NOVA SCOTIA COURT OF APPEAL
Citation: R. v. Stewart, 2003 NSCA 150
Date: 20031230
Docket: CAC 194890
Registry: Halifax
Between:
Peter Edward Stewart
Appellant
v.
Her Majesty The Queen
Respondent
Judges: Oland and Hamilton, JJ.A.; and Cacchione, J. (Ex Officio)
Appeal Heard: October 15, 2003, in Halifax, Nova Scotia
Held: Appeal dismissed per reasons for judgment of Oland, J.A.; Hamilton, J.A. and Cacchione, J. concurring.
Counsel: Donald C. Murray, Q.C., for the appellant
Dana Giovannetti, Q.C., for the respondent
Reasons for judgment:
[1] The appellant was charged with the second degree murder of Brandy Bowman, age 26 months, who died April 17, 2001. At trial, the theory of the Crown was that the appellant had shaken the child in such a violent manner that he meant to cause bodily harm that he knew was likely to cause her death or was reckless whether death ensued or not. The defence suggested that the injuries which resulted in her death could have been incurred in an accidental fall down stairs.
[2] Following an eleven day trial, Justice Margaret J. Stewart of the Nova Scotia Supreme Court found the appellant not guilty of second degree murder, but guilty of manslaughter. Her decision is reported as R. v. Stewart (2003), 212 N.S.R. (2d) 250. She subsequently sentenced him to a term of 54 months imprisonment.
[3] The appellant appeals from conviction. He asks that the conviction be set aside pursuant to s. 686(1)(a) of the Criminal Code and that an order be made quashing the conviction and directing an acquittal pursuant to s. 686(2)(a) of the Code. The grounds of the appeal are two-fold:
1. Whether the trial judge erred in law by failing to properly apply the burden of proof; and
2. Whether the verdict should be set aside on the grounds that it is unreasonable or cannot be supported by the evidence.
The Burden of Proof
[4] The appellant does not take any issue with the trial judge’s summation of the evidence in her decision. In his view, she accomplished a “largely magnificent distilling” of that evidence in her judgment.
[5] Nor does the appellant suggest that the trial judge erred in identifying or stating the applicable law. He acknowledges that her decision set out the correct principles in regard to the presumption of innocence and the requirement for proof beyond a reasonable doubt. He notes that the trial judge was aware that the Crown’s case was founded on circumstantial evidence. Finally, he observes that in her ¶ 83, she reminded herself that in such a case the Crown does not discharge the burden of proof beyond a reasonable doubt unless the trial judge is satisfied beyond a reasonable doubt that the only rational inference to be drawn from the proven facts is the guilt of the accused.
[6] The appellant submits however that after stating that test correctly, the trial judge failed to apply it properly but instead proceeded to compare the theories of the Crown and the defence and to decide which she favoured. According to the appellant, the trial judge required the defence to convince her that Brandy had died in a fall rather than requiring the Crown to prove that the child died from shaken baby syndrome and that she could not have died in a fall.
[7] It would be helpful at this point to briefly summarize some of the evidence before the trial judge and her decision. Brandy Bowman and her mother had moved to the appellant’s home in early March 2001. Brandy’s bedroom was up a steep flight of stairs and there was a door at the bottom of those stairs. After the child had been put to bed the evening of April 14, 2001, the day before Easter Sunday, her mother and the appellant filled plastic Easter eggs with candies. Several were placed on the stairs and others were hidden throughout the residence.
[8] Whatever happened to Brandy happened shortly after her mother left for work at 9:00 p.m. and while the appellant and the child were alone at home. The appellant testified at trial that he had heard three noises, the last one being the door at the foot of the stairs. His evidence was that when he opened the door, Brandy was lying face down on the bottom step, with her head downwards and her feet upwards. He telephoned his father who arrived shortly with his common-law wife Kathy Labrador. They headed for the hospital with the unconscious, gasping child. At trial, the appellant testified that he was “panicked” and Ms. Labrador described him as “hysterical” during the drive. She also testified that the appellant found a bump on Brandy’s head and screamed at her to check it which she did. She maintained that she felt something like a welt on the back of the child’s head which was just in line with the bottom of her ears.
