COURT OF APPEAL FOR ONTARIO
DATE: 20220509
DOCKET: C68120
Feldman, van Rensburg and Coroza JJ.A.
BETWEEN
Her Majesty the Queen Respondent
and
Philip Grandine Appellant
Counsel: Michael Lacy and Bryan Badali, for the appellant Dena Bonnet, for the respondent
Heard: November 1, 2021 by video conference
On appeal from the conviction entered on February 28, 2019 by Justice Faye E. McWatt of the Superior Court of Justice, sitting with a jury, and the sentence imposed on January 7, 2020, with reasons reported at 2020 ONSC 110.
van Rensburg J.A.:
A. OVERVIEW
[1] The appellant appeals his conviction for manslaughter in the drowning death of his wife, Karissa Grandine. He also seeks to appeal his sentence of 15 years in prison, less time served.
[2] This was the appellant’s second trial in relation to Ms. Grandine’s death. In 2012, he had been tried on a charge of first degree murder. The Crown alleged that the appellant, motivated to continue an extramarital affair, surreptitiously drugged the deceased with lorazepam (an anti-anxiety medication whose main effect is sedation) and intentionally caused her death by drowning. The jury found the appellant not guilty of murder, but guilty of manslaughter. On appeal, the verdict was overturned and a new trial was ordered: 2017 ONCA 718, 355 C.C.C. (3d) 120. This court held that the fairness of the trial had been compromised when, in response to a jury question, the trial judge introduced a new route to liability that was inconsistent with how the trial had proceeded and without the jury having heard submissions on the issue from the parties. The new theory was that the appellant could be liable for manslaughter by committing the unlawful act of failure to provide the necessaries of life (contrary to s. 215 of the Criminal Code) in allowing the deceased to take a bath after he knew she had ingested lorazepam.
[3] The appellant was retried by judge and jury on a charge of manslaughter, contrary to s. 236(b) of the Code. The indictment particularized the alleged means as “either … an unlawful act or by criminal negligence”.
[4] On this appeal, the appellant raises three grounds against his conviction. First, he asserts that the trial judge erred in leaving with the jury a theory of liability for which there was no evidence – manslaughter by criminal negligence via omission – that permitted the jury to find the appellant culpable whether or not he administered or provided the lorazepam to his wife, so long as he knew she was sedated when he left the house while she was in the bath. This was essentially the “new” theory of liability that the trial judge wrongly permitted to be advanced in response to a jury question at the first trial; at the second trial, this theory was advanced in the alternative by the Crown from the outset.
[5] The appellant’s second ground of appeal from his conviction is that the pre-trial motion judge erred in refusing to exclude, on the basis of issue estoppel, evidence that was relevant to the original charge of first degree murder but had little or no probative value on the issues at the re-trial, namely evidence of computer searches including the word “autopsy”. He further submits that this evidence ought to have been excluded in any event because it was prejudicial.
[6] The third ground of appeal from conviction is that the trial judge gave the jury an inadequate instruction about the permissible use of out-of-court statements the appellant made after the event, if they found such statements to have been lies.
[7] With respect to his sentence, the appellant argues that the trial judge erred in sentencing him as though he had been convicted of murder, that the trial judge relied on an aggravating factor for which there was no evidence, and that the sentence was harsh and excessive.
[8] For the reasons that follow, I would dismiss the conviction and sentence appeals. As I will explain, there was an air of reality to the “criminal negligence by omission” theory of liability for manslaughter in this case, and it was properly left with the jury. The admissibility of the evidence of the internet searches about “autopsies” was not precluded by the doctrine of issue estoppel, and no objection was taken to its admission at trial. Nor do I accept that the trial judge erred in her instruction concerning the appellant’s statements, particularly where no concern about the wording was raised in the extensive pre-charge conference discussions, and the direction the appellant says ought to have been given would not have assisted him at trial. Finally, there is no basis to interfere with the sentence, which reveals no error in principle and was fit and reasonable in all the circumstances.
B. FACTS
[9] Ms. Grandine drowned in a bathtub on October 17, 2011, with lorazepam in her system. The appellant, her husband, was a registered nurse practitioner who was employed at a long-term care facility and retirement residence. He also served part-time as a pastor. At the time of her death, Ms. Grandine was twenty weeks pregnant with the couple’s first child.
[10] In the preceding months, Ms. Grandine had become aware that the appellant was having an extramarital affair with a parishioner from their church, Ms. Florentino. The entire parish was informed, the appellant had to resign from his position, and the couple had to leave the church, becoming members of another church. The pastor of that church provided weekly marriage counselling to the couple. As part of the counselling, the couple agreed to the installation of filtering software known as “K9” on their home computer, that was controlled by Ms. Grandine, in order to prevent the appellant from watching pornography, which was something that had been identified during the counselling sessions as a problem.
[11] Despite the appellant’s public agreement to do so, and his assertions to his wife that he had terminated his extramarital relationship, by October 2011 he was continuing the affair with Ms. Florentino.
[12] On the evening of October 13, 2011, the pastor met with the couple at their home. He testified that the appellant’s speech was slurred and that he fell asleep several times during the counselling session. Later that evening, the appellant had a lengthy phone call with Ms. Florentino. She testified that the appellant told her the counselling had finished early because his wife was tired and wanted to sleep, and that he asked to see her that night but she declined.
[13] The appellant did not testify at trial, however three of his statements to police were admitted into evidence. The appellant told police that he had to help his wife into bed on the evening of October 13 because she was very sleepy. He noticed that she had begun to vomit and she was complaining she was feeling unwell. At one point, he heard a noise and saw she had fallen out of bed and was on the floor. He had to “kind of carry her – lift her onto the bed”. The next day, October 14, he stayed home from work because he was concerned about his wife.
[14] That same day, the home computer was used through the appellant’s Facebook account. There were computer searches for sex-related websites. The appellant also communicated with Ms. Florentino by text message some 115 times between 2:17 p.m. and midnight.
