CITATION: R. v. Kumar, 2011 ONCA 120

DATE: 20110210

DOCKET: C48821

COURT OF APPEAL FOR ONTARIO

Rosenberg, Armstrong & Rouleau JJ.A.

BETWEEN

Her Majesty The Queen

Respondent

and

Dinesh Kumar

Appellant

James Lockyer and Alison Craig, for the appellant

Gillian Roberts, for the respondent

Heard: January 20, 2011

On appeal from the conviction entered by Justice Edward F. Ormston of the Ontario Court of Justice (Criminal Division), dated December 3, 1992.

By The Court:

[1] This appeal originates in the reinvestigation of cases where Dr. Charles Smith had provided opinions and evidence as to the cause of death of children. In the early hours of March 18, 1992, Gaurov Kumar, then five weeks old, stopped breathing and was rushed to hospital. Breathing was restored, but Gaurov’s brain was badly damaged. Two days later, on March 20, 1992, life support was terminated and Gaurov was declared dead. Based on Dr. Smith’s opinion that the baby was a victim of “Shaken Baby Syndrome” (SBS), the appellant, Gaurov’s father, who had been caring for the baby at the time of his collapse, was charged with murder. He eventually pleaded guilty to criminal negligence causing death, and the murder charge was withdrawn. After a series of investigations of cases where Dr. Smith provided an opinion, in 2008, the Crown consented to an extension of time to allow the appellant to appeal his conviction. The appellant seeks to have fresh evidence admitted, his guilty plea set aside, and an acquittal entered. The Crown agrees with that disposition. For the following reasons, the fresh evidence is admitted. the guilty plea is set aside, the appeal is allowed and an acquittal entered.

THE FACTS

[2] The appellant is now 44 years of age. He immigrated to Canada from India with his wife, Veena, in 1991. They had their first child, Saurob, just before moving to Canada. Their second son, Gaurov, was born on February 11, 1992. On the evening of the birth, Veena had a seizure in the hospital. She was diagnosed with a brain tumour and remained in hospital for one month.

[3] The appellant’s version of events on the evening of March 17, 1992 was as follows. The appellant was at home with his wife and their two sons. At around midnight, Gaurov woke up crying. The appellant fed him milk from a bottle, burped him and placed him back in his crib where he fell asleep. At 12:30 a.m., Gaurov woke up with a scream. The appellant picked him up, and realized he was not breathing and was turning blue. He told his wife something was wrong. The appellant gave Gaurov CPR including mouth to mouth. Gaurov was still not breathing. The appellant called his brother-in-law to ask what to do (he did not understand the 911 system). His brother-in-law told him to call 911. Although neither the appellant nor his wife spoke English very well, both spoke to the emergency operator and soon after emergency personnel arrived at their home. Mrs. Kumar’s brother and sister-in-law also arrived.

[4] The aunt was seen by emergency personnel to give Gaurov “three to four... good hard shakes ... but not ... a violent shake.” Emergency personnel then took over and administered first aid to Gaurov. He was taken to Scarborough Centenary Hospital and his breathing was restored. The appellant explained to the physician as best he could what had happened. Gaurov was diagnosed by Dr. Beaulieu as having suffered from “cardiorespiratory arrest secondary to choking spell”. A secondary diagnosis was also provided “anemia of unknown etiology, possible central nervous system haemorrhage.”

[5] Gaurov was transferred that same morning to the Hospital for Sick Children and put on life-support. On Friday, March 20, 1992, the life support was removed and Gaurov died. His brother Saurob was immediately apprehended at the hospital by Children’s Aid. Saurob was returned to his mother following a family court proceeding on July 2, 1992.

The Autopsy

[6] On March 21, 1992, the day after Gaurov’s death, Dr. Charles Smith conducted the autopsy at the Hospital for Sick Children. He found extensive haemorrhaging within the brain, behind both retinas and around the spinal cord. It was his opinion that these critical medical abnormalities had all the earmarks of injuries sustained after an episode of Shaken Baby Syndrome. Dr. Smith solidified his opinion after conducting further tests, declaring that the injuries were definitely not accidental in nature and because of the absence of external trauma, the injuries were consistent with Shaken Baby Syndrome. Additionally he explained that the injuries were probably a result of continuous shaking as opposed to a single violent shake.

