expert medical testimony that is false, misleading, or without medical foundation. The key to this process is thorough review of available and appropriate medical records and contemporaneous literature concerning the case being examined.

On January 30th, 1992 a meeting took place involving members of the SCAN team, Crown Attorneys, and Dr. Smith to discuss the implications of Justice Dunn’s decision.

This should have been an important “lessons learned” meeting for all involved.

Instead, it appears to have been an exercise in “damage control”.

The case itself had no value as a precedent because the “family court judge” was “at the bottom of the heap” and it would be acceptable to indicate that the team simply disagreed with the decision if it came up at a later date.

Judging from contemporaneous notes, it appears that Dr. Smith took away no lessons from this meeting, nor did others."

Harold Levy...hlevy15@gmail.com;