[9] Their vehicle was intercepted by an ambulance which took Brandy to the South Shore Regional Hospital. Paramedic Joanne Newell did not note any bumps or bruises on the child’s head when she examined her. Dr. Chris Naugler who performed emergency treatment at the hospital did not see any evidence of a head injury. Brandy Bowman was transported to the IWK Health Centre in Halifax and died three days after sustaining her injuries.
[10] At trial, the Crown called 11 medical witnesses, seven of whom were doctors. The staff at the IWK Health Centre who treated Brandy included Dr. Michael Riding, paediatric neuro-radiologist; Dr. Kathryn Morrison, paediatrician and Director of the IWK Health Centre child protection team; Dr. David Clarke, neurosurgeon; and Dr. George LaRoche, paediatric ophthalmologist. According to her decision, the trial judge was also particularly interested in the evidence of Joanne Murphy, anatomical pathologist and Dr. Robert James Macaulay, neuropathologist.
[11] Dr. Morrison defined a “syndrome” as “several findings which, by themselves, may not be specific for a diagnosis but when taken together provide a unified diagnosis.” The constellation of injuries characteristic of shaken baby syndrome consists of inter-hemispheric subdural hemorrhage, retinal hemorrhages, sometimes bone fractures, together with little or no signs of external injury.
[12] All of the medical experts were of the view that the injuries suffered by Brandy Bowman were consistent with those seen in children who have been violently shaken. Drs. Morrison and Clarke described her injuries as those classically associated with shaken baby syndrome. Dr. Macaulay found that the child had sustained “diffuse axonal injury” in the brain stem, a shear injury to the axons of the neurons, which requires major deceleration and can be produced by high speed highway accidents, falls from a second storey window or greater heights, and violent assaults, but not a trip and fall or tumble down the stairs.
[13] After reviewing his findings of Brandy’s injuries including retinal hemorrhages, disc edema, macula edema in both eyes, and retinoschisis in the left eye, Dr. LaRoche was asked why he diagnosed shaken baby syndrome. He replied:
Well, what causes me to say that, and me and everyone else that would look at this, is simply because there’s nothing else that can do that as far as medical knowledge is at the moment.
Questioned whether he had reviewed other literature with respect to such injuries to the eye, he answered that people have looked for other causes for hemorrhages and continued:
And, unfortunately, so far every one of these reports, when you look at the actual report, and not just the summary of it or just not what you want to find out it, its - - it just doesn’t muster up. The evidence is not there that you can create this kind of dramatic picture in the back of one’s eye in other circumstances than the one that we have discussed.
[14] In ¶ 84 of her decision, the trial judge stated that the medical evidence as a whole was “certainly not definitive” as to the mechanisms involved in Brandy’s death. Although not varying in their opinions as to the constellation of injuries being consistent with shaken baby syndrome, the Crown’s medical witnesses did not agree exactly how certain injuries could have been caused. Drs. Riding, Morrison and Clarke were of the view that there had also been a manual strangulation, an opinion which the trial judge rejected. As the trial judge noted, Drs. Murphy and Macaulay testified that there may not have been shaking involved in the child’s death although both added that her death was not accidental.
[15] Each of the Crown’s medical witnesses testified that Brandy’s injuries were not consistent with a fall down stairs. In that regard, several of them pointed out the absence of any significant bruising, swelling, cuts on the head or skull fractures.
[16] The defence did not call any medical expert to contradict the testimony given by the Crown’s witnesses. Its witness, Dr. Lawrence Holt, a kinesiologist, accepted the findings of the pathologists and considered whether Brandy’s injuries could have been caused by a fall. The summary contained in his report stated as follows:
We have examined the circumstances surrounding the catastrophic injuries and death of Brandy Bowman from a movement science perspective. We have considered both environmental and human factors in our analysis. We have generated a reasonable model that explains this event. We believe it could have been due to an accidental fall. We see the injuries as being a series of impacts between the head of Brandy Bowman and the stairs.
