Davis v R

Court of Criminal Appeal (NSW)|2018-10-29|Before: Hoeben CJ, Harrison J, Schmidt J, Hulme J

(a) Admitting evidence for a coincidence purpose; and

(b) Using the coincidence evidence in a process of "backwards reasoning" thereby reversing the onus of proof.

Ground 2 - The verdicts are unreasonable and cannot be supported by the evidence.

GENERAL BACKGROUND AND CROWN CASE

Over two consecutive days in October 2013, three elderly residents from the same ward of an aged care facility in Newcastle were each found unresponsive in a seriously hypoglycaemic state requiring urgent medical attention. It was subsequently established that each had been injected with high doses of insulin in circumstances where there was no medical need for such treatment.
Two of the residents, Ms Gwendolyne Fowler, aged 83, and Mr Ryan Kelly, aged 80, died as a result of the injection. The third resident, Ms Audrey Manuel, aged 91, remained in hospital for a significant period of time. She never regained the same level of functioning as she had enjoyed before the injection. Ms Manuel died from unrelated causes in January 2015.
The Crown case was that it could readily be inferred that whoever injected the victims with the insulin did so deliberately with an intention to kill. The Crown further alleged that the same person administered the insulin to each of the three victims and that that person was the applicant.
As at October 2013, the applicant was aged 26 and was employed as a Team Leader Aged Care Worker at the SummitCare aged care facility at Wallsend. He worked the morning shift from 6.30am until 3pm on both 18 and 19 October 2013.
The Crown case was that the applicant had injected each of Ms Fowler, Ms Manuel and Mr Kelly with insulin. Ms Fowler was injected on 18 October 2013. At the time there was not thought to be any reason for suspicion and she passed away the following day at about 12.50pm. Ms Manuel and Mr Kelly were given injections on 19 October 2013. Mr Kelly died on 29 October 2013.
At trial, there was no dispute that each of the victims was injected with insulin when there was no medical need to do so. The focus of the trial was whether it was the applicant who was responsible for giving the injections. There was no direct evidence that the applicant was the person who gave any of the three insulin injections. The Crown case against the applicant was circumstantial.
The circumstances relied upon by the Crown were that the similarities between the events themselves and the surrounding circumstances exhibited a pattern of conduct from which an inference could be drawn that a single perpetrator had been responsible for all of the offences. It followed, so the Crown submitted, that this person had to have been present at the SummitCare facility on both 18 and 19 October in order to carry out these acts, and within certain timeframes said to be established by the expert evidence.
It was the Crown case that medical evidence established that each injection would have been administered no more than eight hours before each victim was found in a significant hypoglycaemic state. Using this information, CCTV footage and other records kept by the facility, it was the Crown case that police were able to determine that from a total of 319 people seen entering and exiting the premises, 25 people including the applicant were present on each of the days during the relevant time periods when the injections were administered.
The 25 people consisted of 20 staff members (including the applicant) and five visitors. The evidence was that many of the relevant staff members did not work in the ward where the three victims resided.
Other than the applicant, and one staff member who had died in the interim, the Crown called each of the remaining 23 people to give evidence.
In addition to having the opportunity to commit the offences, the circumstantial evidence against the applicant which the Crown relied upon to establish that he was the person who injected all three doses of insulin, included the following:

The fact that the applicant was skilled and experienced in the administration of insulin injections;
As a Team Leader, the applicant had a key to access treatment rooms at the aged facility in which quantities of insulin were stored (the only other people who had such keys, in addition to Team Leaders, were the registered nurses);
Evidence that each of the victims was familiar with the applicant whose responsibilities included administering oral medication to all of the residents in the Mountview ward (in the context of there being no suggestion that any of the victims had remonstrated or protested about the injections being given);
When a search warrant was executed at the applicant's home, needles, syringes and information about insulin were located (including information about the time that particular types of insulin take to achieve a hypoglycaemic effect);
Text messages which were sent by the applicant to colleagues which suggested foreknowledge by him of the impending or prospective demise of Ms Manuel and Mr Kelly;

The judgment (R v Davis [2016] NSWSC 1362)