[7] In his post-mortem report, Dr. Smith also noted an old chronic subdural hematoma in the occipital lobe (located in the back of the skull), the tentorium cerebelli (an extension of the dura on which the lobes rest) and the cervical cord, which he wrote “may have been as old as five weeks i.e. as a result of birth trauma”. Dr. Smith certified the cause of death as “Head Injury”. Dr. Dirk Huyer, a physician who was a member of the Suspected Child Abuse and Neglect Program (the SCAN unit) at the Hospital for Sick Children agreed with Dr. Smith’s conclusion. Dr. Huyer had no training in pathology.

The Appellant’s Arrest and Guilty Plea

[8] On June 26, 1992, the appellant was charged with second degree murder. He was released on bail on July 10, 1992. He was represented by Mr. David Gorrell and Mr. Dhaman Kissoon. In preparation for trial, the defence retained a hospital pathologist, Dr. Jay Naidoo of the Queensway General Hospital, to review Dr. Smith’s findings. Dr. Naidoo had no training in forensic pathology. He verbally advised Mr. Gorrell that he agreed completely with Dr. Smith’s conclusions of baby-shaking, and consequently he was not asked to provide a written report.

[9] Unexpectedly, the Crown offered to withdraw the second degree murder charge if there was a plea to criminal negligence causing death. Subsequently, on December 3, 1992, the appellant entered a guilty plea before Ormston J. of the Provincial Court (Criminal Division) on a new Information alleging Criminal Negligence Causing Death. Justice Ormston accepted a joint submission on sentence and sentenced the appellant to 90 days imprisonment to be served intermittently, followed by two years probation. The transcript of that proceeding is no longer available.

THE FRESH EVIDENCE

Explanation for Guilty Plea

[10] Mr. Gorrell testified at the Inquiry into Paediatric Forensic Pathology (the “Goudge Inquiry”), that the appellant had maintained his innocence from the outset of his case, until the day he attended Mr. Gorrell’s office to provide him with written instructions to the contrary, and to accept the Crown’s offer. Mr. Gorrell prepared a plea “Direction” for the appellant. He believed that one of the main reasons the appellant pleaded guilty may have been to ensure that his family could remain together, and to avoid losing custody of his surviving son forever. He further noted that the appellant ran the risk of deportation had he been convicted of second degree murder or manslaughter and he so advised the appellant.

[11] The appellant filed an affidavit on this appeal. In his affidavit, he explained that after Gaurov’s death, he and his wife were very sad and devastated. He said he cooperated with police because he wanted to know why Gaurov died. He was shocked when he was arrested, and felt confused, frightened and ashamed before his family and the community.

[12] He related in his affidavit (confirmed by medical records) that, in 1991, his son Saurob had had an experience as an infant, similar to what happened to Gaurov, fainting suddenly and turning blue. Three weeks after Saurob’s apprehension by the Children’s Aid Society, CAS authorities placed him in the care of Veena’s brother. Neither the appellant nor his wife was allowed to be alone with Saurob. After the appellant was released on bail, one of the conditions of his release was no contact with Saurob except in the presence of a CAS supervisor.

[13] The appellant explained that he was in a new country with its own culture, and he did not speak English very well. He was told that he would be deported if convicted of murder or manslaughter but assured that the police would not report his case to immigration if he accepted the plea. The appellant explained how the plea would alleviate many pressures for him and his family. At the time, his wife was recovering from surgery and could not cope alone with an infant and no income. The family was afraid of the murder charge, and his defence counsel told him there was no way to challenge the testimony of Dr. Smith. They wanted to put the charge behind them. So after much discussion with his family, the appellant decided to plead guilty even though he maintained that he never harmed Gaurov in any way. Importantly, his probation order did not prevent the appellant from seeing Saurob.

[14] After serving his sentence, the appellant still felt the shame of having to admit to causing Gaurov’s death. He and his wife decided not to have another child. They think about Gaurov every day. While his wife and Saurob became Canadian citizens, the appellant never applied for citizenship for fear that his conviction could lead to an immigration investigation.

Chief Coroner’s Review and the Goudge Inquiry

[15] This case came to light again during a review between 2005 and 2007 by the Office of the Chief Coroner of Ontario of criminally suspicious and homicide cases in which Dr. Charles Smith performed an autopsy or provided an opinion. The results of the review were announced in April 2007, and this case was identified as one of twenty where the external reviewers had significant disagreements with Dr. Smith’s opinion. The reviewer, Dr. Helen Whitwell, a forensic pathologist and neuropathologist from the United Kingdom, concluded in her Autopsy Report Review Form dated December 5, 2006, that while the opinion provided by Dr. Smith was in conformity with accepted opinion in 1992, “[t]his would not be the usual view of the UK forensic pathologists now [although it] may well still prevail with the paediatricians.”