[17] Dr. Holt’s model was illustrated by a diagram in his report. It shows Brandy tripping at the top of the stairs, followed by a complex, twisting and somersaulting fall which ends with her resting face upwards. In his view, such a fall would account for the injuries as found by the pathologist and neuro-pathologist, including certain bruises and abrasions on the child’s body that her mother had not seen prior to this incident and the bilateral retinal hemorrhages and retinoschisis found by Dr. LaRoche. The pathologist had identified the freshest of the bruises as those on each elbow.
[18] The trial judge stated that the issue was whether or not the appellant killed Brandy and if he did, it was not necessary to decide the mechanism by which the death took place. She did not believe the appellant’s evidence as to how events unfolded on the evening Brandy was injured, pointing out, among other things, inconsistencies in the accounts he gave to other witnesses. She then proceeded to determine whether, on the basis of all the evidence, she was satisfied beyond a reasonable doubt that the appellant killed Brandy. The trial judge concluded on the totality of the evidence that the only rational explanation for Brandy’s injuries was shaking in some context, without manual strangulation, by the appellant. Not having been satisfied that the Crown had established the requisite mental element for second degree murder beyond a reasonable doubt, she found the accused not guilty of second degree murder but guilty of manslaughter.
[19] The appellant submits that the trial judge incorrectly applied the burden of proof. The following paragraphs from her decision form the basis of his argument:
[89] Defence urges the acceptance of Dr. Holt's hypothesis of a fall, particularly given the independent corroboration of the posterior injuries - elbows, buttock, forearm, shoulder and what Cathy Labrador felt with her own hand and the fact that the mechanism of what would be happening to the brain is the same whether it accelerates by shaking or whether it decelerates by an impact to some surface like the base of the stairs.
. . .
[97] Acknowledging, medical science is unable to explain the pathology of retinal hemorrhage or retinoschisis but able to address rapid movement of the vitreous in and of itself as pure speculation and able to point to correlation with proven shaking events, as well as certain diseases and infections of the eye, Dr. LaRoche with his 27 years of experience, extensive review of the literature and professional involvement and study of the pathology and protocol of shaken baby syndrome maintained his opinion that with Brandy's constellation of injuries shaken infant was a proper diagnosis.
[98] Difficulties can arise when unique circumstances are postulated that could result in the type of injuries occasioned to a victim although not within the experience of medical experts who testified. However, as stated, the onus on the Crown is not absolute certainty and although the medical evidence was not such as to state that the injuries occasioned could not have been caused by the hypothesized sequence of events, the requisite Crown burden is met particularly through the evidence of Dr. LaRoche and where inconsistent, rejecting the evidence of Dr. Holt as well as that of the accused. When I consider his testimony in combination with other experts of the Crown and on the totality of the evidence, I conclude beyond a reasonable doubt that the only rational explanation for the injuries suffered by Brandy is shaking in some context, without manual strangulation occasioned to Brandy by Peter Stewart.
. . .
[101] Given the paramedic, Joanne Newell's focus, under the circumstances of the clinical history presented, inclusive of viewing bruises and her expertise in feeling for soft tissue head injuries and feeling "no deformities at all" within what was less than 30 minutes, I am satisfied despite Dr. Holt's confidence from his experience that such impact injuries can dissipate in a short time span that Cathy Labrador's positive and confirming reaction to Peter Stewart's hysterics about a lump on the back of Brandy's head is something she believes existed both then and now; however, I am not convinced that it did. [Emphasis added]
[20] The appellant argues that the trial judge considered the testimony of Drs. LaRoche and Holt and then chose which she favoured. In a criminal proceeding where there is a conflict between experts, the decision is not to be based on whose evidence was preferred. According to R. v. Parnell (1983), 9 C.C.C. (3d) 353 (Ont. C.A.), leave to appeal to the Supreme Court of Canada refused February 7, 1984, the proper direction to be given in a case of conflicting evidence was summarized in R. v. Platt, [1981] Crim. L.R. 332 (Ont. C.A.). There the pathologists for the prosecution and the defence had expressed differing opinions on a matter. The direction to the jury to decide whose evidence it preferred was held to constitute misdirection and the conviction was quashed. In Parnell, supra, the Ontario Court of Appeal stated at p. 355 that it was not proper to limit the jury by asking whose evidence was preferred.