In relation to the two charges of murder, the trial judge identified as essential facts to be proved by the Crown that a deliberate act of the applicant caused the deaths of the deceased and that such act was carried out with an intention to either kill or cause really serious bodily harm. His Honour found that the deliberate act relied upon by the Crown was the injection of insulin and the intention asserted by the Crown was that the applicant intended to cause death. In relation to the charge of administering a poison with intent, the trial judge identified the necessity for the Crown to prove beyond reasonable doubt that the applicant deliberately injected insulin into the victim and that he did so with an intention of killing the victim.
The trial judge further defined the issues in dispute as follows: "8 There is no dispute that each of the victims were injected with insulin and that this was done with an intention to kill. There is no dispute that where there was no medical need for insulin to be injected, and therefore the consequences would be injurious to life or health, that it would qualify as a "poison". The dispute is focussed upon the issue of who did it. The Crown, of course, says it was the accused in each of the three cases. The accused says it was not him; it must have been someone else. 9 While the Crown says that the same person administered insulin to each of the three victims, the accused says that it may have been more than one person who did so. The accused, of course, does not have to prove anything. He is presumed to be innocent. It is a matter for the Crown to prove what it asserts."
His Honour made a number of findings as to how the aged care facility was run, its staffing and what provision was made for security and the control of visitors. None of this material was controversial, and none of his Honour's findings on these issues was challenged on appeal. The findings were: "SummitCare Wallsend 14 SummitCare Wallsend is within a three-level building in Bent Street, Wallsend. It provides beds for 142 residents. As at 17 October 2013 it was almost full; there were only three vacant beds. 15 There are two wards on the ground level, Magnolia and Lavender, and two on the first floor, Mountview and Parkview. There are north and south wings to each ward. Each of Mountview, Parkview and Lavender wards had 35 beds and Magnolia ward had 37 beds. 16 The main public entrance is on the ground floor. On the lower ground floor there is a staff entrance, staff room, kitchen, laundry and other facilities accessible only by staff. 17 Twin elevators serviced each floor and there were fire stairs running between the floors. There were also some external stairs although the evidence about them was not very descriptive and they were not shown on any of the plans. Staffing 18 Staff at SummitCare worked in three shifts. The morning shift was from 6.30am until 3.00pm. The afternoon shift was from 2.30pm to 11.00pm. The night shift was from 10.30pm until 7.00am. 19 During the morning and afternoon shifts there was a registered nurse ("RN") working on each floor; that is, one in the Mountview and Parkview wards on the first floor and another working in the Magnolia and Lavender wards on the ground floor. A team leader ("TL") and four assistants in nursing ("AINs") were assigned to each ward. Generally, two AINs worked in each wing of a ward but they would provide assistance in the other wing if required. Ms Carolyn Tranter, the operations manager at SummitCare at the relevant time, spoke of some AINs working "short shifts" in the morning and afternoon. 20 On the night shift there was one RN for the entire facility and two AINs for each ward. 21 At the end of each shift there was a handover of keys and a briefing as to events that had occurred in the ward by the outgoing to the incoming registered nurses and team leaders. 22 SummitCare also employed casual staff and staff provided by nursing agencies when it was necessary to fulfil unplanned rostering requirements. 23 Another staffing issue to note is that new AINs were required to work with a more experienced AIN (a "buddy") to assist their familiarisation with the facility and its residents. Security Staff 24 There was a keypad security device at the staff entrance. All staff had the same generic code to operate it to obtain access to the building ("the staff code"). 25 In the staff tea room there was a fingerprint scanning device that was used to record times of entry and exit of permanent and casual, but not agency, staff for payroll purposes. 26 The staff code was needed to access the fire stairs and to access the lower ground floor by elevator. Visitors 27 There was a keypad security device at the main public entrance on the ground floor. There was a generic code common to all visitors ("the visitors' code"). This was made available to relatives and friends upon a resident being admitted to the facility and it was changed quarterly. There was conflicting evidence given by witnesses who were visitors as to whether it was necessary to use the code to enter the front door during office hours but it seems likely that it was not. 28 There was a requirement that visitors, including tradesmen and contractors, were to sign a visitors' book at reception on the ground floor. But reception was not staffed outside of office hours (8.30am to 5.00pm) or on weekends and sometimes even during office hours reception staff might be temporarily absent. Signing in and out in the visitors' book seems to have been required but not strictly enforced. 29 If a person wishing to enter the facility did not have the code to enter on the keypad at the front entrance they could press a buzzer which would be answered by staff in the Parkview wing who could see on a video monitor the person who wanted to enter. The front door could be unlocked remotely. Ms Tranter said that staff would usually ask who was buzzing and who they wanted to visit and they would be asked to sign the visitors' book. However, she agreed that "who's buzzing to come into the front door may well be a secondary concern" when staff were busy with more urgent tasks. She also agreed that there was no searching or checking bags of visitors. (T69-70) 30 Contractors and delivery drivers were not provided with a code, even if they attended regularly. Staff permitted them access as needed on each occasion they attended. The same applied in respect of garbage collection from the waste room on the lower ground floor. 31 There was also a keypad security device at the entrance to each of the wards. The generic staff and visitors' codes were used. There was some inconsistency in the evidence as to when this requirement was put in place but I am satisfied that it was in place before 18 October 2013: see, for example, the evidence of Ms Diane Dean (T140) and Ms Julie Ross (T599-602) and the accused's statement of 24 October 2013 (Exhibit Y at [11]). 32 A similar keypad security device was at the entry points to the fire stairwells. The staff code was required to access them. There was evidence that some staff used the fire stairs as a shortcut to travel between the wards on the ground and first floors but there was no evidence of residents or visitors doing so. 33 In each of the wards there was a "back of house" area that was secured by keypad. There was a different code, albeit generic, used by staff to access this area. Access by residents and visitors was not permitted. 34 There was some evidence that non-staff had been able to use the staff code. Ms Sutherland, who used to visit her mother in the Mountview ward, said that there was an occasion when her brother gave her what was thought to be the visitors' code but she was informed by a staff member ("Judy") that it was a staff code and that she was not to use it. (T369) RN Esta Luo said that she had seen visitors using a staff code. She agreed in cross-examination that this was "common practice". (T246) When she was asked how she knew this, she said "sometimes I captured them actually putting the staff code". (T247) In re-examination she was asked what she meant by "sometimes" and she referred to a single occasion when she had seen this. On that occasion, the person was recognised by her as being a known visitor to a resident of the facility; somebody who would have been entitled to enter with the use of the visitors' code in any event. (T253) I gained the distinct impression that Ms Luo's initial evidence on the subject overstated the frequency with which this occurred. I accept that it may have happened, but not that often. In the end, I do not believe there is anything significant in this issue. 35 There was a "treatment room" within each of the wards and a key was required for entry. The registered nurse and the team leader held the keys and managers had a master key. A key to a treatment room in one ward could be used to access the treatment rooms in the other wards. The registered nurses also (and only) had keys to the secure cabinet within the treatment rooms in which restricted (S4 and S8) drugs were kept. Everything else in the treatment room, including residents' insulin stored in a refrigerator, was otherwise readily accessible. 36 A key register was maintained. The keys were passed between corresponding staff members in those positions and, when not otherwise in use, were kept in a locked drawer in Parkview. 37 There was an issue raised with most of the staff members who gave evidence as to whether treatment rooms were always closed and locked when not being accessed by staff. No-one admitted to leaving a treatment room open and unattended but a number of witnesses spoke of having seen it. I accept that it occurred from time to time although it does appear to have been a situation that was occasional rather than regular or frequent. There was no evidence that any treatment room was left unattended on 18 or 19 October 2013. Closed-circuit television (CCTV) cameras 38 There were five CCTV cameras installed and working at the relevant time. (There were two more but they were not working.) There was one over the front public entrance to the facility and four on the lower ground floor: one covering the staff entrance on Raglan Street; one at the deliveries area; and two covering the corridor leading from the staff meal room area and around to the elevators. In short, the CCTV cameras covered all of the possible entry points to the building and particularly to the ground and first floors where the wards were. (T81-2) No evidence or suggestion of unauthorised access 39 Finally, in relation to security at SummitCare, it is of significance to note there was no evidence that there were ever: (a) people on the premises who did not have a legitimate reason for being there; (b) people who had a legitimate purpose in visiting certain places within the premises but were found to be in other places; (c) staff who had duties to perform in certain places but were in other places without legitimate reason; (d) people who were not members of staff who entered a treatment room; and (e) people who were staff but not registered nurses or team leaders, or people who were not staff members at all, who were involved in giving any form of medication to patients, whether it was required medication or otherwise. Availability and use of insulin 40 Blood glucose levels were regularly checked for all of the insulin-dependent diabetic residents. According to Dr Tuan Minh Quach (whose evidence will be discussed later), the normal blood glucose level range is considered to be between 4 and 7 mmol/L (millimole per litre). Any level below 4 mmol/L is classified as being hypoglycaemic. 41 Insulin was not accountable so no records were maintained as to usage (aside from patient medication charts) or of stock on hand. It was kept in a non-secure refrigerator in the treatment rooms. 42 In October 2013 it was the policy at SummitCare that only registered nurses could administer injections, including insulin, to residents. A number of witnesses attested to this and the fact that it was previously the case that team leaders could do so as well. The change in policy came about in early 2013. 43 The protocol for insulin administration in place as at 18-19 October 2013 was that a registered nurse would give the injection in the presence of another member of the nursing staff (usually a team leader). They would each identify the patient and verify the correct insulin and dosage and the patient's medication chart would be initialled by both in confirmation. 44 Notwithstanding the change in policy, the accused continued to administer insulin to patients. RN Stephen Zhao said that the accused and one or two other team leaders continued doing so for a week or two after the policy change. SummitCare management issued a further direction that injections were only to be carried out by registered nurses but the accused and a couple of other team leaders persisted, mainly when the registered nurses were busy and they felt they could help out. (T556) 45 Mr Zhao said he had the experience on a few occasions when he went to administer insulin to patients in Mountview, only to be told by the accused that he had taken it upon himself to do so already. (T557) 46 An extract from the medication chart for a particular insulin-dependent diabetic resident of Mountview ward became Exhibit J. There were a number of entries in the week of 16-22 October 2013 where only one person had initialled in relation to the administration of insulin. RN Maria Zubcakova was shown this extract and gave evidence that on 17 October the 7.30am administration had been initialled by the accused and not by her. When she was asked about this by police almost a year later she had no precise recollection but considered it was likely that the accused had administered the insulin on her behalf. She had been busy and the accused had done it in her absence and told her later. She considered the accused was being helpful; she described him as a reliable person who could do it. (T439-440; 443; 447) In cross-examination and despite her other evidence to the contrary, she agreed that this was probably something the accused did after having consulted her. (T444) 47 RN Paulette Hills said that three or four weeks before 18 October the accused had told her that she did not have to worry about giving insulin to a resident of Mountview because he had done so already. She thanked him but cautioned that he was not permitted to do so. (T476)"
The evidence concerning the three victims' state of health, both before and after the administration of the insulin, was also not controversial. An understanding of that material, however, is necessary for an understanding of the issues raised in the appeal, in particular Ground 2. Because the trial judge's findings on that issue were not controversial and not challenged in this appeal, they also can be set out. These unchallenged factual findings were: "Ms Gwendolyne Fowler 48 Ms Gwendolyne Fowler was aged 83. She had been married for 55 years when her husband passed away in 2005. As far as I am aware she had three daughters. Ms Fowler suffered from Alzheimer's disease, diabetes mellitus type 2, rheumatoid arthritis and osteoporosis. She was 157cm tall and weighed 43kg. She was a resident of Room 8 in the south wing of the Mountview ward. 49 Ms Fowler was under the care of Dr Lynn Gay, general practitioner. Dr Gay took Ms Fowler off one of her medications (Diamicron 60mg) on 6 September 2013 and asked that nursing staff at SummitCare test her blood glucose levels (BGLs) daily for two weeks. On 19 September Dr Gay found the daily BGL readings (5.2 to 6.5 mmol/L) were within an acceptable range and nursing staff could revert to weekly testing. These readings (6.0 to 6.3 mmol/L) continued to be within an acceptable range. 50 Dr Gay saw Ms Fowler on 17 October and found her to be in "reasonable health given her current medical condition". There was no reason to expect that on her current medication and dietary plan she would have a significant hypoglycaemic episode. (Exh C tab 5) 51 On the morning of 18 October Ms Fowler's BGL was 5.8 mmol/L. 52 AINs Juneen Avery and Erin Mathews attended to getting Ms Fowler out of bed that morning and assisted her with her "personal cares". She was taken to breakfast in the dining room, walking with the assistance of her walking frame. AIN Mathews fed Ms Fowler and she ate all of her breakfast. AIN Avery described Ms Fowler as "taking direction well and appeared normal" and "her movements were as good a standard as I'd come to expect from her". AIN Mathews noticed nothing untoward about her. (T608-9; 637) 53 After breakfast Ms Fowler was taken to the toilet and then to the lounge room in the south wing. At some stage, and in accordance with her habit, she made her way to sit on a lounge in the foyer of the ward near the nurses' station. At about 11.45am the AINs started to take residents to the dining room in anticipation of lunch which was usually at midday. AIN Mathews assisted Ms Fowler to eat lunch. She ate most of it but appears to have become frustrated with AIN Mathews' encouragement to eat. There is evidence that she made a remark of rebuke and attempted to bite AIN Mathews. (T610; 640) The accused was present in the dining room giving out medications when the verbal outburst occurred and he remarked "Where did that come from?" AIN Mathews was untroubled; she said this would sometimes happen as Ms Fowler would occasionally become frustrated when being encouraged to eat and drink. (T620) AIN Avery, however, thought it was "odd or strange and … out of character for the Mrs Fowler that [she] knew". Notwithstanding this, she thought Ms Fowler appeared her usual self throughout the remainder of the day. (T640) (Ms Avery's shift ended at 3.00pm.) 54 After lunch, Ms Fowler was taken to the toilet and then back to sit in the foyer. AIN Mathews said that when she left her she seemed fine. She last saw her at about 1.45pm. (T620-622) 55 At some stage thereafter Ms Fowler returned to her room. Her daughter, Julie Ross, entered SummitCare at 3.53pm, signed the visitors' book and proceeded to her mother's room. She briefly said hello as she passed a resident near the entry to the ward and popped into a room to say hello to another resident. It must have been close to 4.00pm when she entered her mother's room. The dividing curtain separating the two beds in the room was pulled across. Ms Ross found her mother in a state that prompted her to immediately seek assistance from nursing staff. (Exhibit C tab 11) 56 RN Maria Zubcakova and other staff attended upon Ms Fowler. She was cold, clammy and unresponsive. Her temperature was low and her BGL was found to be 1.1 or 1.3 mmol/L. (Exhibit C tab 9 p.356; 375) In short, she was suffering from hypoglycaemia and hypothermia. The ambulance service was called and Ms Fowler was taken to John Hunter Hospital. 57 Doctors at the hospital confirmed a diagnosis of hypoglycaemia. There was not thought to be any reason for suspicion. The family decided there should be no further investigations or treatment and simply asked that Ms Fowler be kept comfortable. She was returned to SummitCare late that evening for palliative care. This news was passed on at the handover to the Saturday morning shift staff, including the accused. Ms Fowler passed away at about 12.50pm the next day. 58 Dr Gay accepted that Ms Fowler had passed away due to an age related event. On Monday 21 October she signed death and cremation certificates. Later that day she reviewed incoming correspondence from John Hunter Hospital about another of her patients, Ms Audrey Manuel, and noted the similarity of the diagnoses. She contacted Ms Tranter at SummitCare who said she would look into it. She also contacted Dr Quach at the hospital who said it looked like a deliberate poisoning with a large dose of insulin. She recalled the death and cremation certificates and the matter was referred to the Coroner. 59 Ms Fowler's body was returned from the funeral directors and taken to the Newcastle Department of Forensic Medicine. Dr Jane Vuletic, the Senior Staff Specialist in Forensic Pathology, carried out an autopsy examination on 24 October 2013. She concluded that the direct cause of death was "bilateral bronchopneumonia" with the antecedent causes being "hypoglycaemia" and "insulin overdose". Ms Audrey Manuel 60 Ms Audrey Manuel was aged 91 as at 18 October 2013. She had two daughters and a son. She had been a resident in Room 19 in the north wing of the Mountview ward since moving into SummitCare in 2009. She weighed 36kg. 61 Ms Manuel was also under the care of Dr Gay who first saw her in March 2013. She was a dementia patient who also suffered from osteoporosis, depression, glaucoma and gastro-oesophageal reflux. These conditions were being well-managed according to Dr Gay. There was no history of diabetes. (Exh C tab 15) Dr Graham Walter, who was Ms Manuel's treating GP in 2010 to 2013 confirmed that there was no history of diabetes. (Exh C tab 21) 62 Dr Gay last saw Ms Manuel on 17 October 2013. She seemed to Dr Gay to be very well, consistent with her current medical condition. There was no reason to expect a significant hypoglycaemic episode. 63 On the morning of Saturday 19 October she was seen by AIN Jordan Franks at 7.00am and appeared to be fine. Ms Franks assisted her with breakfast and said, "While I was feeding Audrey she appeared in good spirits and her normal mood. She is always happy. She was talking to me." After breakfast she was walking around the wing with her walking frame as she usually did. (T660) 64 At about 10.30am Ms Manuel was sitting in the foyer area of the Mountview ward. She was asked if she would like morning tea and she responded with an affirmative nod, looking towards the inquirer, Ms Kathryn Tindall of the kitchen staff. (T269) Ms Tindall said that Ms Manuel usually liked a bit of a chat but not on this day; she assumed she was just having a bad day. She said, "Aud was a little confused when I spoke to her, a little different to her normal response. It seemed a little slow." (T271) 65 AIN Erin Mathews saw Ms Manuel sitting in the foyer at about 11.15am. She described her as being "a little confused". (T625) 66 Ms Michelle Sutherland arrived at 11.30am to visit her mother, Ms Maureen Wheatley. Ms Wheatley was sitting in the foyer with Ms Manuel. Ms Sutherland gave evidence that Ms Manuel had an untouched cup of tea and a piece of cake on the walker in front of her. She appeared to be dozing or sleepy and did not respond to an attempt to gain her attention. Ms Sutherland decided to watch her for a minute in case there was something wrong. While she sat next to her mother she noticed that Ms Manuel's right hand began to shake and then it became more pronounced. The accused was in the nearby nurses' station and he came out, saying, "I just caught that out of the corner of my eye". His attempts to gain a response were unsuccessful. (T361-365) 67 Other staff came to assist. RN Paulette Hills took charge. She noted that Ms Manuel was drooling, her eyes were open but she appeared on the verge of passing out. She was very pale. At her request the accused carried out full observations which revealed, amongst other things, that she had a low temperature of 33.4˚C and a high pulse. She was taken to her room by wheelchair. A blood glucose reading was taken with a very low result of 1.3 mmol/L. An injection of glucagon raised the level but not acceptably. An ambulance was called and Ms Manuel was taken to John Hunter Hospital. (T464-468) 68 In the Emergency Department at the hospital Ms Manuel was confirmed to be experiencing a hypoglycaemic and hypothermic episode. She was given repeated doses of glucose but the BGL did not respond adequately. The medical registrar, Dr Sven Speich, was advised by a clinical biochemist to carry out blood tests for insulin, C-peptide and glucose. The results were that the insulin level was 21,190 mlU/L (milli-International Units per litre), C-peptide was 0.1 microgram/L, and glucose was 0.8 mmol/L. Based on these results and other clinical observations he concluded that Ms Manuel's symptoms were "caused by the administration of a very high dose insulin in a non-therapeutic setting". (Exh C tab 17) 69 Ms Manuel was maintained on a glucose infusion as her BGL fluctuated over the ensuing five days. She remained under the care of Dr Quach who said that she continued to improve but never attained the same level of functioning she had prior to her admission. She was not able to swallow thin fluid competently and was being fed with thickened food only. She was not able to mobilize independently. Her prognosis was poor. Ms Manuel remained in John Hunter Hospital until 1 November 2013 when she was transferred to Belmont Hospital. She later moved to a nursing home in Waratah. She passed away in January 2015 from unrelated causes. Mr Ryan ("Greg") Kelly 70 Mr Ryan ("Greg") Kelly was 80 as at 19 October 2013. He was a widower with two daughters. He had been a resident in Room 24 in the north wing of the Mountview ward, having moved to SummitCare in February 2012. He was 164cm tall and weighed 60kg. 71 Dr Robert Kisonas had been Mr Kelly's treating general practitioner for a short time prior to October 2013. Mr Kelly's conditions included Alzheimer's dementia, ischaemic heart disease and hypertension. Dr Kisonas was not aware of any history of diabetes and said there would be no reason Mr Kelly would be prescribed insulin. (Exh C tab 30) 72 Ms Margaret Martin, daughter to Mr Kelly, last saw him on 14 October 2013 and described him as being "in his usual health". She was not aware of any problems between Mr Kelly and the staff of SummitCare and thought they all seemed to hold him in high esteem. (Exh C tab 32) 73 AIN Jordan Franks checked on Mr Kelly as part of her morning rounds soon after commencing her shift on 19 October 2013. She also fed him his breakfast in his room at about 8.30am and described him as being "absolutely fine". She also described him generally as being "mentally with it"; she was talking to him as she fed him and he appeared to be following the conversation. (T662; 666) 74 During the course of her shift Ms Franks checked on Mr Kelly three or four times and noticed nothing untoward. (T667) After she returned from her lunch break just before 2.00pm she saw Mr Kelly sitting in a chair in his room; he was asleep which was normal. 75 AIN Cheryl Minter assisted Mr Kelly in getting out of bed this morning and helping him to shower. She described him as being "his normal self". (T690) Later that day she brought lunch to his room but when she later returned he had not touched it; he said he did not want any and seemed tired. However, Ms Minter considered this to be "normal"; she said "he likes to sleep". (T687) 76 RN Paulette Hills walked past Mr Kelly's room at about 2.40pm and saw him standing with his walking frame at the end of his bed. He appeared to be fussing around with some of his personal belongings that were on the foot of his bed. There was nothing she observed that gave her any reason for concern about Mr Kelly's welfare. In fact, she noted that he was in apparent good health for the entirety of her shift. (T475-7) 77 TL Michael Webb went to Mr Kelly's room at about 3.15pm and found him sitting in his chair asleep. Mr Webb woke him. He described him as being a little drowsy; "as normal as you would [be] after being woken". He assisted Mr Kelly to have a sip of drink. He then went back to sleep and Mr Webb left. (T517) 78 At about 4.30pm, Mr Webb asked AIN Azeez Bello to go and wake Mr Kelly so he was ready for dinner. Mr Bello returned and said he was unable to wake him. Mr Webb went to Mr Kelly's room and found him still sitting in the same chair. He appeared to be asleep and did not respond to efforts to wake him. Mr Webb described the sound of Mr Kelly's breathing as very congested but he otherwise did not notice anything untoward. He formed the opinion that he was just in a very deep sleep. Mr Webb left, telling Mr Bello to continue trying to wake him. (T522-523) 79 Mr Azeez Bello had a different account, suggesting that Mr Webb had said "Just leave him, he's been like that all day, we tried to wake him up earlier but he wasn't waking up". Mr Bello was asked if he could be wrong about that and he agreed, but he reverted to his original evidence when cross-examined. (T497; 499) Mr Webb denied having said that. Mr Webb impressed as a very careful witness and a conscientious and experienced nursing assistant. I prefer his evidence over that of Mr Bello. 80 AIN Emily McAlpin, a student nurse, started work shortly before 5.00pm, having been called in because of a staff shortage this day. Her first task was to take evening meals to residents who ate in their rooms and her first port of call was Mr Kelly's room. She found him sitting in his chair, chin forward, drooling, and very cold and clammy. She could not rouse him so she went to get the registered nurse. (T395-397) RN Zubcakova said that it was about 5.30pm that she was approached by Ms McAlpin and asked to come to Mr Kelly's room. She gave a detailed description of what she observed; in short, Mr Kelly was hypoglycaemic and hypothermic. His BGL was undetectable and he was given an injection of glucagon which only raised the level to 0.9 mmol/L. (T431ff) … 82 An ambulance was called and Mr Kelly was taken to John Hunter Hospital. Amongst other things, Mr Kelly was confirmed as being hypoglycaemic. He was admitted and remained in hospital until he passed away on 29 October 2013. … 85 Dr Jane Vuletic carried out an autopsy examination on 30 October 2013. She found the direct cause of death was "aspiration pneumonia" with the antecedent cause being "insulin overdose". She found no evidence to indicate any acute deterioration of Mr Kelly's chronic medical conditions prior to death. (Exh C tab 23)"