[15] The appellant told police that his wife was still not waking up by the afternoon of October 14. He went out to get a haircut, and when he returned an hour later “she still seemed very lethargic”. He decided to take her to the hospital. He described his wife as disoriented. He had to help her get dressed and into the car.
[16] Ms. Grandine was admitted to the hospital the evening of October 14. Her mother and the appellant were at her bedside. Ms. Grandine’s mother testified that the appellant looked anxious to leave the hospital, and he was coming and going from the room. At one point, Ms. Grandine asked him, “did you give me a pill?”, or “did you give me something?”, and “why are you in such a hurry to go home? Is there someone waiting for you at home?” The appellant left the hospital around 2 a.m., and dropped off his mother-in-law at her home.
[17] A few hours later, at around 4:45 a.m. on October 15, and while Ms. Grandine was still in the hospital, there was computer activity on the appellant’s Facebook account on the home computer. The computer opened sex-related websites.
[18] The appellant had been in communication with Ms. Florentino while he was at the hospital, and he met up with her later that morning, going for a walk and for coffee and “making out” in her car.
[19] Ms. Grandine was discharged from the hospital at 1:45 p.m. on October 15. She and the appellant missed a family event that day, after she told her mother she had to rest. That evening, starting at around 6:40 p.m., there was a Facebook entry from the appellant’s account on the home computer. Seven minutes later, the computer searched for “100 milligrams Ativan”, Ativan being the trade name for lorazepam, and for the Wikipedia page for lorazepam, and a few minutes later it searched for sex-related websites.
[20] The next day, October 16, the appellant went to church and then he and Ms. Grandine attended a family gathering.
[21] Ms. Grandine was pronounced dead at 11:57 p.m. on October 17. The appellant told police that they had both gone to work that day and that he had picked up his wife from the subway station. He said that they had strawberry banana smoothies, but did not eat any other dinner, and that they had watched “The Good Wife” on the internet. Ms. Grandine said she felt tired, so they went upstairs, where they had sexual intercourse. Afterwards, Ms. Grandine told the appellant she was going to take a bath – something that the appellant and Ms. Grandine’s mother and sister described as unusual for her. The appellant told police that he had gone for a run, leaving their home around 9:45 p.m. and returning an hour later. He said that when he came home, he found his wife in the bathtub filled with water, with her mouth and nose frothing.
[22] At 10:06 p.m. that evening, following a request for a temporary password, the K9 filtering program was uninstalled from the home computer. The appellant had a five-second call with Ms. Florentino at 10:12 p.m., a roughly 26-minute call with her between 10:15 and 10:42 p.m., and a further four-minute call at 10:42 p.m.
[23] The appellant called 911 at 10:49 p.m. The call lasted about ten minutes. The dispatcher attempted to guide the appellant to lift his wife from the bathtub, however the appellant was unable to do so. In response to the call, members of the Toronto Fire Service were dispatched to the home, arriving around 10:56 p.m. The attending officer testified that he knocked at the front and side doors, but that the appellant did not answer the door until just after 11:00 p.m. EMS arrived shortly thereafter. Ms. Grandine was pronounced dead at the hospital.
[24] A forensic toxicologist analyzed a post-mortem blood sample and found lorazepam in the deceased’s blood. She testified that lorazepam, a benzodiazepine, is a central nervous system depressant. Its main effects are sedation but it can also cause reduced motor coordination and impaired cognition, and can mimic symptoms of intoxication. A blood sample that had been taken on Ms. Grandine’s earlier admission to hospital was also tested after her death, and was positive for lorazepam.
[25] A forensic pathologist testified about the cause of death, the mechanism of death and the mechanism of injury. She concluded that the cause of Ms. Grandine’s death was drowning with a head injury and/or presence of lorazepam as potential contributing factors. She could not rule out an accidental drowning or suicide.
[26] A forensic examination of a computer seized from the Grandine home revealed a number of search queries and site visits for where to buy lorazepam in Toronto without a prescription, whether overdoses of particular drugs such as Imovane or Risperdal would be identifiable in an autopsy, and the quantities at which lorazepam might become fatal. However, it could not be ascertained who conducted these searches.
[27] At the re-trial for manslaughter, it was the Crown’s theory that the appellant embarked on a plan to drug and incapacitate his wife in order to pursue his extramarital affair and to watch pornography without inhibition. According to the Crown, the appellant committed the unlawful acts of administering a noxious substance, trafficking in a controlled substance, or was criminally negligent in permitting his wife to enter the bath when he knew she was heavily sedated.
[28] The defence argued that the Crown had not proved its case beyond a reasonable doubt, and that there was insufficient evidence to conclude that the appellant had obtained lorazepam, administered or provided it to his wife, or was aware that she was heavily sedated on the evening of her death. The defence theory was that Ms. Grandine had voluntarily ingested the lorazepam and then either entered the bath and drowned by suicide, or had an accident in the bath, which led to her drowning.
C. CONVICTION APPEAL
(1) ISSUES
[29] The appellant raises three issues in his conviction appeal:
Did the trial judge err in leaving open a route of liability that had no evidentiary basis, that is, criminal negligence based on omission: that, knowing that his wife had taken lorazepam, the appellant had failed to provide her with the necessaries of life in permitting her to enter the bath?
Did the trial judge misdirect the jury as to the use that could be made of out-of-court statements made by the appellant after the events, if they concluded the statements were untrue?
Did the motion judge err in permitting the Crown to lead evidence respecting certain computer searches that was relevant only to intent for murder? Should the trial judge have excluded the evidence in any event?
[30] The appellant also contends that his sentence was excessive and that the trial judge erred in principle in referring to a breach of trust in relation to his having taken the drugs from his place of employment, in the absence of supporting evidence.