[16] The Chief Coroner’s review eventually led to the Inquiry into Paediatric Forensic Pathology conducted by Justice Goudge of this court. The Inquiry retained Dr. Whitwell to provide a more detailed medico-legal report of Gaurov’s death, and testify before the inquiry. Dr. Whitwell identified a number of issues of potential concern in this case, but did not provide a conclusion about the cause of death, recommending that this case be referred for review. The issues she flagged as concerning were as follows:

First, the significance of the old [subdural hemorrhage], likely an injury from birth, and the question of whether or not re-bleeding can occur, and, if so, how much force, if any, may be necessary.

Second, the possibility that resuscitative shaking could have caused damage, particularly in the context of a chronic subdural hemorrhage.

Third, whether baby Gaurov even had the so-called triad of signs (subdural hemorrhage (SDH), retinal hemorrhage (RH), hypoxic-ischemic encephalopathy (HIE)) as no RH were observed when he was first examined in the emergency room at Scarborough Grace.

Fourth, the specificity of the triad alone, without any other evidence, pathological or circumstantial, of abuse.

[17] Following the Inquiry the appellant brought an application for an extension of time to appeal his conviction. With the consent of the Crown, that application was granted.

Crown Experts post-Goudge Inquiry

(i) Dr. Pollanen

[18] In a lengthy report, Dr. Pollanen has provided a helpful review of the diverse views concerning the validity of Shaken Baby Syndrome. There are a number of controversies surrounding Shaken Baby Syndrome. Most prominent is the meaning that should be attached to the triad of symptoms considered to be indicative of SBS: thin film subdural haemorrhage (SDH), widespread bilateral retinal haemorrhage (RH), and hypoxic ischemic encephalopathy (HIE). In the early 1990’s, when the autopsy was conducted in this case, these findings were widely accepted to be diagnostic of non-accidental head injury. In effect, the infant had been shaken to death with such force that any normal adult would realize that the infant would be seriously injured. The current view of forensic pathologists, although not necessarily clinicians, is that the triad is at worst suspicious, but can no longer be considered absolute proof of traumatic head injury in the absence of other evidence. There are some experts who hold the view that, in fact, there is no such thing as shaken baby syndrome; it is impossible to apply sufficient force to an infant by shaking without there being other injuries, such as trauma to the spine and neck areas.

[19] The role played by evidence of the triad and SBS in convictions of caregivers has been examined at length on two occasions by the England and Wales Court of Appeal, (Criminal Division) in R. v. Harris, [2005] EWCA Crim 1980 and R. v. Henderson, [2010] EWCA Crim 1269. That court having heard extensive evidence from many of the leading experts noted the evolution of thinking about the triad and SBS. In Henderson, the most recent case, at paragraph 6, the court made the important point that the courts cannot finally resolve a medical controversy: “the conclusion of any court as to the medical evidence, whether at first instance or on appeal, is dependent upon the evidence before that court”. These comments are particularly apt in this case. The issues of the triad and SBS were not fully argued before us because the Crown concedes that the appeal must be allowed. The fresh expert evidence, which we describe in greater detail below, provides a basis for finding that the conviction in this case was unsafe without having to resolve any controversy about SBS.

[20] Dr. Pollanen noted that Gaurov had no scalp injuries, no skull fracture, and no diffuse or traumatic axonal injury (forms of severe damage to nerve cells (axons) in the brain caused by force). With respect to the triad, Dr. Pollanen noted that in Gaurov’s case, retinal haemorrhages had not been observed when Gaurov’s eyes were first examined on arrival at the hospital. He hypothesized that the retinal haemorrhages developed later due to raised intracranial pressure from hypoxic-ischemic encephalopathy, rather than shaking. He excluded resuscitative shaking (as may have been observed by the emergency personnel firefighters who first attended) as the cause of the injuries to the brain since the infant was “dead” at the time of the shaking.

[21] Dr. Pollanen concluded that he was unable to establish that Gaurov was killed by shaking, and that although there are medical findings that could be interpreted as evidence of shaking or head injury, those findings were not sufficiently determinative of that conclusion given the current state of medical and scientific literature. Dr. Pollanen also cautioned that the science of paediatric head injury is still evolving, and that his views as described in his report are based on past and current literature, and may well be assessed differently in the future. Dr. Pollanen suggested that other experts should be consulted in other areas including neuropathology, child abuse paediatrics, neuroradiology, ophthalmology and biomechanics.