His Honour also reviewed the evidence of the medical experts and concluded that the injections occurred within eight hours of the victims displaying significant symptoms of hypoglycaemia. His Honour found that it was quite likely that the period was shorter than that, but it was safer to act upon an eight hour period. He therefore concluded that Ms Fowler was injected at some time after 8am on Friday, 18 October, Ms Manuel was injected at some time after 3.30am on Saturday, 19 October and Mr Kelly was injected on that same day, at some time after 7am. His Honour's approach to the expert evidence and his conclusions have been challenged in this appeal.
The expert evidence as to the timeframe, within which the injections of insulin would have occurred, provided the basis for the police investigation. The police also had regard to SummitCare records and the CCTV camera film to establish who was at the facility between 17 - 19 October 2013 and at what times. His Honour noted that before the records were handed over to police, they were cross checked by SummitCare staff to ensure their accuracy. His Honour further noted that police did not conduct a physical search of the facility because the offending was not brought to their attention until the evening of 21 October. The crime scene had by that time become "contaminated". Using that information, police prepared a spreadsheet which became Exhibit D. This was prepared to show the times at which various persons were in attendance at the SummitCare facility, and extended from 5am on Friday, 18 October to 6pm on Saturday, 19 October.
His Honour found that there was no dispute at trial as to the accuracy of the content of Exhibit D. What was challenged, however, was whether the timeframe covered by the spreadsheet was extensive enough to have included all persons who had the opportunity to administer insulin injections to the victims. The proposition underlying the preparation of Exhibit D, i.e. that there was only one offender involved in all three offences was also challenged.
His Honour noted that Exhibit D referred to 319 persons. Of those 319, police were able to identify all but 35 persons. It was the Crown case that the inability to identify 35 persons did not invalidate the utility of the spreadsheet because each of those unknown persons only attended the facility on one of the two days under consideration. His Honour accordingly accepted that if the "one-perpetrator" theory was made out, the 35 unidentified persons could be excluded.
His Honour noted that through a process of elimination, based upon the "one-perpetrator" theory and upon the expert opinion that the injections occurred no more than eight hours before the victims were found with symptoms, police determined that there were potentially 25 people, including the applicant, who could be responsible for the offences. Of those 25 people, 5 were visitors to the facility and 20, including the applicant, were members of staff.
His Honour took into account the applicant's statements to police between October and December 2013. His first statement was made on 24 October 2013 and set out his background and duties at SummitCare. His Honour's summary of the content of the applicant's statements is uncontroversial and has not been challenged: "118 The following is derived from statements made by the accused to police on 24 October (Exhibit U), 6 November (Exhibit V) and 10 December 2013 (Exhibit W). 119 The accused was aged 26 as at October 2013. He started a traineeship as an Assistant in Nursing at St Joseph's Nursing Home at Sandgate in 2006. He worked at Uniting Care Koombahla at Wallsend from 2007 to 2009. He then took a casual Team Leader position at Redhead Gardens Nursing Home where he worked for six months in 2009 before taking a position at William Cape Gardens Nursing Home at Wyong where he remained before taking up his position at SummitCare, Wallsend in September 2011. (Exhibit U [3]-[5]) 120 In October 2013 the accused was employed as a Team Leader Certificate IV Aged Care worker. He worked 6.30am to 3.00pm shifts three days one week and four days the next. In his statement of 24 October 2013 he said that he had worked in the Mountview ward for more than six months with the same staff and his duties included: • Monitoring staff • Medication rounds • Simple wound care • Resident progress notes and related documents • Resident assessment regarding care • Resident funding assessments • Liaising with doctors • General observations of residents 121 The accused said he carried out a medication round at 7.00am until about 8.30 to 9.00am. He then assisted the registered nurse with a Schedule 8 drug round. He carried out another medication round at about 12.00pm until 1.00 to 1.30pm. He said he regularly administered medications to Ms Fowler, Ms Manuel and Mr Kelly including on 18 and 19 October. None of them received any form of medication by injection. (Exhibit U [6], [9]-[10], [14], [18]-[21], [23]) 122 The accused said that up until early 2013 the Certificate IV Team Leaders were permitted to administer insulin to residents. This was done via a "wind up insulin pen". (He gave a detailed description as to the procedure.) (Exhibit V [12]-[14]) 123 In January 2013, management directed that all insulin injections were to be administered via a needle and syringe and he said this was something that a team leader was not authorised to do. Since the change in policy he had "not injected any patients with any substances", "except for the very odd occasion" which he estimated at once or twice a month. This would occur when the registered nurse was busy and had asked him to do it for them. He said that the last occasion this occurred "would have been sometime around the start of October 2013". The only patient that he could recall being asked to inject with insulin since January 2013 was Mr Rex Joyce. (Exhibit V [15]-[17], [19]) 124 The accused nominated RN Stephen Zhao and RN Maria Zubcakova and "a couple of casual registered nurses" as those who had asked him to administer insulin. He would do so by way of needle and syringe, drawing the insulin up from a vial and injecting it to the front stomach area of the patient. (Exhibit V [18]) 125 The accused described Ms Fowler, Ms Manuel and Mr Kelly as "not problem residents". He said their behaviours were consistent with similar people in SummitCare. They seemed to live a trouble-free life in the facility and did not have any issues that caused him concern. (Exhibit U [24])"
The applicant made a second statement to police in which he specifically referred to his interaction with the victims on 18 and 19 October 2013: "Ms Gwendolyne Fowler 126 The accused said in his second statement that he gave Ms Fowler her morning medication on 18 October at about 7.45am in the dining room. In the third statement he said it was at about 7.30am. (Exhibit V [21]; Exhibit W [14]) 127 He again saw Ms Fowler at lunch in the dining room and specifically recalled that she attempted to bite the AIN who was trying to feed her. He saw her at about 2.20pm when she was sitting with Ms Manuel on a lounge near the nurses' station; they seemed to be chatting and were fine. He saw them again in the same place as he left for the day at 3.00pm and again "both appeared to be in good health". When he returned to work the next morning he was told that Ms Fowler had taken some sort of a turn and that she was in a palliative state. He checked on her during the course of the morning to ensure she was comfortable; the last occasion being at around 10.00am. (Exhibit V [21]-[25]) Ms Audrey Manuel 128 The accused said he gave Ms Manuel her morning medication at about 8.30am on 19 October in the dining room. He said, "She just seemed fine and was her usual self". She did not complain of any unusual pain or of feeling unwell. He said, "there was nothing that I saw that morning that gave me any indication that Audrey was unwell". (Exhibit W [28]-[31]) 129 The accused saw Ms Manuel shortly after 11.00am on 19 October when she was walking down the hallway of the north wing towards the foyer area. It was Ms Manuel's habit to walk around with her walking frame and to sit on the lounge near the nurses' station. She was sitting there when she "took her turn" at about 11.30am. The accused was in the nurses' station at the time and went to assist. It was while attending to Ms Manuel that he was informed that Ms Fowler had passed away. (Exhibit U [22]; Exhibit V [26]-[28]) 130 The accused said that his initial assessment was that Ms Manuel had suffered "a TIA which is a type of small stroke". He said: "It was my belief that despite being taken to hospital in a poor condition there was nothing to have suggested to me at that time that she would pass away. It is my experience that some patients can have a TIA or neurological events and can make a good recovery." (Exhibit V [33]) Mr Ryan Kelly 131 The accused said that he gave medications to Mr Kelly at around 8.00am on Saturday 19 October (or 8.40am according to the third statement). In the second statement, the accused said that Mr Kelly "seemed his normal self and was generally fine". Aside from his medical conditions he was "in otherwise reasonable health". In the third statement he said that Mr Kelly did not complain of any unusual pain or of feeling unwell and that "there was nothing that I saw that morning that gave me any indication that Ryan was unwell". He described Mr Kelly as a patient "who's level of health was basically the same all of the time". He would walk around and wander a bit but for the most part sit in his room and read, or just sleep. (Exhibit V [29]; Exhibit W [22], [24) 132 The accused gave Mr Kelly his lunchtime medication in his room at about 12.15pm (second statement) or 12.50pm (third statement) as he ate his lunch. Mr Kelly seemed "his normal self" and "there was nothing that I saw or he told me which indicated to me that he was in anyway unwell". He last saw him at around 2.15pm standing at the end of his bed appearing to be fiddling or fussing around with some paperwork. He said Mr Kelly "seemed okay to me at this time and… gave me no reason to believe that he was about to take ill". He said: "At the time that I left work around 3.15pm I had fully expected to come back to work the following day to find Ryan in his room in the same medical condition that I had last seen him when I left work the previous day. I had no reason to believe that Ryan's medical condition was anything other than it had been for the months before. I had no reason to consider that within a number of days Ryan would be dead." (Exhibit V [30]-[31]; Exhibit W [25]-[26]) 133 The accused said that when he returned to work at 6.30am on Sunday 20 October he was told at the handover that Mr Kelly had taken a turn; a test revealed that he had a low blood sugar level; and that he had been taken to hospital. (Exhibit V [29])"
Police interviewed the applicant on a third occasion on 21 August 2014. The interview was recorded and was very lengthy, having occupied over seven hours. By agreement between the parties, it was edited and about one-third of the recording was tendered. His Honour took into account the applicant's demeanour during the interview and noted that one of the major topics discussed was the exchange of text messages between the applicant, Ms Debbie Wilson, Ms Raylene Collins and Ms Monique Christensen. His Honour summarised the rest of the statement. That summary is not controversial and has also not been challenged in the appeal: "137 The accused was asked about working at SummitCare generally and in the Mountview ward in particular. He described the work as busy but mainly from the start of his shifts at 6.30am until 1.00pm. (Q 877) 138 There were questions about a search warrant that was executed at the accused's residence on 5 December 2013. (Q1034-1041) Curiously, there is nothing in the edited form of the interview that is before me that deals with some seized items that were tendered: a copy of a chapter from "MIMS Full Prescribing Information" dealing with Human Insulin, about 15 syringes and about the same number of needles. (Exhibits AC and AD) 139 In another part of the interview in which Detective Sergeant Faber asked the accused what he thought had happened in relation to the injection of Ms Fowler, Ms Manuel and Mr Kelly (Q1654ff) the accused clearly struggled to provide an answer (which is, of course, equally consistent with his innocence). He said, in effect, that he could not think of somebody doing such a thing. (Q1664) He did not think that any of the victims would have been capable of injecting themselves. (Q1729-30) 140 The accused thought each of the victims might have questioned somebody giving them an unexpected injection, even himself. (Q1736ff) He denied having injected any of them for any reason. (T1755) 141 He doubted that a fellow resident might have injected the victims. (Q1783-93) It was possible, he thought, that a visitor to the facility might have. If the person was a stranger it would be less likely but at busy times they might go unnoticed. (Q1794-1806) 142 Detective Sergeant Faber went through a list of names of staff the police had identified as being present at SummitCare in the relevant period in which the injections were thought to have been administered and asked the accused whether he had any reason to suspect them. He did not suggest that any of them could be responsible, although there were some he did not know very well. Mr Watson submitted that this was indicative of innocence in that a guilty person would be likely to deflect responsibility on to somebody else. (T844) However, another way of looking at it is that a person dishonestly deflecting responsibility onto to somebody who could be shown to be innocent might then attract further suspicion to themself. 143 The accused could not think of any reason why somebody would inject the three victims. They were "perfect residents"; "easy to look after"; and "they weren't really a bother". (Q2087) 144 Detective Sergeant Faber asked if the accused thought that the perpetrator could have had euthanasia in mind and he did not completely dismiss the notion: "They were up and down all the time but you know Ryan Kelly one day he was fine the next day he was bad, one you know, it's probably the same for all three of them they were up and down all the time but you know somebody could, somebody you know but yeah." (Q2090) 145 The accused was asked how he felt about being on the list of possible perpetrators: "Q2112 How does it make you feel to be on that list? A Um I don't know, a bit depressed because you know I shouldn't be on it but you know I'm on it. Q2113 Why shouldn't you be on it? A 'Cause I know that I didn't do anything like this.""
The applicant was arrested at his home on 17 December 2014. He largely exercised his right to silence and his Honour did not draw any adverse inference from that. The applicant did tell police "I didn't do it" and specifically that he did not kill Ms Fowler, Mr Kelly nor attempt to kill Ms Manuel. He also reiterated that he could not think of anyone who could have committed the offences.
His Honour dealt with the text messages in some detail. He considered first the communication with Ms Wilson, who was an AIN at Summitcare. She worked night shift on 17 October and on Friday, 18 October. The text was initiated by Ms Wilson at 11.14am on Saturday, 19 October. Ms Wilson gave her interpretation of the meaning of the text in evidence and the applicant gave his interpretation in his statement to police of 21 August 2014.
There were also texts with Ms Raylene Collins, who was an AIN, who worked the night shifts on 17 - 19 October 2013 in the Parkview ward with Ms Wilson. At 4.13pm on Saturday, 19 October Ms Collins received a text message from the applicant. Following that message, there was a further exchange on the Saturday evening at approximately 8pm. Ms Collins gave evidence as to her understanding of the texts.
There was an exchange of texts with Ms Monique Christensen on Saturday, 19 October 2013 at about 10pm. Ms Christensen was a friend of the applicant.
The trial judge set out in detail the questions and answers put by police to the applicant in relation to the content of the texts with Ms Wilson and Ms Collins at [169]-[182] of the judgment. It is fair to say that the applicant's responses were not particularly convincing. An example is as follows: "169 … He was asked about having said in that statement that Mr Kelly was apparently well when he, the accused, left SummitCare at the end of his shift as compared to having referred to the prospect of Mr Kelly dying in the text message some four hours earlier. He said: "Look, I understand there's a contradiction between what I've said in my statement and those text messages and I don't really have an answer to it from the fact that it might look bad to you but I can't explain it." (Exhibit Z at Q116)"
His Honour set out his conclusion in relation to the text messages as follows: "The text messages 253 If they stood by themselves the text message exchanges would not be decisive in the context of the standard of proof being beyond reasonable doubt. The Crown did not contend otherwise. They must be looked at as another strand in the cable of the circumstantial case. 254 I am satisfied that they do suggest foreknowledge by the accused of the impending or prospective demise of Ms Manuel and Mr Kelly. On his own account, particularly prior to being confronted with the messages, he had no reason to think that either of them would die when he sent the messages. At 11.27am on 19 October Mr Kelly had not demonstrated any symptoms and at 1.23pm Ms Manuel had been taken to hospital with something he claimed to police he thought was not life-threatening. Yet in his private communications with Ms Wilson and Ms Collins he was suggesting they were the next residents on the ward to die. By 8.00pm, he expressed confidence in his prediction about Ms Manuel ("Its gonna be audrey")."
Other bases for that conclusion in relation to the texts are set out at [255]-[261] of the judgment. There, his Honour concluded: "260 I accept that there might be "black humour", as it was described, in conversations between nurses and carers who work in a stressful environment in which death is constantly present. However, the accused's attempts to portray the text messages in this way fails when it might be more expected that if he was going to predict someone's death he would nominate people who were the most ill amongst the residents of Mountview; not people who were, despite their respective medical conditions, comparatively stable, or in the case of Ms Manuel, thought to have taken a turn which was not anticipated to be catastrophic. 261 In my view, the references to the imminent demise of Ms Manuel and Mr Kelly support the Crown's circumstantial case in that they are indicative of the accused knowing that they had both received injections of insulin."
In relation to similarities between the events and the application of coincidence reasoning, the trial judge gave the following analysis: "193 In the course of his closing address the Crown Prosecutor made submissions about similarities in the three events and the circumstances in which they occurred which pointed to the same person being the perpetrator in each case. It was submitted that if such similarities were accepted as establishing, or at least supporting, that conclusion, then the evidence would have "significant probative value". The Crown Prosecutor refrained from addressing on the s 101(2) issue in light of the fact that this was a judge-alone trial. (T803ff; 829) 194 The situation is rather unusual in that s 98 is concerned with admissibility of evidence and the reality is that no question of admissibility arises. The evidence upon which the Crown relies was always going to be led in the trial without objection and that is what occurred. The only issue is whether coincidence reasoning may be deployed. At the risk of being technical (because the parties did not address the issue in this way), it may well be characterised as an issue of whether the evidence pertaining to each of the three counts in the indictment is cross-admissible. 195 I did not understand Mr Watson to have ever challenged the use of the evidence in relation to one event as coincidence evidence in relation to the other events. Certainly he never articulated a challenge in the form of an "objection". His arguments were confined to the strength of the Crown's arguments as to the similarities. Perhaps this was an argument about whether the evidence had "significant probative value". In a moment I will indicate my assessment on that subject. There was no suggestion of any prejudicial effect. 196 In these circumstances it is doubtful that I need to rule on the topic at all. However, for abundant caution I can indicate that I am satisfied that the evidence does have significant probative value for the reasons that appear elsewhere; it outweighs the non-existent prejudicial effect; and accordingly I rule that the evidence in relation to individual counts is available to be used as coincidence evidence in relation to the other counts."
His Honour then considered alternative hypotheses put forward by the defence as to the perpetrator being someone other than the accused. His Honour noted that the defence had questioned the legitimacy of the theory that there was only one perpetrator.
In relation to that issue, his Honour found: "Multiple perpetrators 198 There was little advanced in support of the contention that there might have been more than one person responsible for these crimes beyond a bare assertion that it might be possible. 199 These crimes were far from typical; they were confined to a very small pool of potential victims in a relatively confined place and in a relatively short timeframe. Moreover, they needed to be perpetrated by a person with particular knowledge and ability and who had a connection with and access to the victims. I am satisfied that when regard is had to all of the circumstances and their similarities the conclusion that there was only the one perpetrator is inescapable."
The trial judge identified the various hypotheses put forward on behalf of the applicant to the effect that the perpetrator was not him. They were: 1. An unidentified person who entered without being captured on CCTV. 2. A resident of the SummitCare aged care facility. 3. A visitor to the SummitCare aged care facility. 4. Someone else as the perpetrator. 5. Another staff member (excluding the applicant) as the perpetrator. 6. A woman as the perpetrator. 7. "Anything might be possible" by reference to the evidence.
His Honour set out why those alternative hypotheses lacked substance. In relation to the "anything might be possible" alternative, his Honour had regard to the lack of motive and the adequacy of the police investigation which had been criticised by defence counsel. On that last issue, his Honour concluded: "236 With respect to Mr Watson, these suggestions as to what the police might have done are quite unrealistic and whether they would have yielded anything useful appears most unlikely. Although it is not at all determinative of this issue, I note that none of these matters were raised with Detective Sergeant Faber. A major topic of his evidence-in-chief was the adequacy of the investigation but he was not cross-examined at all. 237 Another point raised in the closing address on behalf of the accused was that there was a lack of thorough testing sought by the police as to the type of insulin found in the samples obtained from the victims. The point was that if there had been, it might have been possible to show that the insulin was not of a type that was on hand in the treatment rooms in the SummitCare wards. Further, it might have been possible to identify a shorter timeframe for when the injections may have occurred. Little, if anything, turns on this. The evidence is what it is. It is not a matter of there being potentially exculpatory evidence that was not available."
As a last step in his analysis, the trial judge put together the elements of the circumstantial case against the applicant. As the starting point, his Honour concluded that the same person had committed the three offences: "238 The starting point in examining the possibility that the accused was the person responsible for injecting each of the victims is to acknowledge that it is an undisputed fact that somebody did so and that I am satisfied that it was the one person. The concision with which this can be stated should not belie its significance. It demonstrates that what would otherwise seem improbable - that a person could inject a harmful substance into three residents in the Mountview ward without anyone noticing - did in fact occur."
The other circumstances to which his Honour had regard were: 1. The applicant was on duty in the Mountview ward when the injections were given. 2. The applicant was skilled and experienced in the injection of insulin. 3. The applicant administered insulin to two insulin dependent residents of Mountview ward on the mornings of 17 and 18 October. 4. The applicant was found in possession of needles, syringes and information about insulin. 5. The applicant had access to insulin. 6. The applicant had the opportunity to give the injections. 7. The relationship between the applicant and the victims. 8. Other potential perpetrators have been eliminated. 9. The effect of the text messages when taken with the other circumstances.