(2) PRE-TRIAL MOTION RE: ISSUE ESTOPPEL
[31] Three pre-trial motions in his case were heard by Code J., with reasons reported at 2018 ONSC 5646. The third motion, which is pertinent to this appeal, was based on issue estoppel. The defence moved to exclude various categories of evidence from the trial on the basis that such evidence was relevant only to whether the appellant had the intent to kill, and its admission could contradict the finding of lack of mens rea for murder that was implicit in the jury’s verdict in the first trial. Part of the evidence sought to be excluded was the internet research on the couple’s home computer about fatal amounts of lorazepam, and research concerning autopsies.
[32] The motion judge dismissed the issue estoppel motion. He concluded that all of the categories of evidence the defence sought to exclude were relevant to the manslaughter trial, including the evidence that someone had used the Grandines’ home computer to conduct research into the drug lorazepam (as well as other drugs), such as research as to fatal amounts of lorazepam and research concerning autopsies. Applying the principles from the leading case on issue estoppel, R. v. Mahalingan, 2008 SCC 63, [2008] 3 S.C.R. 316, he was satisfied that the Crown was seeking at the second trial to use the evidence that the defence sought to exclude “solely in relation to the three essential elements of manslaughter: the intentional unlawful act of surreptitiously drugging Ms. Grandine; causation of death by drowning due to intentionally placing her in the bath, assisting her into the bath, or standing by and doing nothing while she got into the bath in a dangerously disabled state; and objective foresight of the risk of bodily harm”: at para. 33. He observed that the jury at the subsequent trial could be given a limiting instruction, warning against misuse of the evidence in relation to issues decided in the accused’s favour at the first trial. He noted that the Crown proposed that the appellant’s motive to drug and incapacitate his wife was his desire to be free to pursue the extramarital affair and to view pornography, but if the Crown intended to allege a motive involving some future intent to kill, it would be necessary to raise the matter in advance with the court to seek a ruling as to whether this would violate issue estoppel.
(3) DISCUSSION
1) Did the trial judge err in leaving open a route of liability that had no evidentiary basis?
[33] The appellant submits that the trial judge committed a reversible error by leaving manslaughter by criminal negligence via omission as a route of liability in this case. The jury was invited to find the appellant guilty of manslaughter regardless of whether it concluded beyond a reasonable doubt that he had administered the lorazepam to the deceased. The appellant contends that this route of liability had no air of reality on the evidence: there was no evidence that the appellant knew the deceased had consumed lorazepam, and was, as a result, at such risk from taking a bath that it would have engaged his statutory duty under s. 215 of the Criminal Code to provide his spouse with the necessaries of life. The appellant points to the sentencing findings of the trial judge in which, in rejecting the premise that the deceased had voluntarily taken lorazepam, she also stated, “there is no evidence in this trial that the defendant knew that Ms. Grandine was taking or had taken lorazepam on the night of her death”.
[34] The Crown submits that there was an air of reality in this case to the omission route of liability – that, separate from whether the appellant administered or trafficked the drug, the appellant at least knew the deceased had ingested lorazepam, and that he left her to take a bath despite understanding its effects and knowing she was sedated. It was the defence theory at trial that Ms. Grandine had taken the lorazepam herself, and had drowned by suicide or had accidentally drowned after entering the bath while sedated. In addition to the evidence relied on by the defence to suggest that Ms. Grandine had taken lorazepam, the Crown points to various aspects of the evidence from which there were available inferences that the appellant knew Ms. Grandine had done so, and that he knew she was at risk based on his medical training and her condition days earlier, and had nevertheless failed to prevent her from entering the bath while sedated.
[35] The question is whether there was evidence that, if believed, could “reasonably support an inference of guilt” on the particular theory of liability: R. v. Al-Enzi, 2021 ONCA 81, 401 C.C.C. (3d) 277, at para. 150. A trial judge has a duty to charge a jury on any alternative route of liability that arises from the evidence when there is an air of reality to that route, and therefore some evidence on which a properly instructed jury, acting reasonably, could convict: R. v. Largie, 2010 ONCA 548, 101 O.R. (3d) 561, at para. 141, leave to appeal refused, [2010] S.C.C.A. No. 460 (Karl Largie), and [2011] S.C.C.A. No. 119 (Gavra Largie). An instruction on a theory of liability that does not have an air of reality will constitute reversible error: R. v. Ranger (2003), 67 O.R. (3d) 1 (C.A.), at para. 132.
[36] In this case, the trial judge instructed the jury that there were three possible routes of liability for manslaughter: first, that the appellant had caused Ms. Grandine’s death through the illegal act of administering lorazepam to her without her knowledge, thereby causing her death when she became incapacitated and drowned in the bathtub; second, that he had committed the illegal act of trafficking (which included administering) lorazepam, a controlled substance, to his wife, thereby causing her to become incapacitated and to die by drowning; or third, that, whether Ms. Grandine knowingly or unknowingly ingested the lorazepam, the appellant, knowing his wife was in a sedated state, and owing her a duty of care, failed to provide the necessaries of life to her by permitting her or assisting her in getting into the bath where she drowned.
[37] The trial judge instructed the jury that the Crown could prove the third route of liability either by an act, which required the Crown to prove that the appellant gave, offered or administered the lorazepam to his wife, or by an omission – that, knowing that his wife had taken lorazepam, the appellant left her to take a bath and in doing so failed to provide her with the necessaries of life.
[38] It is the criminal negligence by omission basis of liability that the appellant contends ought not to have been left with the jury. Unlike at the first trial, this theory of liability was squarely before the court from the outset of the second trial. The appellant’s focus in relation to this ground of appeal is not on whether or not there was evidence that the deceased took the lorazepam herself; at trial, the defence pointed to various aspects of the evidence that they said supported this conclusion. Rather, the appellant contends that there was no evidence that the appellant knew his wife had ingested lorazepam, or that she was sedated to the extent that it would have been dangerous to leave her alone in the bath.
[39] I would not give effect to this ground of appeal. As I will explain, there was an adequate evidentiary record to warrant leaving this route to liability for manslaughter with the jury.