(ii) Dr. Halliday

[22] Dr. Halliday disagreed with Dr. Pollanen’s opinion that the retinal haemorrhages were likely not present on initial examination at the hospital. He believed that they were there but not seen by the examiner. Further, in his opinion, the (likely) birth-related old subdural haemorrhage “does not offer any particular insights” to the case.

[23] Dr. Halliday wrote that the evidence would cause a physician to be “very suspicious” that the appellant’s case represents a case of non-accidental head injury or abusive head trauma (modern terms for Shaken Baby Syndrome). He explained that today, the finding of the “triad” is considered to be suspicious, but not pathognomonic (necessarily characteristic) of abusive head injury. He also noted that the fact that his aunt shook Gaurov after he had collapsed added complexity to the case, since the autopsy would reflect that event as well. Dr. Halliday concluded that, in the particular circumstances of the case, the cause of death is undetermined.

Appellant’s Experts post-Goudge Inquiry

[24] Following receipt of the Crown reports, the appellant retained three experts, Dr. Chris Van Ee, a biomechanical engineer from Michigan, Dr. Jan Leetsma, a neuropathologist from Chicago, and Dr. Waney Squier, a neuropathologist from Oxford.

(i) Dr. Van Ee

[25] Dr. Van Ee’s expertise does not enable him to provide an opinion as to cause of death for Gaurov, but he was able to provide an opinion on whether shaking could have caused Gaurov’s head injuries. In his report, Dr. Van Ee explained that severe rotational acceleration/deceleration forces can cause subdural bleeding. Dr. Van Ee reported that science does not support Shaken Baby Syndrome theories for two reasons: (a) shaking by a human cannot alone cause head injuries in a normal baby, (b) vigorous shaking of a baby would generally result in trauma to the spine and neck areas; injuries that have not been historically observed in suspected Shaken Baby Syndrome cases.

[26] Dr. Van Ee noted that Gaurov did not have any notable chest or neck injuries. He concluded that in light of all the facts, in his opinion there was no bio-mechanical evidence that Gaurov’s injuries were caused by shaking.

(ii) Dr. Leestma

[27] In his report on Gaurov’s case, Dr. Leestma discussed his own research as well as that of others that has discredited Shaken Baby Syndrome. He stated that based on today’s science, Shaken Baby Syndrome must remain an unproven hypothesis with no scientific, medical or legal significance.

[28] Dr. Leestma reviewed the circumstances of Gaurov’s death and concluded that he likely died as a result of his birth injury (increased intracranial pressure caused by chronic subdural fluid collections with recent bleeding) and not from abusive head trauma whether with or without shaking. He noted that while non-accidental injury cannot be excluded, there is no positive evidence for it. A defensible and logical conclusion is that the child died from natural disease processes and their complications.

(iii) Dr. Squier

[29] In her report, Dr. Squier noted that Dr. Smith’s diagnosis seems to have been based on the triad of subdural haemorrhages, retinal haemorrhages and brain swelling. Dr. Squier spent some time in her report reviewing the causes of the triad, both natural and non-accidental. In her view, the presence of Gaurov’s old and chronic subdural haemorrhages was “very significant”. She noted that birth-related subdural bleeding is common and that re-bleeding is commonly seen by microscopy in any baby with a chronic subdural membrane. This re-bleeding results from the vulnerability of the thin walled vessels in the membrane to leaking and bleeding, which may occur without trauma.

[30] Dr. Squier concluded that the evidence indicates Gaurov developed a subdural haemorrhage at birth. Neuropathological observations are those of fresh bleeding into an older, healing subdural membrane consistent in appearance with having originated at birth. The history suggested that the baby choked and may have aspirated vomit or milk. Choking could have been initiated by a seizure or have been a spontaneous event in the baby compromised by the presence of a healing subdural membrane. She concluded that there were no features which indicate violence or inflicted injury.

The Parties’ Positions

[31] As a result of these expert reports the Crown’s position is that the prosecution can no longer establish a cause of death, let alone a homicidal cause of death. Given that the Crown’s case relies entirely on medical evidence and the appellant has explained his guilty plea, the Crown concedes that the conviction is unreasonable and should be set aside and an acquittal entered.