Relevant legal principles

The High Court has summarised the correct approach to be taken in a circumstantial evidence case in The Queen v Baden-Clay [2016] HCA 35; 258 CLR 308 (French CJ, Kiefel, Bell, Keane and Gordon JJ): "46 The prosecution case against the respondent was circumstantial. The principles concerning cases that turn upon circumstantial evidence are well settled. In Barca v The Queen, Gibbs, Stephen and Mason JJ said: "When the case against an accused person rests substantially upon circumstantial evidence the jury cannot return a verdict of guilty unless the circumstances are 'such as to be inconsistent with any reasonable hypothesis other than the guilt of the accused': Peacock v The King. To enable a jury to be satisfied beyond reasonable doubt of the guilt of the accused it is necessary not only that his guilt should be a rational inference but that it should be 'the only rational inference that the circumstances would enable them to draw': Plomp v The Queen; see also Thomas v The Queen." 47 For an inference to be reasonable, it "must rest upon something more than mere conjecture. The bare possibility of innocence should not prevent a jury from finding the prisoner guilty, if the inference of guilt is the only inference open to reasonable men upon a consideration of all the facts in evidence" (emphasis added). Further, "in considering a circumstantial case, all of the circumstances established by the evidence are to be considered and weighed in deciding whether there is an inference consistent with innocence reasonably open on the evidence" (emphasis added). The evidence is not to be looked at in a piecemeal fashion, at trial or on appeal."
In Lane v R [2013] NSWCCA 317 the Court of Criminal Appeal (Bathurst CJ; Simpson and Adamson JJ) spoke of the "very real distinction between drawing an inference from proven facts and engaging in speculation": "109 … In Seltsam Pty Ltd v McGuiness; James Hardie & Coy Pty Ltd v McGuiness [2000] NSWCA 29; 49 NSWLR 262, Spigelman CJ considered this very question. He acknowledged that it is often difficult to distinguish between permissible inference and conjecture. Quoting from Jones v Great Western Railway Co (1930) 144 LT 194, his Honour adopted a definition of inference as "a deduction from the evidence" which, if reasonable, may have the validity of legal proof. 110 He referred also to Caswell v Powell Duffryn Associated Collieries Ltd [1940] AC 152, quoting as follows: "Inference must be carefully distinguished from conjecture or speculation. There can be no inference unless there are objective facts from which to infer the other facts which it is sought to establish. In some cases the other facts can be inferred with as much practical certainty as if they had been actually observed. In other cases the inference does not go beyond reasonable probability. But if there are no positive proved facts from which the inference can be made, the method of inference fails and what is left is mere speculation or conjecture." Spigelman CJ stated the test as: "... whether, on the basis of the primary facts, it is reasonable to draw the inference." To similar effect were observations of Gibbs, Stephen and Mason JJ in Barca v The Queen [1975] HCA 42; 133 CLR 82 at 104-105."
The determination of this appeal is governed by s 6(1) of the Criminal Appeal Act 1912 (NSW): "6(1) The court on any appeal under s 5(1) against conviction shall allow the appeal if it is of the opinion that the verdict of the jury should be set aside on the ground that it is unreasonable or cannot be supported having regard to the evidence or that the judgment of the court of trial should be set aside on the ground of the wrong decision of any question of law or that on any other ground whatsoever there was a miscarriage of justice and in any other case shall dismiss the appeal; provided that the court may notwithstanding that it is of the opinion that the point or points raised by the appeal might be decided in favour of the appellant dismiss the appeal if it is satisfied that no substantial miscarriage of justice has actually occurred."
In their joint judgment in Filippou v The Queen [2015] HCA 29; 256 CLR 47 French CJ, Bell, Keane and Nettle JJ said: "4 As will appear, the Court of Criminal Appeal is required to deal with an appeal from judge alone in three stages. The first is to determine whether the judge has erred in fact or law. If there is such an error, the second stage is to decide whether the error, either alone or in conjunction with any other error or circumstance, is productive of a miscarriage of justice. If so, the third stage is to ascertain whether, notwithstanding that the error is productive of a miscarriage of justice, the Crown has established that the error was not productive of a substantial miscarriage of justice. … The nature of a criminal appeal from a judge alone 6 Section 133 of the Criminal Procedure Act 1986 (NSW) provides that a judge who tries a criminal proceeding without a jury may make any finding that could have been made by a jury on the question of guilt of the accused and that such a finding has the same effect as a verdict of a jury. In that sense, "finding" means an ultimate finding of guilt as opposed to a finding of fact leading to the finding of guilt. The section also provides that the judge must include in his or her reasons for judgment the principles of law applied and the findings of fact on which the judge relies; and that, if any Act or law requires a warning to be given to a jury in such a case, the judge is to take the warning into account in dealing with the matter. As was held in Fleming v The Queen, the requirement to take a warning into account necessitates that the judge expressly refer to the warning in his or her reasons for judgment. … 9 As was also explained in Fleming, perforce of s 133 of the Criminal Procedure Act, each of the three limbs of s 6(1) of the Criminal Appeal Act is capable of application to the verdict of a judge alone. For the purposes of the first limb, the question is whether, upon the evidence on which the judge acted, or upon which it was open to the judge to act, the judge's finding of guilt is "unreasonable" or "cannot be supported". For the purposes of the second limb, the question is whether the judge has erred in law in the sense of a departure from trial according to law. Under the third limb, the question is whether for any other reason there has been a miscarriage of justice. … 11 Beginning with the first limb of s 6(1) of the Criminal Appeal Act, it is clear from the terms of s 133(1) of the Criminal Procedure Act that the effect of the latter provision is to equate a judge's finding of guilt to a jury's finding of guilt "for all purposes". It follows from the natural and ordinary meaning of the words of s 133(1) that, for the purposes of an appeal against conviction under s 5 of the Criminal Appeal Act, a judge's finding of guilt is to be treated as if it were the same as a jury's finding of guilt."