[40] As a preliminary matter, the trial judge’s statement that there was “no evidence” that the appellant knew Ms. Grandine had taken lorazepam on the night of her death must be considered in context. The trial judge engaged in a detailed review of the evidence to make the necessary findings of fact for sentencing, pursuant to s. 724 of the Code. The defence had argued, for this purpose, that the jury convicted the appellant on the basis of the criminal negligence by omission route of liability – that is, that Ms. Grandine self-administered the lorazepam and the appellant knew she was sedated when he left the house while she was in the bath. The trial judge rejected this submission and, after methodically evaluating all of the evidence, concluded for the purpose of sentencing that the appellant administered the lorazepam to Ms. Grandine in the days before and on the night of her death.
[41] The issue raised by the appellant is whether this route to liability had an air of reality based on the evidence. As I will explain, there was evidence from which the jury could have found the appellant guilty on the criminal negligence by omission route, and the trial judge considered this evidence in making her findings of fact for sentencing. That the trial judge did not ultimately make findings for the purpose of sentencing, consistent with this route to liability does not mean that it had no air of reality. In fact, there was evidence that, if believed, could reasonably support an inference of guilt on the criminal negligence by omission theory of liability.
[42] First, there was evidence from which the jury might have concluded that Ms. Grandine self-administered the lorazepam. Although the Crown urged the jury to reject this conclusion, it was the defence theory that Ms. Grandine had taken the lorazepam, and had caused her own death intentionally or accidentally. In his closing submissions to the jury, defence counsel argued that there was “substantial evidence” to support this theory. He pointed to the evidence that suicide by drowning was consistent with the post-mortem examination; that Ms. Grandine’s discovery in August 2011 that her husband was having an affair with someone she knew and considered a friend was “life-altering” to her; and that in September, there were computer searches where, according to the defence, it was Ms. Grandine who was researching how to purchase lorazepam over the internet. At a counselling session on October 6, Ms. Grandine was suspicious the extramarital affair had resumed and she had confronted her husband with the accusation. Then on October 10, there was evidence that the Facebook account associated with Ms. Grandine searched “will 100 milligrams of Ativan be fatal?”. The defence asserted that it was Ms. Grandine who had conducted this search. The defence relied on all of this evidence to support the inference that Ms. Grandine had obtained and knowingly taken the lorazepam before she entered the bath and drowned.
[43] Second, I accept the Crown’s submission that there was evidence in this case to support the finding that the appellant knew that his wife had consumed lorazepam, and that he appreciated the risk if he were to leave his wife unattended in the bath. This evidence included computer searches conducted on October 15 and 17 on the couple’s home computer that demonstrated the appellant knew or suspected his wife had taken lorazepam, together with the appellant’s general knowledge of the effects of such drugs through his professional training and employment, and his specific knowledge of the risk to Ms. Grandine based on his observations of her condition a few days earlier.
[44] While the October 10 computer search was conducted using Ms. Grandine’s Facebook account, on each of October 15 (after Ms. Grandine had been released from hospital) and October 17 (the evening of her death), there were other searches respecting “Ativan” and “lorazepam”, where, based on their proximity in time to searches for sex-related websites and activity on the appellant’s own Facebook account, it would have been open to the jury and reasonable to infer that the searches were performed by the appellant. On October 15, at about 6:40 p.m., there was Facebook activity associated with the appellant’s account; seven minutes later, there was a Yahoo search for “100 milligrams of Ativan”, and less than two minutes after that, someone accessed Wikipedia using the term “lorazepam”. Less than five minutes after that, whoever was using the computer accessed sex-related websites. On October 17 at about 9:08 p.m., a sex-related website was accessed and then 24 minutes later, whoever was using the computer accessed the web page “Ativandrug.htm” at the site “rxlis.com”, and then less than four minutes later, accessed the web page “Ativan.html” at the site “halfdoctor.com”. If the jury reasonably concluded that it was the appellant who conducted the searches about Ativan on October 15 and 17, it was an available inference that he did so because he suspected, and then confirmed, that the drug was implicated in Ms. Grandine’s hospitalization and was affecting her on the night that she died.
[45] Finally, the evidence supported the conclusion that the appellant would have had both general and specific knowledge of his wife’s risks in consuming lorazepam on the evening of her death. He was a registered nurse practitioner working at a retirement residence where the evidence was that such drugs were routinely administered, such that he would have been familiar with their general side-effects. He had also had the experience only three days earlier of observing his wife in a state of sedation. He had described in detail to the police how she seemed very tired, and that she fell out of the bed and he had to lift her onto the bed. She was difficult to wake up and she seemed disoriented. He also told the police that he knew his wife was tired the night of her death, and that she did not routinely take baths, but told him she was having a bath. It was a reasonable inference that the appellant knew his wife had taken lorazepam on the earlier occasion and, knowing that she was exhibiting similar signs of sedation as three nights earlier, and having conducted searches, he knew or believed she had taken it again. In these circumstances, knowing the serious effects of the drug on his wife that had led to her hospitalization only days earlier, the appellant had a duty to intervene when he knew she was going to take a bath.
[46] It would have been open to the jury to conclude, based on the searches for information about lorazepam and their timing, together with the appellant’s experience and training in dealing with such drugs and his specific observations of his wife’s condition, that he knew she had taken lorazepam the night of her death and, appreciating the risk, had failed to prevent her from taking a bath and then went out leaving her alone in the bath. Whether or not the jury accepted some or all of the available inferences, the evidence was sufficient to meet the “air of reality” test in this case and to justify leaving this route to liability with the jury.
The adequacy of the jury charge
[47] At the hearing of the appeal, the appellant’s counsel advanced the further argument that the trial judge’s charge was insufficient because she failed to relate the evidence or to effectively explain to the jury how this theory of liability would apply on the evidence. In essence, the appellant pointed to the trial judge’s failure to relate specific pieces of the evidence to the criminal negligence by omission route of liability as indicating that there was, in fact, no evidence for the trial judge to relate.