[32] Counsel for the appellant submits that the fresh evidence has explained the guilty plea and that the appellant should be permitted to withdraw the plea to avoid a miscarriage of justice. The fresh evidence also establishes that the science concerning SBS and the triad has evolved to the point where the conviction, which was based solely on expert evidence, was unreasonable. The appellant goes further and submits that looking at the other circumstantial evidence, it is clear that the appellant did not cause the death of his child.

ANALYSIS

The Guilty Plea

[33] As this court explained in R. v. Hanemaayer (2008), 2008 ONCA 580, 234 C.C.C. (3d) 3 and R. v. T. (R.) (1992), 1992 CanLII 2834 (ON CA), 10 O.R. (3d) 514, even though an appellant’s plea of guilty appears to meet all the traditional tests for a valid guilty plea, the court retains a discretion, to be exercised in the interests of justice, to receive fresh evidence to explain the circumstances that led to the guilty plea and that demonstrate a miscarriage of justice occurred. In our view, this is one of those cases. The circumstances are compelling. At the time he pleaded guilty, the appellant was facing a charge of second degree murder. He was relatively new to Canada and was unfamiliar with the language and the legal system. At the time of the infant’s death, his wife had just returned from hospital after major surgery for a brain tumour. He was facing loss of his liberty for at least ten years, loss of custody of his remaining child and deportation. Competent counsel had been unable to obtain opinion evidence to refute the opinion of the then leading expert in the province that the appellant had intentionally caused the death of his child. Like in Hanemaayer, the appellant faced a terrible dilemma. The justice system now held out a powerful inducement: a reduced charge, a much-reduced sentence (90 days instead of a minimum of ten years), all but the elimination of the possibility of deportation, and access to his surviving child. Given the persuasive value of the fresh expert evidence that shows that the conviction was unreasonable, this is a proper case to set aside the guilty plea to avoid a miscarriage of justice.

The Fresh Expert Evidence

[34] It is important to place the expert evidence in its proper context. This appeal differs from some of the other cases recently heard by this court where an opinion by Dr. Charles Smith was at the heart of the case. The opinion given by Dr. Smith in 1992 was one that would have been supported by many other experts and, in fact, was apparently supported by a pathologist consulted by the defence at the time. Medical science has now advanced to a point where the existence of the triad of symptoms alone, while suspicious, is not diagnostic of non-accidental head injury. When the appellant was charged and pleaded guilty there was nothing but the triad to support the prosecution case of intentional infliction of injury.

[35] There are further particular problems in the appellant’s case. Based on Dr. Pollanen’s opinion, it may be that the triad was not even present, in which case even the suspicion of Shaken Baby Syndrome is not supported. The existence of the old subdural haematoma, which probably occurred at birth, is another complicating factor. It is the opinion of one of the defence experts that the death could have been caused by natural disease processes and their complications, possibly associated with the chronic subdural haematoma.

[36] For the reasons given earlier, this is not an appropriate case to resolve the issues surrounding the triad and Shaken Baby Syndrome. Based on the record presented in this case and given the concessions by Crown counsel, it is sufficient to say that the appellant’s conviction is unreasonable. There is no circumstantial evidence to support a finding that the appellant was responsible for his child’s death and the medical evidence is inconclusive. What evidence there is shows the appellant to have been a loving parent who pleaded guilty because of the enormous stress he was under at the time. As appellant’s counsel put it in their factum, the pressure proved too much. A portion of the appellant’s affidavit puts the matter well:

My family, including my wife, wanted me to accept the Crown’s offer. My wife was still recovering from her surgery and could not cope alone with one infant and no income. We all wanted to put my charge behind us. We were all scared of the murder charge. My lawyer told me that we did not have any way to challenge the testimony of Dr. Smith. So I agreed, after much discussion with my family, to plead guilty as I did. It was the hardest decision I ever had to make. I do not want my guilty plea to ever be interpreted to mean that I did anything to harm Gaurov. I did not. My wife knows this too.

[37] To hold the appellant to his guilty plea in these circumstances and given the fresh expert evidence, would constitute a miscarriage of justice.

DISPOSITION

[38] Accordingly, the fresh evidence is admitted, the guilty plea is set aside, the conviction is quashed and an acquittal entered. We wish to thank counsel, Mr. Lockyer, Ms. Craig and Ms. Roberts for their extremely valuable assistance in resolving this very difficult case.

Signed: “M. Rosenberg J.A.”

“Armstrong J.A.”

“Paul S. Rouleau J.A.”

RELEASED: “MR” FEBRUARY 10, 2011