Preliminary matter

At the commencement of oral submissions in the appeal, senior counsel for the applicant submitted that one of the hypotheses available to the applicant, which had not been rebutted by the Crown, was that the insulin could have been accidentally administered to each of the victims, i.e. that the Crown had failed to establish that in each case the insulin had been deliberately administered. On behalf of the applicant, it was submitted that although such an alternative had not been raised in terms at the trial, the applicant could rely upon it in these proceedings because it simply involved a question of emphasis and it was a matter which the Crown was obliged to prove in any event.
This submission cannot be accepted. There are a number of reasons for this.
It is clear that the trial was not run on that basis. In his opening, the following exchange took place between the trial judge and counsel for the applicant: "HIS HONOUR: Mr Watson do I take it from what you've said that you say that it is not necessarily the case it was the same person who administered or injected the insulin to each of the three people. WATSON: Yes. HIS HONOUR: Okay. WATSON: And your Honour will pick up that we're not contesting the way these people were treated. We don't contest that the three victims were injected with high doses of insulin. That's not part of our case at all. Certainly our case simply put is we don't know who did it but it wasn't us." (T.21.33)
In his closing address, counsel for the applicant was even clearer on this issue: "WATSON: Your Honour they're my submissions. At the end of all of this there is - your Honour wouldn't be satisfied beyond reasonable doubt that the accused has committed the offences or that the prosecution has satisfied your Honour with all the elements of the charges that the accused has faced that there is - as I said to your Honour in my opening address that this is a situation where the accused recognises in the circumstances that the offending behaviour has certainly been carried out by somebody, but not him, and it is a circumstantial case." (T.844.3)
There was never a submission at trial to the effect that there was a possibility in respect of one or more of the victims that there was an accidental administration of insulin. There was no evidence to that effect other than speculation by Dr Gay, Ms Fowler's general practitioner, at an early point in time that there might have been a medication error. That was never pursued or expanded on at trial.
Significantly, the reports of two of the experts, Drs Quach and Tran, were tendered and they were not required for cross-examination. In his report of 7 November 2013 (AB 490) Dr Quach commented on the very high levels of insulin found in Ms Manuel and Mr Kelly as follows: "24 Kelly's level of insulin was high and outside the expected range for a patient receiving insulin. It maybe possible that he was given an insulin dosage in excess of a normal therapeutic given to any insulin dependent patient. 25 Manuel's level of insulin was extremely high far in excess of normal therapeutic dose given to any insulin dependent patient." [Quite apart from the fact that neither Ms Manuel or Mr Kelly was insulin dependent, Dr Quach was able to eliminate accidental administration of the insulin. (AB 490)]
Dr Tran's report of 7 November 2013 was to similar effect. He reported: "14 … I am able to say that although a high reading, a level in the vicinity of 92.6mlU/L may be in some cases, within an acceptable range for a particular patient. However, the result for Manuel is markedly above this acceptable level and would be unlikely to be encountered through normal insulin treatment processes for diabetes." (AB 498) "27 The result for Fowler is excessive and beyond the normal acceptable levels for an insulin dependent or insulin treated patient and is unlikely to be encountered through normal insulin treatment processes for diabetes." (AB 500) Both Dr Quach and Dr Tran referred the results of their tests to NSW Police.
In further confirmation that accidental administration was never suggested at trial, none of the witnesses who were responsible for injections of insulin was questioned, i.e. the registered nurses, about whether there was or was likely to have been an accidental injection of insulin by one of them. The issue was never explored with the registered nurses because it was not an issue at trial. The trial was run without demur on the part of the defence on the basis that the necessary intention to kill would be inferred from the act alone. This act was the deliberate injection of a high quantity of insulin to non-insulin dependent residents on each occasion over the course of a two day period.
The case for the defence was that the applicant did not inject the victims. There was an acknowledgement by the defence that somebody committed the crime but not him. Inherent in that acknowledgement is an acceptance that the act of injecting the victims was deliberate, and not accidental.

(a) Admitting evidence for a coincidence purpose

The applicant relied upon the following statements of principle in support of this ground: DSJ v R; NS v R [2012] NSWCCA 9 (Whealy JA): "50 An important cornerstone in the structure of the Evidence Act 1995 is s 55. This deals with the topic of relevant evidence. The evidence that is relevant in a proceeding is evidence that, if it were accepted, could rationally affect (directly or indirectly) the assessment of the probability of the existence of a fact in issue in the proceedings. 51 Another key section is s 56(1) which provides that, except as otherwise provided by this Act, evidence that is relevant in a proceeding is admissible in a proceeding. Subsection (2) provides that evidence that is not relevant in the proceeding is not admissible. 52 I have earlier set out the definition of probative value - the extent to which the evidence could rationally affect the assessment of the probability of a fact in issue. 53 As is recognised and stated by the High Court in Smith v The Queen (2001) 206 CLR 650, the first question to be asked when an issue of the present kind arises is whether the evidence is relevant. … As Gleeson CJ observed in Smith v The Queen, although questions of relevance may raise "nice questions of judgment", no discretion falls to be exercised. Evidence is either relevant or it is not. If the evidence is not relevant, no further question arises about its admissibility. The simple fact is that irrelevant material may not be received. This is a proposition that is fundamental to the law of evidence and well settled. 54 In determining the basic issue of relevance in a criminal trial, it is especially important to identify the ultimate issues. These will ordinarily be expressed in terms of the element of the offence with which the accused stands charged in light of the defence, if known. However, behind these ultimate issues, there will often be many issues about facts that are relevant to facts in issue (Smith v The Queen at 654; Papakosmas v The Queen (1999) 196 CLR 297 at 307: especially at 312 and 321-322). 55 In my opinion, it is plain that s 98, in its terms, poses this simple question: whether the evidence being considered is capable, to a significant degree, of rationally affecting the assessment (ultimately by a jury) of the probability of the existence of a fact in issue. Again, in its terms, it requires the trial judge to assess whether the evidence has capacity to that extent and for that purpose." (Emphasis added by the applicant)
R v Gale; R v Duckworth [2012] NSWCCA 174; 217 A Crim R 487 (Simpson J, McClellan CJ at CL and Fullerton J agreeing): "25 At its heart, s 98 is a provision concerning the drawing of inferences. The purpose sought to be achieved by the tender of coincidence evidence is to provide the foundation upon which the tribunal of fact could draw an inference. The inference is that a person did a particular act or had a particular state of mind. The process of reasoning from which that inference would be drawn is: • two or more events occurred; and [step 1] • there were similarities in those events; or there were similarities in the circumstances in which those events occurred; or there were similarities in both the events and the circumstances in which they occurred; and [step 2] • having regard to those similarities, it is improbable that the two events occurred coincidentally; [step 3] • therefore the person in question did a particular act or had a particular state of mind." [step 4/ultimate inference]
The applicant submitted that it followed from those statements of principle that the particular fact in issue must be identified with precision before any decision can be made as to its admissibility under s 55 (relevance) or s 98 (significant probative value) of the Evidence Act 1995 (NSW) (the Act). The applicant accepted that defence counsel at trial did not dispute that the injections had been administered deliberately and that the issue at trial was whether in each case it was the applicant who was responsible for injecting the victims.
The applicant submitted that both the Crown and the trial judge proceeded on an erroneous construction of s 98 by failing to identify the particular issue which was sought to be established by the use of coincidence reasoning, with the result that the inferential reasoning process set out by Simpson J was limited to the first three steps set out above. The failure to identify the ultimate inference [step 4] meant that the determination under ss 98 and 101(2) of the Act proceeded on an erroneous basis, i.e. that coincidence reasoning could be used to "narrow down" the range of potential suspects.
The applicant accepted that the trial judge had correctly noted that the issue was really one of whether coincidence reasoning could be deployed, as opposed to whether the evidence sought to be used for this purpose was itself admissible and that a central element of the Crown case against the applicant was that "one person" had been responsible for injecting each of the victims.
The applicant submitted that in line with what was said in DJS and Duckworth, the correct approach to s 98 was to determine first the particular fact in issue which was sought to be established (for the purposes of determining the question of relevance under s 55). The applicant submitted that in respect of each count, the only fact in issue was whether he was the person who had administered the relevant injection or injections, i.e. the ultimate inference of Simpson J's inferential reasoning process "therefore the person in question did a particular act or had a particular state of mind".
In further support of that proposition, the applicant referred to R v Merritt [1999] NSWCCA 29 at [39] where the Court (Wood CJ at CL; James and Adams JJ) said: "39 … The essence of coincidence evidence is that there is established to have been two events in circumstances where the existence of one event establishes the likelihood that the appellant was responsible for, or involved in, the other separate event or to establish that the other of those events did not occur by accident …" (Emphasis added by applicant)
The applicant submitted that this statement was apposite to this case and that on the facts of this case, "the existence of one event" needed to be established with sufficient specificity to inculpate the applicant, it being the case that the applicant did not contend that the events occurred coincidentally or else by accident.
The applicant submitted that put another way, the trial judge needed to satisfy himself that the applicant had been responsible for the administering of at least one of the relevant injections before the judge could deploy coincidence reasoning to infer that, because the applicant was guilty on one count, he was guilty on all three counts.
The applicant submitted that such an approach was consistent with what was said by this Court in Gilham v R [2012] NSWCCA 131 which concerned a situation in which three people, being the father, mother and brother of the appellant, had been stabbed to death in the family home. The appellant admitted to having killed his brother but denied being responsible for the death of his parents. The defence case was that the appellant's brother had stabbed and killed their parents, set the bodies alight and that the appellant had then killed his brother. The applicant submitted that the principal fact in issue in Gilham was the identity of the person who had stabbed the parents, with the only two possible perpetrators being the appellant and his deceased brother. The applicant noted that the Crown argued in Gilham that there were similarities between the stab wounds in each of the three deceased persons and that this pointed to the perpetrator of each homicide being the same person.
In support of that proposition, the court in Gilham (at [328]) said: "328 The identification of the proposed use of the similarity evidence as falling within the coincidence rule is not purely academic. The coincidence rule requires that the court admitting the coincidence evidence be satisfied that the evidence has, either by itself or in conjunction with other evidence, "significant probative value". The coincidence evidence had probative value in the sense that it could have affected the jury's assessment of whether one person stabbed each of the deceased. Proof of that fact would have strongly supported the Crown case. But the question whether the evidence had significant probative value was never explored at the trial. Relevant to that question is the varying number of stab wounds; the fact that most stabbings are to the chest area; and the consideration that some of the similarities in the pattern of wounding may be accounted for by the use of the same knife, as distinct from their infliction by the same attacker. For the purposes of this appeal, it is not necessary to decide whether the evidence had significant probative value. We note, however, that that question may fall to be determined at a subsequent trial."
The applicant submitted that in Duckworth, after explaining the inferential reasoning process involved in the use of coincidence evidence, Simpson J set out the discrete steps for arriving at a determination as to the admissibility of that evidence under s 98.

The first step is to identify the "particular act of a person" or the "particular state of mind of a person" that the party tendering the evidence seeks to prove;
The second step is to identify the "two or more events" from the occurrence of which the party tendering the evidence seeks to prove that the person in question did the "particular act" or had the "particular state of mind";
The third step is to identify the "similarities in the events" and/or the "similarities in the circumstances in which the events occurred" by reason of which the party tendering the evidence asserts the improbability of coincidental occurrence of the events;
The fourth step is to determine whether "reasonable notice" has been given of the intention to adduce the evidence (or if reasonable notice has not been given, whether a direction under s 100(2) ought to be given dispensing with the requirement);
The fifth step is to make an evaluation of whether the evidence will, either by itself or in conjunction with other evidence already given or anticipated, "have significant probative value";
In a criminal proceeding, if it is determined that the evidence would have "significant probative value", the sixth step is the determination whether the probative value of the evidence "substantially outweighs" any prejudicial effect it may have on the defendant (s 101(2)).
The sixth step necessarily involves some analysis both of the probative value of the evidence in question and any prejudicial effect it might have: R v RN [2005] NSWCCA 413, and a balancing of the two.

The applicant submitted that in Duckworth Simpson J had observed that in relation to this step-wise process for determining the admissibility of coincidence evidence in criminal proceedings, "the final and central question is whether the probative value of the evidence "substantially outweighs" its obvious "prejudicial effect" (s 101(2) of the Act). It is to be understood that the "prejudicial effect" of a particular item of evidence is the risk that a jury might in some way misuse it (Duckworth at [34]).
By reference to those matters, the applicant submitted that the trial judge had erred in finding that "the evidence does have significant probative value for the reasons that appear elsewhere; it outweighs the non-existent prejudicial effect" (at [196]). The applicant submitted that this error flowed from a failure to consider the particular fact in issue which was sought to be established by the inferential reasoning process involved in the use of coincidence evidence. The applicant submitted that the similarities between the three events themselves and the surrounding circumstances bore little or no relation to the applicant, which meant that the purported coincidence evidence could not meet the threshold test under s 98 because it lacked probative value, let alone significant probative value.
The applicant submitted that in the present case, the probative force of the proposed coincidence evidence was severely diminished by the lack of any direct evidence linking the applicant to any one of the so called "attacks". He submitted that if the Court were satisfied that the applicant had been responsible for one or more of the relevant injections, for example because someone had witnessed him administering that particular injection, then coincidence reasoning could be deployed to draw the inference that he was responsible for the others, even though there was no direct evidence. The applicant submitted that that proposition is reinforced by the wording of s 101(2) which proscribes evidence adduced by the prosecution "about a defendant" from being used "against the defendant" unless its probative value substantially outweighs any prejudicial effect it may have "on the defendant".
The applicant submitted that while it was true that a determination as to whether proposed coincidence evidence has significant probative value for the purposes of ruling on its admissibility under this section, may be reached having regard to all of the evidence adduced or sought to be adduced by the party seeking to tender it, it may also be necessary to have regard to an alternative explanation arising on the evidence (DJS at [78] per Whealy JA).
The applicant submitted that there was a possibility, reasonably open on the facts, that these offences had been committed by two or more people, whether in pursuit of a joint criminal enterprise or otherwise, and that the existence of this alternative explanation would further diminish the probative value of the proposed coincidence evidence.