[48] I do not accept this argument. I have already concluded that there was an air of reality to the criminal negligence by omission route of liability. The obligation of the trial judge, once a route to liability is in play, is to provide a legally adequate instruction with respect to its elements and to help the jury understand how the evidence relates to the elements. A trial judge has substantial discretion in the structure of the charge, so long as the jury has been “left with a sufficient understanding of the facts as they relate to the relevant issues” or “would adequately understand the issues involved, the law relating to the charge the accused is facing, and the evidence they should consider in resolving the issues”: R. v. Jacquard, [1997] 1 S.C.R. 314, at para. 14; R. v. Cooper, [1993] 1 S.C.R. 146, at p. 163; see also R. v. Daley, 2007 SCC 53, [2007] 3 S.C.R. 523, at paras. 27-31. I am satisfied that the jury was equipped in this case to properly consider this ground of liability.
[49] First, the trial judge provided a legally correct instruction. She directed the jury that, with respect to this theory of liability, they had to be satisfied beyond a reasonable doubt that, knowing that his wife had taken the lorazepam, the appellant left her to take a bath and, in doing so, failed to provide the necessaries of life to her. She then directed that the jury had to be satisfied beyond a reasonable doubt that the appellant showed a wanton or reckless disregard for the lives or safety of others, in that what he did or failed to do was a marked and substantial departure from what a reasonably prudent person would do in the same circumstances. The trial judge then broke down “marked and substantial departure”, instructing that Crown counsel must prove beyond a reasonable doubt either: (i) that the appellant was aware of an obvious and serious risk or danger to the lives or safety of others, but went ahead anyway despite his knowledge of that risk or danger; or (ii) that he gave no thought to the obvious and serious risk or danger to the lives or safety of others.
[50] Second, the trial judge conducted a careful review of the evidence in her charge. The Crown observes, and I agree, that the evidence in this case related to all of the three theories of liability. The live issue for the jury was what inferences should be drawn from the evidence. While the appellant is correct that the trial judge did not summarize the evidence relating to each of the elements immediately after setting out each route to liability, her review of the evidence together with her correct legal instructions, in my view, properly equipped the jury for their task.
[51] While not determinative, it is relevant and important that defence counsel did not object to this part of the charge. Although he made several submissions on the charge, defence counsel did not ask for the evidence to be directly related to each theory. As Strathy J.A. observed in R. v. Minor, 2013 ONCA 557, 303 C.C.C. (3d) 382, at para. 89, “counsel may have made a tactical decision not to request further instructions out of concern that in so doing the judge might re-emphasize damaging evidence”.
[52] For these reasons, I would dismiss this ground of appeal.
2) Did the trial judge misdirect the jury as to the use that could be made of out-of-court statements made by the appellant after the events, if they concluded the statements were not true?
[53] After Ms. Grandine’s death, the appellant made a number of statements to a variety of people, including on the 911 call, to the first responders, to the police and to the pastor. The trial judge instructed the jury on the use they could make of untrue statements uttered by the appellant. Referring to two examples from the evidence, she said as follows:
I want to discuss for a moment anything you find as a fact that the accused told anyone that you also find as a fact was untrue.
As one example, you will recall that the accused said that he went for a jog for approximately one hour, but said nothing to the officers about having spoken to Florentino for approximately 30 minutes of that time. Some of you may consider that by failing to tell police that he had been on the telephone for this period of time, he lied to them.
You may wish to consider whether the accused was being truthful about the efforts he made to get his wife out of the bathtub when he says he found her. You heard the 911 call.
If you decide that the accused lied to the police, or to anyone else, you must be careful how to use that evidence.
If you decide that the accused told a deliberate lie, then you are entitled to consider what inference, if any, you should draw from the fact that he lied. You should also consider that Mr. Grandine may have had no involvement with his wife’s death, but lied for one or more other reasons. It is for you to decide. Use your common sense.
In determining what inference, if any, you should draw from the accused having lied to the police, or to anyone else, you must consider the timing of his statements, the circumstances in which those statements were made, and any other evidence that might explain any lie you find as a fact that he told.
Finally, on the subject of the accused’s out of court statements, you may give anything you find Mr. Grandine said out of court as much or as little importance as you think it deserves in deciding this case. It is for you to say. Bear in mind, however, that anything you find he said is only part of the evidence in this case. Consider it along with and in the same way as all the other evidence.
[54] The appellant focuses on the two statements referred to as examples by the trial judge and makes three arguments, the first two of which were not addressed in his factum. He submits that the statements alleged to have been false had no inculpatory value in his manslaughter trial; that the trial judge erred in failing to instruct the jury about what inferences they could properly draw; and that the trial judge erred in failing to give a limiting instruction consistent with the instruction approved in this court’s decision in R. v. O’Connor (2002), 62 O.R. (3d) 263 (C.A.), at paras. 17-27.
[55] The appellant’s point of departure is a passage from O’Connor in which the court explains the potential relevance of an accused’s untrue statements. At para. 26, O’Connor A.C.J.O. explained:
False exculpatory statements made by a person upon being informed that a crime has been committed will in some circumstances be consistent with that person being conscious of having committed the crime, and may point to guilt in the same way that other after-the-fact conduct such as flight, a threat to a witness, or concealment of evidence can be probative of guilt. The circumstances in which a false statement is made may show an intent to mislead the police or others or an intent to deflect suspicion and may be evidence of a conscious mind that he or she has committed an offence…
[56] The appellant contends that the fact that he might have lied when he said he was out for a run had no probative value in this case. His failure to inform the police that he had been on the phone with Ms. Florentino did not necessarily demonstrate consciousness of guilt; rather, he may have wanted to conceal his conversation because he was embarrassed that he was pursuing his extramarital affair at the time of his wife’s death. As for the untrue statement about having tried to remove his wife from the bath, this was irrelevant in the second trial.