[21] The appellant also submits that ¶ 89, 98 and the statement in ¶ 101 of her decision that she was “not convinced” that there had been a bump on Brandy’s head shows that the trial judge had improperly placed the burden on the defence to prove that Brandy had been injured falling down stairs. He says that the trial judge committed the same errors as had been made in R. v. Robert (2000), 143 C.C.C. (3d) 330 where the Ontario Court of Appeal allowed an appeal from conviction for intentionally causing fire to a dwelling.
[22] In Robert, supra, the Crown’s case at trial had been entirely circumstantial and the appellant, who had been present when the fire started, had claimed the fire had been accidental. The trial judge had examined each of the possible causes for the fire offered by the defence. Sharpe, J.A., writing for the Court of Appeal stated at ¶ 21 and 22:
The trial judge took as the starting point the presence of the appellant at the scene of the fire and then scrutinized the case for the defence by asking whether it established innocent cause for the fire. The appellant was, in effect, required to satisfy the trial judge that an accidental cause was made out as a "reasonable inference" or as a "reasonable conclusion" from the "proven facts". With respect, that was not the issue from the perspective of the appellant. It was for the Crown to show beyond a reasonable doubt that there was no other reasonable inference than the guilt of the accused. The appellant was entitled to an acquittal if there was a reasonable doubt on all of the evidence, a conclusion sustainable at a threshold significantly lower than a "reasonable inference" from "proven facts".
In my view, the trial judge took a formula, designed to test the case for the Crown, and applied it to the accused. This set the standard too high. It is trite law that an accused need only raise a reasonable doubt as to guilt to gain an acquittal. There is no affirmative obligation on an accused to prove anything by way of reasonable conclusion or reasonable inference. As Martin J.A. stated in R. v. Campbell (1977), 38 C.C.C. (2d) 6 (Ont. C.A.) at 22, ". . . reasonable possibilities in the accused's favour may give rise to a reasonable doubt".
[23] The appellant further argues that Dr. LaRoche never explained how bilaterial retinal hemorrhages occur but only that there was a correlation between its occurrence and certain situations. He says that the trial judge did not give Dr. Holt’s evidence the same consideration given to that of Dr. LaRoche. He maintains that so long as Dr. Holt’s explanation was not disproven by the Crown, it was not possible for the trial judge to come to a conclusion of guilt. Moreover, the appellant says that the emphasized portion of ¶ 98 shows that the trial judge had had a reasonable doubt which obliged her to acquit, and that she erred in law when she failed to do so.
[24] With respect, I am unable to accept any of these submissions. It is apparent from her decision that the trial judge approached the burden of proof as required by R v. W.(D.), [1991] 1 S.C.R. 742 and R. v. Lifchus, [1997] 3 S.C.R. 320. She asked herself whether she believed the accused; if not, whether his evidence raised a reasonable doubt; and if not, whether on the totality of the evidence which she accepted she was left with a reasonable doubt.
[25] The essence of the appellant’s arguments on the burden of proof is that the evidence of Dr. Holt was capable of providing proof of, or at least a rational basis upon which to believe that Brandy had been injured in an accidental fall. However, an examination of her decision shows that the model put forward by Dr. Holt on behalf of the defence was not accepted by the trial judge as either raising a reasonable doubt or establishing an alternative rational explanation. She was severely critical of both Dr. Holt’s expertise and his methodology. At ¶ 95 of her decision, the trial judge spoke of the kinesiologist’s “efforts to comment on what he read as medical findings in the literature” relating to internal injuries and continued:
. . . Cross examination revealed how Dr. Holt had not been analytical of the methods of calculation provided in some [of] the literature and how he lacked expertise to interpret the findings from various medical instruments and medical findings relating to injuries reported in the literature resulting from various falls. It is not a question of various medical experts conclusions being compared and challenged rather it is just Dr. Holt quoting from various articles and as a non medical person interpreting research and quoting findings. In assessing the medical information as to the cause of injury, Dr. Holt's expertise is in movement studies and limited to "the process" with exposure to resulting sports injuries not inclusive of cerebral.