Consideration

Section 98 relevantly provides: "98 The coincidence rule (1) Evidence that 2 or more events occurred is not admissible to prove that a person did a particular act or had a particular state of mind on the basis that, having regard to any similarities in the events or the circumstances in which they occurred, or any similarities in both the events and the circumstances in which they occurred, it is improbable that the events occurred coincidentally unless: (a) the party seeking to adduce the evidence gave reasonable notice in writing to each other party of the party's intention to adduce the evidence, and (b) the court thinks that the evidence will, either by itself or having regard to other evidence adduced or to be adduced by the party seeking to adduce the evidence, have significant probative value. …"
Section 101 of the Act relevantly provides: "101 Further restrictions on tendency evidence and coincidence evidence adduced by prosecution (1) This section only applies in a criminal proceeding and so applies in addition to sections 97 and 98. (2) Tendency evidence about a defendant, or coincidence evidence about a defendant, that is adduced by the prosecution cannot be used against the defendant unless the probative value of the evidence substantially outweighs any prejudicial effect it may have on the defendant. …"
The constraint sought to be imposed by the applicant on the application of s 98 is not justified by its wording or by authority. There is no requirement in the application of s 98 in this case for there to be evidence directly linking the applicant to any of the instances involving the wrongful injection of insulin. To impose such a requirement flies in the face of the wording of s 98 and the proper approach to a circumstantial case. While it is true that the fact in issue to which the coincidence evidence relates has to be clearly identified, the fact in issue does not necessarily have to involve a direct link between the applicant and one of the episodes of insulin injection.
It is clear from the Crown case and his Honour's judgment that a key issue at trial was whether the Crown could prove that one person alone was responsible for injecting each of the victims. It is for that reason that his Honour treated that issue as a matter which the Crown had to prove beyond reasonable doubt (judgment [199], [238], [262]). That issue was an essential intermediate fact which had to be established by the Crown to the criminal standard. The applicant acknowledged that proposition in his submissions.
A coincidence notice had been served by the Crown, relying upon similarities in each of the three injection of insulin events. This needed to be established as a step in establishing that the applicant did the acts set out in the indictment. The Crown's case was that it was improbable that these events and the circumstances occurred coincidentally and involved the acts of different persons. The Crown's position was that these were deliberate acts by the applicant alone. This was the issue to which the coincidence reasoning was directed.
For obvious reasons, there was no application to sever the indictment and so the evidence in respect of each of the three counts was properly before the trial judge in the trial, regardless of the coincidence notice. The issue, as his Honour understood, was whether coincidence reasoning could be utilised in respect of the three counts, the facts of which were already before him. Because this was a judge alone trial, it was agreed that a ruling in relation to the permissibility of using coincidence reasoning would be made at the end of the trial as part of his Honour's judgment.
The Crown in its closing address made submissions about the similarities in the three insulin injection events and the circumstances in which they occurred which pointed to the same person being the perpetrator in each case (T.803, AB 2118). The Crown went through the circumstantial evidence relied upon by it to establish that the single perpetrator was the applicant.
As his Honour noted, defence counsel at trial never articulated a challenge in the form of an objection to the use of coincidence reasoning. His arguments were confined in his closing address to the strength of the Crown arguments as to the similarity. For abundant caution, his Honour nevertheless indicated that he was satisfied that the evidence had significant probative value for the purposes of s 98 of the Act (judgment [195]-[196]) for reasons which appeared elsewhere in his judgment. He further ruled that s 101(1) of the Act was also satisfied (particularly in the context of a judge alone trial). As a result, he concluded that the evidence in relation to the individual counts was available to be utilised as coincidence evidence in relation to the other counts. Importantly, as his Honour again observed, there was little advanced by defence counsel at trial in his closing address in support of the contention that there might have been more than one person responsible for the three crimes beyond a bare assertion that that was possible (judgment [198]).
The trial judge found that the three crimes were "far from typical; they were confined to a very small pool of potential victims in a relatively confined place and in a relatively short timeframe". He found that they needed to be perpetrated by a person with particular knowledge and ability who had a connection with, and access to, the victims (judgment [199]).
The trial judge was satisfied that having regard to all of the circumstances and similarities, the conclusion that there was only one perpetrator was inescapable (judgment [199]).
The thrust of the applicant's challenge to his Honour's reasoning, as set out above, was that the similarities between the insulin injection events and the surrounding circumstances bore little or no direct relationship to him. It was on this basis that the applicant submitted that the coincidence evidence "lacked probative value, let alone significant probative value". This was the basis for the applicant's overall submission that in relation to this ground of appeal, the trial judge's use of evidence for coincidence purposes to infer one person was responsible for injecting each of the victims, was wrong.
As already set out, such an interpretation is not justified either by the wording of s 98 or by authority. His Honour's approach was in line with the approach of this Court in R v Ceissman [2010] NSWCCA 50 at [9] where the principal fact in issue ultimately was the identity of one of the offenders and whether the Crown could prove beyond reasonable doubt that it was the offender in that case. Latham J (McClellan CJ at CL and Schmidt J agreeing) said: "9 Having regard to Zhang, the first question to be determined was whether the evidence upon which the Crown relies as coincidence and/or tendency evidence is capable of rationally affecting the probability of the existence of a fact in issue. A fact in issue, in the sense that the Crown relies upon it to prove its case, is whether the offenders were the same on each occasion: see R v Fletcher [2005] NSWCCA 338 at [49] ; 156 A Crim R 308 at 319. The principal fact in issue ultimately was the identity of "the taller man". The second limb of the exercise (assuming an affirmative answer to the first) was to evaluate, in the light of all the available evidence, the likelihood that the jury would assign significant probative value to the coincidence and/or tendency evidence."
The principal fact in issue in this case was whether it was one person who injected all three victims, that person being the applicant.
The chain of reasoning advanced by the Crown was: 1. The coincidental nature of the "related events" allowed the inference to be drawn that the three offences were committed by the same offender; and 2. The circumstantial evidence relied upon by the Crown established the applicant to be that one offender.
This is in line with the approach in R v Ceissman where Latham J said: "15 In my view, the Judge erred in a number of ways. First, in considering whether the evidence was capable of rationally affecting the probability of the respondent's participation in the offences, the Judge conflated the two steps in the analysis set out in Zhang. The chain of reasoning sought to be advanced by the Crown is: (i) The co-incidental and tendency nature of the "related events" allows the inference to be drawn that the offences were committed by the same offenders. (ii) The respondent is one of the offenders in two of the robberies (DNA on items worn during the commission of those offences). (iii) The respondent was the taller of the two men in all five robberies (Mr McCarthy's evidence). (iv) The respondent is implicated in four of the robberies by circumstantial evidence (the "bad teeth" of one of the offenders described by two witnesses in respect of separate robberies, telephone intercepts between McCarthy and the respondent disclosing planning of the offences, and the pink gloves worn in two robberies). (v) The combination of the foregoing establishes the respondent's guilt of all of the offences beyond reasonable doubt. 16 The evidence of the "related events" is capable of rationally affecting the probability that the offences were committed by the same offenders (i). The test is one of capability, that is, is it open to the jury to conclude from the "related events" that the offences were committed by the same offenders: R v Shamouil [2006] NSWCCA 112 at [61] to [65]. The answer to that question in this case is clearly yes. The second stage of the analysis is to combine the "related events" with the other evidence relied upon ((ii) to (iv)) and pose the question, having regard to all of the evidence, is the jury likely to assign the evidence of the "related events" significant probative value. …"
The Court in R v MR [2013] NSWCCA 236 (Beech-Jones J with whom Hoeben CJ at CL agreed) made it clear how important it is to distinguish between the two steps: "95 The next step is to address s 98(1)(b). Consistent with the opening words of s 98(1), the fact addressed at this point is whether MR was involved in each of the offences (see Ceissman at [15] to [16]). To this point I have passed over the significance of there being a different number of offenders in the armed robberies, being four in the first and fourth, two in the second robbery and three in the third. However, the differing number of offenders is potentially a matter of some significance. All other matters being equal, it is a matter that weighs against the evidence of the events being significantly probative of MR's involvement in each offence. Nevertheless, as noted, each of Zhang (at [139]), Ceissman (at [16]) and most importantly DSJ (at [72]), recognises that the evidence of the events can be considered "on its own account, or by having regard to other evidence adduced or to be adduced by the tendering party". In this case, the uncertainty in terms of admissibility that arises from the differing number of offenders with the second and third armed robberies dissipates when regard is had to those parts of the evidence which directly connect him to each of those offences, albeit that evidence is of varying strength and leaving aside questions of credibility or reliability. Of course whether the jury would regard that evidence connecting him to each offence as overcoming any uncertainty that arises from the differing number of offenders is a different matter."
When one applies that reasoning to the facts of this case, the first step in the chain of reasoning was to establish that the same offender administered each of the three injections. The second step was laying out the circumstantial evidence upon which the Crown relied to show the applicant's involvement in the three offences. The third step was showing that the combination of those two matters meant that on the Crown case it was established beyond reasonable doubt that he was involved in all three offences.
It was open to his Honour to conclude in this instance that having regard to the similarities in the three insulin injection events, and the circumstances in which they occurred, the evidence was capable of rationally affecting the probability of a fact in issue, namely whether the same offender was involved in each offence. The test posed by s 55 was therefore satisfied. As was said in R v MR: "94 Further, the evidence relied on was capable of establishing other similarities, being the involvement of K in three of the four armed robberies and N and Z in two of them (and all of them in the carjacking). Otherwise it is to be remembered that the offences all took place in a short period of time in approximately the same area of Sydney. Applying the first stage of the approach from Zhang, as explained in Ceissman at [16], this evidence is capable of rationally affecting the probability of a fact in issue, namely whether the same group or subgroup of offenders was involved in each offence. The test posed by s 55 is satisfied.
The next step in addressing s 98(1)(b) of the Act and consistent with the opening words of that section is to address the question of whether the applicant was involved in each of the offences. However, most importantly at this step, the evidence of the events can be considered "on its own account or by having regard to other evidence to be adduced by the tendering party" (DSJ v R; NS v R at [72]; R v MR at [95]). This includes having regard to the circumstantial evidence that the Crown relied upon in the case against the applicant to connect him to the offences including for example, the process that was undertaken in eliminating other people and the circumstances that were addressed at judgment [238]-[261].
Having regard to all of the evidence adduced by the Crown, the evidence of the related events had the capacity to be of importance or of consequence in establishing the fact in issue, namely the involvement of the applicant in each of the offences. It was, therefore, clearly open to his Honour to find that the requirements of s 98(1)(b) and s 101 were satisfied and to use the evidence as coincidence evidence.
As earlier indicated, there is no requirement in order for the evidence to have significant probative value, that there be direct evidence linking the applicant to a particular injection incident. Circumstantial evidence can be just as convincing and reliable as a case based on direct evidence (El Hassan v R [2007] NSWCCA 148 (Hunt AJA; with whom Johnson and Latham JJ agreed) at [27]). Further the circumstantial evidence implicating the applicant could, as it did in nearly all instances in this case, relate to all three offences not just one of them. The exception to this was the text messages which related only to Ms Manuel and Mr Kelly.
The applicant's submissions on this issue conflate concepts relating to the admissibility of coincidence evidence and the process of reasoning that can then be deployed once evidence has been admitted for a coincidence purpose.
The applicant's submission that the combined operation of ss 98 and 101 of the Act precluded the use of coincidence evidence unless and until the trial judge was satisfied to the criminal standard of proof that the applicant was responsible for one or more of the insulin injection incidents considered separately should not be accepted. Sections 98 and 101 only relate to issues going to the admissibility of coincidence evidence. There was no requirement at the stage of determining admissibility for his Honour to make any determination of guilt in relation to any count beyond reasonable doubt.
In assessing significant probative value, his Honour was simply required to take the evidence at its highest and determine if it had the capacity to be of importance or of consequence in establishing the fact in issue (IMM v The Queen [2016] HCA 14; 257 CLR 300). Further, even in terms of deploying coincidence reasoning, where counts are cross-admissible as coincidence evidence, there is no requirement that the tribunal of fact must find one of the offences proved beyond reasonable doubt before they can use that count as a basis for coincidence reasoning in relation to other counts (Folbigg v R [2005] NSWCCA 23; 152 A Crim R 35 (Sully J with whom Dunford and Hidden JJ agreed) at [103]).
Ground of Appeal 1(a) has not been made out and should be dismissed.

Ground 1(b) - The trial judge erred in a process of "backwards reasoning" thereby reversing the onus of proof.

The applicant submitted that the trial judge engaged in a process of "backwards reasoning" (or "reverse reasoning") to establish the guilt of the applicant on all three counts and that this process was erroneous. The applicant submitted that in large part, the trial judge reasoned that the applicant was guilty of all three counts because he was effectively the "last man standing" after other suspects were excluded.
The applicant submitted that the reasoning process which was adopted by the trial judge was that which was employed by the investigating police to narrow down the list of potential suspects and later by the Crown in its case against the applicant. The applicant submitted that this mode of reasoning was flawed because:

It "put the cart before the horse".
It effectively reversed the burden of proof; and
It did not give proper weight at all to the surrounding circumstances and applicable directions of law (circumstantial evidence and the drawing of inferences).