[57] This argument is misconceived. First, the appellant’s out-of-court statements were relevant. They were exculpatory and, if believed, could have assisted the appellant in his defence. The jury was given a modified R. v. W.(D.), [1991] 1 S.C.R. 742, instruction in respect of all of the appellant’s statements. To the extent that a statement was not believed, it was open to the jury to consider, if there was independent evidence of fabrication, whether the appellant lied in order to distance himself or otherwise to deflect responsibility for the offence. The question is not whether some other inference could have been drawn from the appellant’s lie. It was up to the jury to decide what to conclude from the evidence. The trial judge properly instructed the jury, if they found the appellant had lied, to consider other possible explanations for the lie.
[58] The real question here is whether the jury was properly instructed about the use of such evidence. This leads to the appellant’s second argument. The appellant contends that the trial judge’s instruction was deficient. He says that when the trial judge told the jury that they were “entitled to consider what inference, if any, [they] should draw from the fact that he lied”, there was a risk of moral reasoning prejudice: the jury might well have reasoned that, if the appellant was lying about going for a run (and not disclosing his call to Ms. Florentino) and about having tried to help his wife, they might have inferred guilt from this discreditable conduct. The appellant submits that the trial judge ought to have circumscribed the jury’s use of this evidence (presumably by a specific instruction in line with the instruction in O’Connor).
[59] There is no merit to this argument. The jury was cautioned repeatedly throughout the charge about the permissible use of prior discreditable conduct evidence, including the appellant’s phone calls with Ms. Florentino the night Ms. Grandine died. A specific caution about discreditable conduct immediately preceded the trial judge’s instructions on the appellant’s out-of-court statements. I see no risk from the trial judge’s instructions that the jury, having been cautioned against such reasoning, would have concluded that the appellant’s lies, showing discreditable conduct, established his guilt.
[60] The appellant’s third argument is that the trial judge erred in failing to give the necessary cautions provided for in O’Connor. The appellant submits that the jury ought to have been instructed that where he made out-of-court statements that they concluded were lies, they could not use such evidence to convict him unless they were also satisfied on the basis of independent evidence that he had lied to conceal his participation in the offence. Absent this instruction, the jury may well have jumped directly from finding that the appellant lied to concluding that he was guilty of the offence, the very prejudice this instruction is designed to guard against.
[61] “The question on appeal … is not whether an O’Connor instruction would have been appropriate, but whether the instruction given prejudiced the appellant’s right to a fair trial”: R. v. Polimac, 2010 ONCA 346, 254 C.C.C. (3d) 359, at paras. 85, 106, leave to appeal refused, [2010] S.C.C.A. No. 263. I am satisfied that the appellant’s right to a fair trial was not prejudiced by the instruction given in this case.
[62] First, the fact that the trial judge did not specifically tell the jury that they could conclude that the appellant’s statements, if false, were evidence of guilt, and did not point out the evidence they could consider in determining whether he had fabricated the statements, did not prejudice the appellant. There was certainly a basis for the jury to use the fact of the appellant’s fabrication as affirmative evidence of guilt in this case. “Independent” evidence of fabrication or concoction simply means evidence other than the evidence that contradicts or discredits the version advanced by the accused. The circumstances in which the statement is made, its timing and content, including the degree of detail provided, may provide independent evidence of fabrication. Such evidence existed in the timing of the statements when the appellant was not a suspect, and the specific details he provided. The failure to point out circumstances supporting a finding of fabrication may enure to the benefit of the accused: Polimac, at para. 106. It would not have helped the appellant to have these details outlined for the jury.
[63] A similar argument was made in R. v. Selvanayagam, 2011 ONCA 602, 281 C.C.C. (3d) 3. As in this case, the trial judge did not instruct the jury to consider the appellant’s exculpatory statement, if disbelieved, as positive evidence of guilt. Nor was the jury cautioned that they could not use disbelief of the accused as evidence of guilt unless they found independent evidence that he had fabricated his statements. Feldman J.A. concluded that there was evidence that was independent of the falsity of the impugned statements that the jury could have considered on the issue of concoction had they been so instructed, and that because this evidence of possible concoction existed, it would not have helped the appellant to have it outlined to the jury by the trial judge. Such an instruction would have called to the jury’s attention the evidence of concoction, and the conclusion that it constituted evidence of consciousness of guilt: at para. 33.
[64] Additionally, no objection was raised by defence counsel despite lengthy pre-charge discussions. The failure of counsel at trial to raise any concern about how this issue was addressed, after having been provided with a draft charge and having engaged in a detailed pre-charge conference discussion, is important: Polimac, at para. 96. In this case, as in Polimac, “[t]he issue arising out of the alleged falsehoods told by the appellant was left to the jury as counsel wanted it left to the jury”, and “[a]n instruction replicating the O’Connor analysis would not … have improved the position of the defence”: at para. 106.
3) Did the court err in permitting the Crown to lead evidence respecting certain computer searches that was relevant only to intent for murder?
[65] This ground of appeal began as an attack on the motion judge’s decision on the appellant’s pre-trial motion based on issue estoppel. Recall that the defence had moved to exclude a number of categories of evidence from the second trial, on the basis that such evidence was relevant only to the charge of murder, on which the appellant had been acquitted. The defence argued that the introduction of such evidence at the second trial was barred by the principle of issue estoppel as this evidence had no relevance to the live issues at the manslaughter trial.
[66] During oral argument, this ground of appeal was expanded to include a second submission: that the trial judge ought to have excluded the impugned evidence on the basis that its prejudicial effect outweighed its probative value.
[67] At trial, Sgt. Manoharan testified that he found data in “unallocated space” on the computer hard drive that potentially indicated web searches for “death by insulin overdose, autopsy homicide”, “risperdal autopsy”, “Imovane overdose, autopsy” and “autopsy toxicology lorazepam”.