[26] Her concerns led to the trial judge’s rejection of Dr. Holt’s evidence where inconsistent with that of other medical experts. In ¶ 97 and ¶ 95, she referred to Dr. LaRoche’s years of experience, his own analysis of the medical literature and his professional involvement and study relating to shaken baby syndrome. However, I am not persuaded that she simply compared the theories of Drs. LaRoche and Holt and chose the one she preferred contrary to Platt, supra. As is apparent, the trial judge appreciated that Dr. Holt’s hypothesis rested on a number of assumptions and his interpretation of literature which was not accepted by other experts. She accepted his evidence as having very limited value. In ¶ 98 the trial judge stated that on the totality of the evidence she concluded beyond a reasonable doubt that the only rational explanation of Brandy’s injuries was shaking in some context by the appellant.
[27] Moreover, I am unable to agree that the trial judge made the same errors as led to a new trial in Robert, supra. According to ¶ 9 of that decision, the trial judge there did not explicitly reject the evidence of that appellant or express a clear view as to his credibility. In the case before us the trial judge did not believe the appellant and was not satisfied with the evidence of Dr. Holt which she expressly rejected whenever inconsistent with that of other experts. As a consequence, unlike the situation in Robert, supra, no other alternative rational explanation acceptable to the trial judge had been made out or remained alive to raise a reasonable doubt.
[28] Nor have I been persuaded that ¶ 98 of her decision discloses that the trial judge had concluded that she had had a reasonable doubt and yet failed to acquit. The portion of ¶ 98 on which the appellant relies is emphasized below:
Difficulties can arise when unique circumstances are postulated that could result in the type of injuries occasioned to a victim although not within the experience of medical experts who testified. However, as stated, the onus on the Crown is not absolute certainty and although the medical evidence was not such as to state that the injuries occasioned could not have been caused by the hypothesized sequence of events, the requisite Crown burden is met particularly through the evidence of Dr. LaRoche and where inconsistent, rejecting the evidence of Dr. Holt as well as that of the accused. . . . (Emphasis added)
[29] I do not agree with the appellant’s interpretation of this passage as an acknowledgment by the trial judge that a reasonable doubt had been raised by the scenario that Dr. Holt had pieced together to explain Brandy’s injuries. In my view, when read in the context of the preceding sentence and paragraph and the trial judge’s unfavourable comments regarding Dr. Holt’s hypothesis, the emphasized words were simply comment by the trial judge on the difficulties which arise when medical experts could not be unanimous in their opinions. Furthermore, I cannot accept that the passage was intended to show reliance on the evidence of Dr. Holt as the basis for a reasonable doubt when that evidence is so heavily discounted within the same sentence.
[30] While he acknowledged that the Crown is not required to meet a standard of absolute certainty, the appellant suggested that the evidence of the Crown’s medical experts was insufficient to convict because the Crown had not proved the mechanism of how Brandy was killed. I reject any such suggestion. In my view, the burden on the Crown to persuade the trial judge beyond a reasonable doubt does not include a requirement that the precise mechanism be established, only that the death resulted from an unlawful act which led to the injuries which caused death.
[31] I also reject the appellant’s suggestion that the criminal law standard of proof beyond a reasonable doubt cannot be met whenever expert witnesses are not on all fours on every aspect of their opinions or where the medical opinion evidence is anything less than absolute. The Crown’s factum responded neatly to this argument as follows:
Legal reasoning does not require absolute certainty concerning the elements of the offence, and the absence of reasonable doubt does not require a foundation of absolute certainty in the evidence. What intervenes between the evidence and the verdict is the Court’s assessment of how the evidence in its totality stacks up and whether this composite assessment admits of reasonable doubt in the judicially trained mind. This is not a mechanical translation from evidence to verdict. It is an inductive and common sense process, guided by experience. It is experience, not deductive reasoning, which is the life of the law.
To give effect to the appellant’s submission would result in the imposition of a requirement that witnesses, whether lay or expert, be unanimous in their version of events or opinions before a guilty plea could ever be entered. This is not possible in many cases and it is precisely the role of the trial judge to assess the evidence and whether the Crown has persuaded her beyond a reasonable doubt.