In order to substantiate those submissions, the applicant relied upon the following extract from the judgment (judgment [238]) where the trial judge said: "238 The starting point in examining the possibility that the [applicant] was the person responsible for injecting each of the victims is to acknowledge that it is an undisputed fact that somebody did so and that I am satisfied that it was the one person. …"
By reference to that part of the judgment, the applicant submitted that the combined operation of s 98 and 101 of the Act precluded the use of coincidence reasoning unless and until the trial judge was satisfied to the criminal standard of proof that the applicant was responsible for one or more of the charged acts considered separately.
In support of that proposition, the applicant submitted that the question of whether in such a case it was necessary for the tribunal of fact to be satisfied beyond reasonable doubt as to the existence of each fact from which the inference of guilt is to be drawn, was considered by Dawson J in Shepherd v R [1990] HCA 56; 170 CLR 573 at 579 and 581: "4 Circumstantial evidence is evidence of a basic fact or facts from which the jury is asked to infer a further fact or facts. It is traditionally contrasted with direct or testimonial evidence, which is the evidence of a person who witnessed the event sought to be proved. The inference which the jury may actually be asked to make in a case turning upon circumstantial evidence may simply be that of the guilt of the accused. However, in most, if not all, cases, that ultimate inference must be drawn from some intermediate factual conclusion, whether identified expressly or not. Proof of an intermediate fact will depend upon the evidence, usually a body of individual items of evidence, and it may itself be a matter of inference. …" 5 … it may sometimes be necessary or desirable to identify those intermediate facts which constitute indispensable links in a chain of reasoning towards an inference of guilt. Not every possible intermediate conclusion of fact will be of that character. If it is appropriate to identify an intermediate fact as indispensable it may well be appropriate to tell the jury that that fact must be found beyond reasonable doubt before the ultimate inference can be drawn. … 8 … And of course it is quite correct to say that an intermediate fact which is an indispensable step upon the way to an inference of guilt, whether it be a fact derived from a single piece of evidence or a conclusion of fact drawn from a body of evidence, must be proved beyond reasonable doubt if the ultimate inference is to be the only reasonable hypothesis." (The applicant's emphasis)
The applicant submitted in relation to Ground 1 that the trial judge's use of the evidence for a coincidence purpose - to infer that "one person" was responsible for injecting each of the victims - constituted a wrong decision on a question of law and that this error gave rise to a miscarriage of justice in the sense of a departure from a trial according to law. The applicant submitted that this constituted a substantial miscarriage of justice because the trial judge then relied upon this inference to reason backwards to the guilt of the applicant, effectively reversing the onus of proof and thereby depriving the applicant of a chance of acquittal which was fairly open to him.

Consideration

Consistent with the process of reasoning that has been referred to in relation to Ground 1(a) above, what his Honour had to be satisfied of beyond reasonable doubt in this case, because it was an essential intermediate fact, was that the coincidental nature of the three insulin injection events allowed the inference to be drawn that the three offences were committed by the same offender and that the circumstantial evidence relied upon by the Crown implicated beyond reasonable doubt the applicant as that offender. This was the approach that his Honour adopted. It clearly does not involve as contended by the applicant his Honour engaging in a process of "backwards reasoning".
Implicit in the first ground of appeal is the proposition that there was "a possibility reasonably open on the facts that these offences had been committed by two or more people whether in pursuit of a joint criminal enterprise or otherwise" and the existence of this alternative explanation further diminished the probative value of the proposed evidence.
It is accepted that the trial judge in forming a view as to whether the evidence had significant probative value, must have considered by reference to all of the Crown's evidence, whether there emerged a real possibility of an explanation inconsistent with guilt of the party tendering it. The court must ask whether that real possibility substantially alters its view as to the otherwise significant capacity of the coincidence evidence to establish the fact or facts in issue. However, it is clear from his Honour's reasoning and the circumstances and similarities which his Honour took into account, that his Honour did not consider the alternative raised by the applicant as a real possibility. Such a finding beyond reasonable doubt was clearly open to his Honour.
This ground of appeal has not been made out.

Ground 2 - The verdicts are unreasonable and cannot be supported by the evidence.

(i) The Crown timeline in Exhibit D.

The applicant submitted that although defence counsel at trial did not object to the admission of Exhibit D, this document only had evidentiary value insofar as the "one perpetrator" theory (which was submitted to be erroneous) was accepted. The applicant submitted that nevertheless the trial judge erred in accepting what was thought to be established by Exhibit D because the exhibit had a number of errors and was flawed in how it was set out, which rendered it unreliable.
The first issue identified by the applicant was that the timeline upon which Exhibit D was based was incorrect. The applicant submitted that the evidence concerning the onset of hypoglycaemia in the case of Ms Fowler, was incorrectly assessed because it did not have regard to the likelihood that she was experiencing symptoms at lunchtime. The applicant submitted that if such symptoms were present at lunchtime, then the timeline established by Exhibit D was wrong and that the injection of insulin to Ms Fowler could have been administered earlier than the earliest time considered by investigating police and set out in Exhibit D, i.e. 4am.
The applicant submitted that there was support in the evidence of Professor Carter and Professor Yue for the proposition that the timeline provided for in Exhibit D should have been much more extensive than from 5am on Friday, 18 October to 6pm on Saturday, 19 October. In those circumstances, the applicant submitted that the time at which Ms Fowler started exhibiting symptoms of hypoglycaemia was an intermediate fact which was an indispensable step upon the way to an inference of guilt.
The applicant submitted that the Crown case depended to a significant extent upon establishing the following matters as "links in a chain": 1. The identity of the persons who were present at the facility at some time over the relevant timeframe. 2. The exclusion of those persons who were there on only one of the two days in question. 3. The exclusion of those persons who were present on both days, but outside of the relevant timeframe in respect of one or more of the victims.
The applicant submitted that the Crown's failure to factor into the timeline that Ms Fowler started becoming hypoglycaemic at lunchtime was fatal to the Crown case. This was because the possibility that she had been injected as early as 4am on 18 October could not be excluded as an inference reasonably available on the evidence.
The applicant submitted that apart from that problem, there were other difficulties affecting the accuracy of Exhibit D. The applicant submitted that his Honour erred in relying upon Exhibit D because it was not a sound basis for inferring what it sought to establish, i.e. the identity of the people whose presence at the facility during the relevant times cast them under suspicion.
The applicant submitted that it was wrong to exclude the 35 people who could not be identified on the basis that they were not present at the aged care facility on both days. The applicant submitted that such a finding amplified the error in accepting the "one perpetrator" theory and diminished the proper focus which should have been directed to each count individually. The applicant submitted that those problems with Exhibit D rendered it unreliable and not a firm basis from which to draw inferences and make findings.

(ii) Further submissions on expert evidence

The applicant submitted that there were a number of other issues with the expert evidence which severely limited the support which Exhibit D could provide for the case against him.
In support of that submission, the applicant analysed the evidence of Dr Song and set out reasons why his conclusions were unreliable and that it was not open to his Honour to find that the insulin found in the blood serum samples of Ms Manuel and Mr Kelly had "an identical chemical makeup". The applicant submitted that Dr Song's testing was limited and therefore it was not open to make that finding on the evidence.
The applicant challenged the findings by his Honour that he had access to insulin. The applicant submitted that there was no evidence to suggest that the insulin in fact came from any of the treatment rooms in the facility so that the fact that he could access these rooms was largely irrelevant. The applicant submitted that alternatively, there was no evidence or suggestion that insulin was difficult to obtain or that there would have been any difficulty in him bringing insulin to the facility undetected. The applicant submitted that on that alternative basis, the inference that he had some kind of special access to insulin which others did not, was of little consequence.

(iii) The conversation with Mr Kelly

The applicant noted that investigating police were able to speak to Mr Kelly on 22 October 2013 before he died and that they asked him who had given him the injection. The applicant relied on the following conversation between Mr Kelly and Detective Little which was recorded by Detective Stephen in his notebook: "Little: Who gave you the injection? Kelly: He pulled it out. Little: Was it a nurse who injected you? Kelly: No (shook head). Little: Was it a doctor who injected you? Kelly: No it wasn't a doctor. Little: Do you remember someone injecting you? Kelly: (Moaning) (Little passes Kelly a pen and some paper) Little: Can you write down who injected you? (Kelly tries to write but struggles) Little: Was it a man or a woman? Kelly: Woman. (Observed Kelly wrote "Coldy") Kelly: Injected three days ago. Little: Did the person have black skin or white skin? Kelly: Radio (Kelly drew symbols that looked like hands and a cross) Kelly: Yes syringe I remember. Little: Do you know the person who injected you? Kelly: I think about it tid [sic] morning. Little: Do you know who made you sick? Kelly: The other people, they saw it there the same situation."
The applicant also referred to the evidence of Mr Kelly's granddaughter, Ms Kathleen Lowe, who visited him on a number of occasions when he was in hospital. She gave evidence that when she went to see him on 19, 20 , 21 and 25 October he was not responsive. On 24 October, which was his birthday, he regained some form of consciousness but was unable to respond to her verbally. She was able to converse with him on 27 October but he was unable to recognise her and thought that her husband was one of his grandsons.
In cross-examination, she agreed that Mr Kelly had not been mistaken as to her husband's gender, but had just mistaken him for another male.
The applicant submitted that the evidence of Mr Kelly that a "woman" administered the injection to him should have been given weight in the judgment of the matter for the following reasons: 1. Ms Lowe gave evidence that Mr Kelly's condition was variable and she did not visit him on 22 October and was therefore not in a position to corroborate the observations of the detectives. 2. Detective Stephen conceded in cross-examination that there was nothing in his statement about Mr Kelly crying, or his eyes rolling from side to side. 3. Detective Stephen conceded that Mr Kelly was responsive to the nature of the questions he was being asked. 4. Detective Little had taken some photos of Mr Kelly on a camera phone. Detective Stephen agreed in cross-examination that in hindsight it would have been prudent to record the conversation with Mr Kelly on video so as to allow the court to make its own assessment of his lucidity or otherwise.
The applicant submitted that Mr Kelly's recorded answers were responsive to the nature of what he was being asked. The applicant drew attention to his response on 22 October that the injection had taken place "three days ago", i.e. 19 October. This was correct according to the expert medical evidence. The applicant submitted that this evidence gave rise to a reasonable possibility that Mr Kelly had been injected by a woman and that this evidence should have engendered a real doubt that he was responsible for injecting Mr Kelly.

Factors said to distinguish the applicant

The applicant noted that in his reasons the trial judge had listed factors which he regarded as inculpating him by comparison with other people who were present at the aged care facility during the relevant timeframes. The applicant submitted that these factors provided very little support for the Crown case, particularly if his Honour was wrong in adopting the "one perpetrator" theory.
The applicant challenged the following factors:

(a) Presence

The applicant challenged his Honour's finding (judgment [239]) that the applicant was on duty in the Mountview ward when the injections were given. The applicant noted that issue had already been taken with the Crown timeline in relation to Ms Fowler and that there were a number of other employees of the aged care facility in the Mountview ward at those times.

(b) Tendency Type Reasoning

The applicant referred to the judgment at pars [241] and [246] which contained the following findings: "241 … Thus, he continued to demonstrate his preparedness to inject insulin into residents right up to the time of the injections given to the three victims. …" … 246 … he was involved in … the unauthorised activity of giving insulin injections to insulin-dependent residents …"
The applicant submitted that in those findings the trial judge erred by engaging in tendency type reasoning, the relevant tendency being that of the applicant to inject residents with insulin when he was not authorised to do so. By reference to ss 97(1) and 101(2) of the Act and the statements of principle by the High Court in Hughes v The Queen [2017] HCA 20; 92 ALJR 52 the applicant submitted that his Honour had used the evidence about the applicant continuing to inject patients, even after the change in policy, for a tendency purpose. The applicant submitted that this had the effect of affording the evidence a significance which was not appropriate and which consequently had a bearing on the ability of the evidence, taken as a whole, to reasonably support the convictions.
The applicant submitted that in line with Hughes v The Queen, two matters should have been considered by his Honour for the purposes of s 97(1)(b), namely: 1. The extent to which the evidence supported the tendency; and 2. The extent to which the tendency made more likely the facts making up the charged offence.
The applicant submitted that even if the evidence provided some support for the inference that he had a tendency to inject insulin-dependent residents in spite of the fact that he was not authorised to do so, that tendency did not make it more likely that he administered the relevant injection to the victims or in circumstances where he intended to cause harm. The applicant submitted that there was a material difference between injecting a person who required insulin in breach of the new facility policy and injecting a person who did not require insulin when the likely result would be death or serious injury. The applicant submitted that primarily because of this difference, using any evidence of breaching an internal insulin policy for a tendency purpose risked giving the evidence too much weight. The applicant submitted that the evidence did not have significant probative value and that accordingly s 97 ought to have prevented the evidence being used for a tendency type purpose. The applicant also noted that this was not a factor that was peculiar to the applicant in that he was not the only team leader who had continued administering insulin to patients (as medically required) after the policy changed earlier that year.

The applicant noted that his Honour took into account that he was found in possession of needles, syringes and information about insulin when his home was searched on 5 December 2013 (judgment [242]-[244]). The applicant submitted that his Honour gave excessive weight to this material, particularly when one took into account that no video recording of the execution of the search warrant was included in the exhibits. This left many unanswered questions in relation to what other material was found in the applicant's room and in the "black bag".

(d) Opportunity - access to insulin and relationship with victims

The applicant submitted that having access to insulin was not a factor peculiar to the applicant. The applicant submitted that before his Honour could find at [246] that he "clearly had the opportunity [to give the injections], probably more than anyone else" a more precise timeframe for when the injections were actually administered should have been proved. The applicant submitted that without knowing whether Ms Fowler and Ms Manuel were injected before the applicant's arrival on the mornings in question, it was erroneous to infer that he had the "opportunity" to administer the injections. The applicant submitted that it was also not open to his Honour to find that he probably had more opportunity to administer injections than anyone else in circumstances where there were registered nurses on duty at the time in the Mountview ward.
The applicant submitted that his Honour erred in accepting the Crown's characterisation of the applicant as the victims' "primary medication giver" so that "they would be less likely to object to him giving them an injection than anyone else, aside from a registered nurse". The applicant submitted that there was no basis in the evidence to suggest that the victims would have necessarily known the applicant's position as a team leader. In support of that submission, the applicant relied on the observations by Ms Fowler's general practitioner as to the extent of her dementia and that the other two victims also suffered a degree of dementia. The applicant submitted that the applicant was not the only person who had a relationship with the victims which included providing them with their medication.