[68] The appellant does not assert that the motion judge identified and applied the wrong test in dismissing the issue estoppel motion; he accepts that the admissibility of the various categories of evidence he sought to exclude at the second trial on the basis of issue estoppel was governed by the principles set out in Mahalingan. Rather, the appellant asserts that this evidence was only logically relevant to an intent to kill, and not to any live issue in his manslaughter trial. If the jury accepted that the appellant had conducted these searches, the only reasonable inference would be that he was contemplating using the drugs to cause death. Such an inference would necessarily contradict the implied finding of the jury at the first trial that the appellant did not have the intent to kill. The appellant also submits (for the first time on appeal) that the evidence ought to have been excluded as prejudicial; the evidence could have been used by the jury to conclude that the appellant was a bad person, and therefore guilty. The evidence was also minimally probative, as the Crown had ample evidence of other searches for lorazepam that did not include the prejudicial term “autopsy”.
[69] I would not give effect to this ground of appeal. The central question was whether the evidence in question was relevant to an issue in the manslaughter trial. I disagree with the appellant’s submission that this evidence was only logically relevant to an intent to kill, and not to any other live issue. The autopsy searches were only a portion of the computer searches. And from many of the searches, according to the Crown, the inference was that the appellant was trying to find out what a therapeutic dose of lorazepam would be, or described another way, what a non-fatal dose would be. His intent was to incapacitate, not to kill, his wife. Autopsy searches were another way in which to pinpoint what a fatal dose would be, and to avoid administering such a dose. The searches also provided some evidence that the appellant knew and understood the impact of the drug, such that when he left his wife alone to take a bath on the night in question, tired and with lorazepam in her system, there was objective foresight of the risk of bodily harm.
[70] I see no error in the motion judge’s conclusion that the computer search evidence, including the specific evidence focussed on in this appeal, was relevant – in the sense that its admission was not precluded on the basis of issue estoppel – for the second trial.
[71] The real concern articulated on appeal was that the evidence of computer searches including the term “autopsy” was prejudicial because it invited improper reasoning by the jury, raising the spectre that the appellant was in fact intending to kill his wife. The appellant asserts that there was other evidence available, including searches for non-fatal doses of Ativan, that supported the Crown’s theory that the appellant was looking to incapacitate, but not to kill, his wife. He contends that the trial judge erred in failing to exclude the prejudicial evidence from the trial.
[72] The problem with this argument is that no such request was made below. Having brought a pre-trial motion with the specific aim of excluding this evidence, the appellant did not ask the motion judge to rule on a request that the evidence be excluded as prejudicial. Nor did the motion judge’s decision addressing the issue estoppel question end the matter of the admissibility of this evidence. It remained open to the appellant at trial to object to the admission of the evidence he now says ought not to have been admitted, including on the basis that its prejudicial effect outweighed its probative value. He did not do so. I do not accept the argument that the trial judge ought to have exercised her discretion to exclude this evidence based on its prejudicial effect in the absence of any objection by the defence at trial. The appellant had not one, but two chances to make this point below. He did not raise the issue at either opportunity, which “speaks volumes” about whether, in the context of this trial, this evidence was prejudicial and ought to have been excluded: R. v. Staples, 2022 ONCA 266, at para. 41.
D. SENTENCE APPEAL
[73] Because there were various routes to conviction for manslaughter, the trial judge was required to make findings of fact for sentencing pursuant to s. 724 of the Code. For reasons reported at 2019 ONSC 6129, she concluded that the appellant had administered the lorazepam to his wife the night of her death, as well as a few days earlier, intending to incapacitate her. He had a hostile animus towards his wife and a motive to incapacitate her. He was unhappy about the discovery of his extramarital affair since he was forced to resign in disgrace from his position at his church. Despite the insistence by his pastor and the deceased that he stop his affair, he had not done so. He was enamoured with both Ms. Florentino and with pornography. The trial judge concluded that, “On occasions in which his wife was incapacitated, the way for him to pursue both these fascinations were cleared and unimpeded.” The trial judge also concluded that, while the amount the deceased had ingested was within the therapeutic range, the effect of the drug was that if she was not unconscious around the time that she drowned, she was at the very least sedated. The appellant had not gone out for a run. Instead, he had contacted the computer filter company from the home computer for a temporary password at 10:06 p.m. and uninstalled the program on the night of his wife’s death. The appellant called 911 without draining the water from the bathtub or lifting his wife’s head out of the water.
[74] The appellant makes three submissions in relation to his sentence. First, he submits that, in imposing a custodial sentence of 15 years, the trial judge erred by mischaracterizing the nature of the offence as one of “almost murder”. Put differently, the appellant argues that the trial judge erred in finding that there was no air of reality to the suggestion that the offences were “anything like an accident or close to it”, and in aggravating the appellant’s sentence on this basis.
[75] Second, the appellant submits that the trial judge erred in principle in referring to a breach of trust in relation to his having taken the drugs from his place of employment, in the absence of supporting evidence. Finally, he submits that the trial judge erred in imposing a sentence that was demonstrably unfit in light of other sentences for similar offences, in particular in comparison with the sentence received by the offender in R. v. Roncaioli, 2011 ONCA 378, 271 C.C.C. (3d) 385.
[76] The Crown contends that the sentence imposed was fit. In crafting an appropriate sentence, the trial judge was required to consider the appellant’s moral blameworthiness. Manslaughter captures a wide range of conduct, from almost accident to almost murder, and as a result, a wide range of sentences have been imposed for this offence. In categorizing the appellant’s conduct, it was reasonable for the trial judge to find that, “in terms of moral culpability, [the appellant’s behaviour fell] closer to almost murder than to anything on the other end of the spectrum of manslaughter cases.” In making this statement, the trial judge explained that this case did not involve a momentary action that resulted in death and immediate remorse. Instead, the appellant prepared for the offence by undertaking computer searches, testing the drug on himself and the deceased beforehand, and administering the drug. These are reasonable findings entitled to deference on appeal.