[32] In summary on this ground I do not agree that the trial judge, in reaching her decision, failed to apply the burden of proof properly.
Unreasonable Verdict
[33] On a hearing of an appeal against conviction, this court may allow the appeal where it is of the opinion that the verdict should be set aside on the ground that it is unreasonable or cannot be supported by the evidence: s. 686(1)(a) of the Criminal Code.
[34] The standard of review to determine whether a trial court reached an unreasonable verdict was established in R. v. Yebes, [1987] 2 S.C.R. 168. McIntyre, J., writing for a unanimous court held at p. 186:
The Court must determine on the whole of the evidence whether the verdict is one that a properly instructed jury, acting judicially, could reasonably have rendered. While the Court of Appeal must not merely substitute its view for that of the jury, in order to apply the test the court must re-examine and to some extent reweigh and consider the effect of the evidence. This process will be the same whether the case is based on circumstantial or direct evidence.
This approach was maintained in R. v. Burns, [1994] 1 S.C.R. 656 at p. 663 and confirmed in R. v. Biniaris (2000), 143 C.C.C. (3d) 1 at ¶ 42 as the binding test for appellate review in determining whether a verdict is unreasonable or cannot be supported by the evidence.
[35] The appellant included several arguments under this ground of appeal. He points to ¶ 100 of the decision which stated that certain abrasions and bruises occurred after Brandy ate some of the candies that had been stuffed in the plastic eggs and then placed on the staircase. According to the appellant, the trial judge improperly drew this inference from the appellant’s denials in response to questions on cross-examination and there was no evidence that the child had ingested candies. However, Brandy’s mother testified that when she left, the containers on the staircase were all snapped closed. Police officers who examined the premises testified that some of the plastic eggs were open and that candies were found out on the stairs. Shown photographs of open containers, the appellant responded that they could have opened spontaneously. Only the appellant, Brandy, the dog and the cat were home when the child sustained injuries. In these circumstances, I cannot agree that the trial judge engaged in speculation or that she drew an unreasonable inference from the evidence on this point.
[36] The appellant submits that it was unreasonable and perverse of the trial judge to have rejected Kathy Labrador’s evidence of a bump on the back of Brandy’s head. He says that this independent witness was unshaken on cross-examination and that the absence of any external head injury was a factor for several of the Crown’s medical witnesses in discounting the possibility of a fall downstairs. He also pointed to other items of evidence such as the red birthmark on the base of the child’s neck which seemed to the mother to be larger than usual, certain bruises and abrasions, and the evidence of Dr. Murphy that it is possible to suffer an impact which is not visible on an autopsy.
[37] It is without question that a trial judge may believe all, some, or none of the testimony given by a witness. Here, the trial judge chose to reject the evidence of Kathy Labrador on this point. I do not accept that she did so arbitrarily or because it was inconvenient to the theory of the Crown. In stating that she was not convinced that there had been a bump, the trial judge set out the circumstances under which the witness says she felt a welt in context, namely during a panicked drive to the hospital and in confirmation as requested by an hysterical appellant.
[38] The assessment of a witness’ credibility is a matter within the province of the trial judge. The trial judge did not believe the evidence of the appellant, the child’s caregiver the night Brandy was injured and the only person with her. Her rejection of his testimony regarding a fall down the stairs which would have raised an alternate rational explanation if believed, has not been challenged on appeal.
[39] The other matters raised by the appellant on this ground are various pieces of evidence. They are not findings made by the trial judge.
[40] The record shows that Drs. Riding, Morrison, Clarke, Murphy, Macaulay and LaRoche testified that Brandy died because of injuries consistent with shaken baby syndrome. The same medical experts opined that her death was not caused by a household fall. On the evidence before her, it was open to the trial judge to reach the conclusion she did, that the only rational explanation for the injuries suffered by Brandy Bowman was shaking in some context by the appellant.
[41] Having applied the standard of review in Yebes, supra, re-examined and to some extent re-weighed the evidence, I have not been satisfied that the trial judge reached an unreasonable verdict or one that was unsupported by the evidence.
Conclusion
[42] I would dismiss the appeal.
Oland, J.A.
Concurred in:
Hamilton, J.A.
Cacchione, J.