[77] The standard of review of sentence is deferential. Except where a sentencing judge makes an error of law or principle that has an impact on the sentence, an appellate court may not vary the sentence unless it is demonstrably unfit: R. v. Lacasse, 2015 SCC 64, [2015] 3 S.C.R. 1089, at para. 58; R. v. Friesen, 2020 SCC 9, 391 C.C.C. (3d) 309, at paras. 25-26. The appellant has not demonstrated any reversible error in his sentence, nor in all the circumstances is the sentence unfit.
[78] First, the trial judge did not sentence the appellant as though he had committed an intentional killing. “[M]anslaughter captures a wide and disparate range of conduct… As a result, a wide ‘range’ of sentences have been imposed for this offence”: R. v. Alexander, 2014 ONCA 22, 315 O.A.C. 49, at para. 41. The trial judge was required to fashion a sentence that was fit for the offence and the offender.
[79] The Crown had proposed a custodial sentence of 13 to 15 years, while the defence position was 5 to 7 years, mainly on the basis that the offence did not involve extreme physical violence. The trial judge, referring to her findings of fact on sentencing, concluded at paras. 48-51 that while the appellant did not have a history of physical violence, this was a violent offence, involving deception and premeditation:
In committing this offence, he did perpetrate a form of emotional and physical violence on his wife… His behaviour was diabolical and violent. It was planned and premeditated and took place not just on the night of her death, but days before when he experimented with the lorazepam on her and on himself to see what the effects of the drug would be… He knew the first night that she was ill and groggy as a result of the drug. He knew then that she was affected by the drug seriously enough to go to the hospital. He knew on the night Ms. Grandine died that her symptoms were the same as on the day he had tested the drug on her. Just before her death, he allowed her to enter the bath waters knowing she was seriously affected by the drug. He did it all to communicate with his lover and use pornography that night.
This was an assault – and act of violence generated directly from the anger Mr. Grandine bore towards Mrs. Grandine. The assault was physically innocuous. Nonetheless, it had fatal consequences when it was combined with a tub full of water…
The defendant’s actions after he returned to the home and found his wife dead were callous and an attempt to cover up what he had done…
[80] It was in the context of these remarks that the trial judge noted, at para. 51, that the case did not have strong mitigating factors and that “in terms of moral culpability, Mr. Grandine’s behaviour falls closer to almost murder than to anything on the other end of the spectrum of manslaughter cases”. In making these comments, the trial judge was not sentencing the appellant for “almost murder”; rather, she was contrasting the circumstances of this case with other manslaughter cases relied on by the defence, where the moral culpability was less, including other domestic homicide cases involving “drugging on the sudden” and an offender’s immediate remorse. The trial judge found that there were numerous aggravating factors, including the appellant’s motivations for the offence (that he “wanted the freedom to have sex with his girlfriend and watch pornography”), the vulnerability of the victim and her pregnancy, the fact that the attack was planned and premeditated, and the appellant’s callous actions before and after her death. All of these factors were relevant to the appellant’s moral culpability, and the trial judge reasonably concluded that the conduct here, together with the few mitigating factors, warranted a sentence at the higher end of the range.
[81] The appellant’s second complaint is that the trial judge identified the appellant’s breach of trust in having taken the lorazepam from the retirement residence and patients he was employed to serve, when there was no proof beyond a reasonable doubt that he had taken the lorazepam from his workplace and no evidence any patient had been affected.
[82] I disagree. The trial judge made a finding of fact for sentencing that the appellant conducted online searches for where to acquire lorazepam and then took the lorazepam from his workplace. There was significant circumstantial evidence, outlined at paras. 68-73 of the trial judge’s reasons, from which the inference could be drawn that the appellant took lorazepam from his workplace. This included the evidence that the appellant had access to the drug at his workplace and was responsible for receiving and logging prescription drugs, including lorazepam, securing and dispensing them, and then disposing of unused or expired drugs. The trial judge noted that there was no evidence that he had purchased lorazepam illegally.
[83] From this evidence, the trial judge drew the reasonable inference that the appellant took the lorazepam from his workplace, and this supported a finding of a breach of trust by the appellant to his employer and patients. Even if, as the appellant asserts, this would not amount to a breach of trust by the appellant to his employer and patients, this was one of many aggravating factors identified by the trial judge, and in the context of this case, is not significant. The trial judge also identified, as an aggravating factor, that the appellant’s actions were “a breach of trust expected in any domestic relationship”. Whether or not there was an additional breach of trust in relation to the procurement of the lorazepam is of little relevance to the appellant’s moral culpability in respect of the manslaughter death of his wife. In any event, the incorrect identification of a factor as an aggravating factor constitutes a reversible error on sentencing only if the error had an actual impact on the sentence: Lacasse, at paras. 44, 83. The absence of the factor challenged by the appellant would not have materially affected the sentence in this case.
[84] The appellant’s final argument is that the sentence was unfit, particularly having regard to the sentence imposed on the offender in Roncaioli. He repeats the argument made at first instance that the circumstances of that case are on all fours with the present case and that a similar sentence ought to have been imposed.
[85] In Roncaioli, the offender received a sentence of seven years’ imprisonment for manslaughter. The sentence was upheld by this court on the offender’s appeal. The Crown did not appeal the sentence. The offender, a doctor, had injected his wife with toxic amounts of anesthetics which, in conjunction with her consumption of alcohol, was a significant cause of her death.
[86] I agree with the trial judge that Roncaioli is distinguishable. She distinguished this case on the basis that the offender in Roncaioli was 67 years old at the time of the offence, 72 years old at the date of the trial, and 75 when he appealed. She also noted that the court considered the doctor’s long and successful career prior to his wife’s death, and that the deceased had been using alcohol and drugs before her death. The victim was also not pregnant at the time of her death.
[87] The appellant has not demonstrated any error in the sentence, or that the sentence was unfit.
DISPOSITION
[88] For these reasons, I would dismiss the conviction appeal and I would grant leave to appeal sentence, but dismiss the sentence appeal.
Released: May 9, 2022 “K.F.” “K. van Rensburg J.A.” “I agree. K. Feldman J.A.” “I agree. Coroza J.A.”