(e) The view of the trial judge in relation to these factors

The applicant submitted that it was not open to his Honour to make the findings at [252] of the judgment and in particular, that "nursing and care staff who were present at the relevant times, remained possibilities but there is nothing to suggest anyone of them in particular might have been responsible". The applicant submitted that the effect of that finding was to use the absence of evidence in respect of other potential perpetrators to make up for a shortfall in the evidence which could be used to inculpate him. The applicant submitted that this effectively reversed the onus of proof. The applicant submitted that the only circumstance which was peculiar to the applicant related to the items which were recovered from his residence when police executed the search warrant on 5 December 2013.

(f) The text messages

The applicant noted that his Honour summarised the evidence relating to the text message exchanges with Ms Wilson, Ms Collins and Ms Christensen at [147]-[181] of the judgment and set out his conclusions in relation to the text messages at [253]-[261].
By reference to that material, the applicant submitted that his Honour was correct to disregard the text messages sent by him to Ms Christensen as equivocal but that his Honour still fell into error in finding that the balance of the messages supported the inference that he had foreknowledge of the impending or prospective demise of Ms Manuel and Mr Kelly because he knew that they had both received injections of insulin.
In order to substantiate that challenge, the applicant relied upon the following: 1. His Honour failed to have sufficient regard to the different contexts in which the applicant made his "inconsistent" statements. The applicant submitted that there was a material difference between an exchange in the nature of "black humour" which took place between colleagues and statements to police in the context of an investigation into the death of three people. 2. His Honour unreasonably downplayed the extent to which these "black humour" exchanges were regarded as routine by those working in the industry. 3. The proposition that the person who had actually committed these offences would send messages of this kind was speculative, if not illogical. 4. The purportedly "predictive" messages were confined to trying to guess who was going to be the third person to die in circumstances where two had already died. 5. These messages were instigated by Ms Wilson. 6. Ms Wilson gave evidence that she too believed that Mr Kelly was unwell at the time. 7. The inference that the text messages indicated special knowledge on the part of the applicant was difficult to sustain.
The applicant submitted that if a person had been responsible for killing three people and had sent texts "predicting" two of their deaths, those messages would have been at the forefront of the mind of that person whereas here, it seems obvious that he had forgotten about these texts altogether.

(g) Shortcomings in the Crown case: lack of motive and inadequacy of the investigation

The applicant accepted that the Crown was not required to establish motive in order to obtain a verdict of guilty. He submitted that nevertheless in this case the failure of the Crown to establish a motive on his part was significant. This was in a context when there was no direct evidence which provided positive support for the case against him and the circumstantial case depended on the adoption of the "one perpetrator" theory.
The applicant submitted that his Honour's observation at judgment [237] that "the evidence is what it is. It is not a matter of there being potentially exculpatory evidence that was not available" tended to obfuscate the fact that the Crown case against the applicant was weak so that the shortcomings in the police investigation were a relevant matter for the trial judge to take into account when considering whether it had been proved beyond reasonable doubt that the applicant was the same person who had committed all these offences. The applicant submitted that even if the concept of reverse reasoning were a proper approach, given the deficiencies in the Crown case, the state of the evidence and the erroneous timeline, investigating police had not successfully eliminated other available individuals and/or suspects.
The applicant submitted that as defence counsel at trial had observed in his closing address, the items found at the applicant's residence underscored the imperative of carrying out a physical search of the SummitCare facility as soon as possible. The applicant submitted that any assertion that the police considered and explored all possibilities during their investigation should have been viewed with suspicion by the trial judge, given how little time it took for a search warrant to be executed at the applicant's home. This was particularly so in the absence of any evidence to suggest that police scrutinised other potential suspects to the same degree as the applicant.
The applicant submitted that for all these reasons articulated in support of Ground 2, the verdicts were unreasonable and not supported by the evidence at trial. The applicant submitted that it was not open to his Honour to be satisfied beyond reasonable doubt as to his guilt.

Consideration

The principles upon which an appellate court approaches a ground of appeal from a jury verdict are well settled and are as set out in SKA v The Queen [2011] HCA 13; 243 CLR at [11]-[14] and the authorities therein. The task of this Court is to review the whole of the record of the trial and to make an independent assessment of the evidence, both as to its sufficiency and its quality. If, after giving full weight to the primacy of the tribunal of fact, the appellate court is left in reasonable doubt as to the verdict, it is only where the advantage of the tribunal of fact in seeing and hearing the evidence is capable of resolving that doubt that the Court can conclude that there was no miscarriage of justice.
As Hayne J said (Gleeson CJ and Heydon J agreeing) in Libke v The Queen [2007] HCA 30; 230 CLR 559 at [113]: "113 But the question for an appellate court is whether it was open to the jury to be satisfied of guilt beyond reasonable doubt, which is to say whether the jury must, as distinct from might, have entertained a doubt about the appellant's guilt. It is not sufficient to show that there was material which might have been taken by the jury to be sufficient to preclude satisfaction of guilt to the requisite standard."
The Crown case was a circumstantial one. As such, the evidence must not be analysed in a piecemeal way. The whole of the evidence, and the inferences which are available to be drawn from it, must be considered and weighed (R v Hillier [2007] HCA 13, 228 CLR 618 at [46]). An approach to the assessment of a circumstantial case which concentrates upon certain parts of the evidence to the exclusion of others is erroneous (Gilham v The Queen [2012] NSWCCA 131; 224 A Crim R 22 at [466]; Perish, Anthony v R; Perish, Andrew v R; Lawton, Matthew v R [2016] NSWCCA 89 at [428]).
In Dickson v R [2017] NSWCCA 78 Bathurst CJ (with whom Johnson and Fullerton JJ agreed) said at [86]: "86 In considering the issue in a case such as the present where the Crown relies on circumstantial evidence, it is important to bear in mind that the task of the appellate court is to consider and weigh all the circumstances in considering and deciding whether there is an inference consistent with innocence reasonably open on the evidence. The evidence is not to be looked at in a piecemeal fashion: R v Baden-Clay (2016) 90 ALJR 1013; [2016] HCA 35 at [46]-[48]."
The decision of Filippou v The Queen [2015] HCA 29; 256 CLR 47 makes clear that those principles also govern an appeal like the current one, invoking the first limb of section 6(1) from a finding of guilt after a trial by judge alone: Bell v R [2017] NSWCCA 207 at [20]-[23]; Haines v R [2018] NSWCCA 11 at [12]).
To the extent that the argument in support of a ground of appeal under s 6(1) takes issue with the correctness of the trial judge's intermediate findings of fact, this Court must have regard to the reasons stated by the trial judge; Filippou v The Queen at [12], [81]-[83]. "[12] … It follows perforce of s 133(1) of the Criminal Procedure Act that, in the case of an appeal against a judge's finding of guilt, the finding is not to be disturbed under the first limb of s 6(1) of the Criminal Appeal Act unless there is no or insufficient evidence to support the finding, or the finding is otherwise unreasonable or the evidence was all one way …"
In this case the applicant takes issue with a significant number of his Honour's intermediate findings of fact. It is, therefore, necessary to have regard to his Honour's reasons. The question for the Court is not whether it is satisfied that his Honour's findings of fact are correct, but whether his Honour's findings were not reasonably open (Filippou v The Queen at [56]).
For ease of reference, I have used the same system of headings as was used in the applicant's submissions.

(i) Exhibit D

The injection of the three victims took place on 18 and 19 October 2013. A spreadsheet was prepared by police in order to show the times at which various people were in attendance at SummitCare facility from 5am on Friday, 18 October to 6pm on Saturday, 19 October 2013 in accordance with the CCTV footage from the facility. This spreadsheet became Exhibit D.
It was tendered without objection from the applicant's counsel and the Crown stated when it was tendered: "And I can indicate to your Honour with respect to Exhibit D that there is no dispute between the parties as to the accuracy of the particular document, so what is recorded times and persons your Honour is not in dispute." (T.78, AB 1393)
Consistent with this, his Honour noted in his judgment at [110]: "110 … There was no dispute with the accuracy of the content of this spreadsheet, although some minor typographical and inconsequential errors were noted during the trial."
The spreadsheet was prepared by police upon the basis of the theory that one person was responsible for all of the injections and in light of the expert opinion that they occurred no more than eight hours before each victim was found in a significant hypoglycaemic state (judgment [114], [115]).
Based on these premises, it was made clear at trial that police identified 25 potential suspects (including the applicant) who could be responsible. These people were recorded on Exhibit D in red italics (T.88, AB 1403). His Honour also created and attached to his judgment, using Exhibit D, a more concise spreadsheet showing only the persons present on both days. Importantly, there was no cross-examination of the officer in charge by defence counsel at trial in respect of Exhibit D.
Despite the approach adopted at trial, the applicant now submits that his Honour erred in accepting what was sought to be established by Exhibit D because the timeline upon which the spreadsheet was based was wrong in relation to Ms Fowler. It is also now separately asserted that Exhibit D "contained errors and was otherwise confusing". Despite that latter assertion, these errors were not further identified in submissions nor was it articulated how the document was confusing. Accordingly, it is not proposed to further address that aspect of the submission.
In order to challenge the timeframe for Ms Fowler's injection, considerable reliance was placed upon one answer by Professor Carter, one of a number of experts called in respect of the issue of the timing of the injections. The timeframe relied upon by police and also found beyond reasonable doubt by his Honour on the basis of all of the medical evidence was that the three injections all occurred at most eight hours before significant hyperglycaemic symptoms were displayed by each of the victims (judgment [262]).
His Honour concluded that he was satisfied on the basis of "reasonable conformity to the expert opinions" that the injections occurred within eight hours of the relevant victim displaying significant symptoms of hyperglycaemia (judgment [105]). His Honour observed that it was quite likely that the period was shorter but that it seemed safer to act upon the eight hour period.
There was no issue that Ms Fowler was displaying significant hyperglycaemic symptoms when she was found cold, clammy and unresponsive at about 4pm. However, the applicant asserts that Ms Fowler's behaviour at lunchtime (getting frustrated and trying to bite a care giver) if considered abnormal for her, might indicate that this was the "onset of hyperglycaemia" or when she started "becoming hyperglycaemic". The applicant therefore contends that the eight hours can be taken backwards from lunchtime to the effect that there was a possibility that the injection was administered as early as 4am. The relevant answer from Professor Carter needs to be carefully examined in its entirety and also in the context of subsequent answers given by him.
Professor Carter said unequivocally that Ms Fowler was hyperglycaemic about eight hours after breakfast. He only noted that there were big question marks as to whether there were "features of hyperglycaemia" four hours after breakfast, i.e. at lunchtime. Professor Carter later expressly stated that in the case of Ms Fowler, it was "likely that the injection was given around breakfast time, or some time after that".
The relevant evidence from Professor Carter was as follows. The reference to "BSL" in the evidence is to "Blood Sugar Levels": "Q. Through the morning there was an incident at lunch where she yelled and may have tried to bite somebody. There's confusion as to whether that's normal or aberrant for her, but that's pretty much her observations through the day, breakfast and the like, correct? A. Correct. Q. Knowing that fasting BSL at 8 o'clock at 5.8 and knowing her found at 3.55 in clearly hypoglycaemic, hypothermic situation, does that assist you with that other information, professor, to put a time or a more precise time on your view or your opinion as to when she was given exogenous insulin? A. Yes, it does. The biggest problem here is not knowing what insulin was administered. There are three main possibilities. One is that she was just injected with a short-acting insulin. Another, that she was injected just with a long-acting insulin. And the third is that she was injected with a mixture of both short and long-acting insulins. The timeframe of action of those insulins has been summarised in my paragraph 6. It appears that almost eight hours after breakfast Mrs Fowler was found to have symptoms strongly suggestive of hypoglycaemia, low blood sugars, and subsequent blood tests confirmed that the blood sugar was low. At lunchtime, so four hours after breakfast, the erratic behaviour and perhaps attempting to bite somebody, if that were regarded as abnormal, that could be a feature of hypoglycaemia, but if that was not regarded as abnormal behaviour, there would be no clinical suggestion of hypoglycaemia. So unequivocally she was hypoglycaemic about eight hours after breakfast. Big question marks as to whether there were any features of hypoglycaemia four hours after breakfast. So if we accept that her behaviour at lunchtime was within normal limits for her, we're looking at what insulin and when it was administered could account for the hypoglycaemia around 4pm. So if she were given insulin at 8am and she didn't really develop features of hypoglycaemia for nearly eight hours, that is not compatible with the injection of just short-acting insulin at 8am. It would be compatible with an injection at lunchtime, so four hours earlier, but eight hours later it would be compatible with an injection of mixed insulin around breakfast time, 8am, even though that's got 25 per cent or 30 per cent of the total amount of insulin is short-acting insulin in the commercially available mixed insulins, it could be that mixed insulin was given at 8am. I don't think we can say that hypoglycaemia did not develop until eight hours after breakfast. It presumably developed somewhere between four and eight hours after if it's regarded that she was -her behaviour was normal for her at 12 noon. So it could be that she had been sweaty and low temperature and shaky for some time before she was found at 3.55pm. So my conclusion from the summary you've given is that the hypoglycaemia Was consistent with an injection around breakfast time of a mixed short and long-acting insulin or a long-acting insulin, but not just a short-acting insulin, or an injection of an insulin at lunchtime. It could be short-acting alone, it could be a mixed insulin and it could be a long-acting insulin but less likely in view of the timeframes, but I couldn't exclude that." (T.695.49-696.48, AB 2009-2010) "Q. Paragraph 5 you talk about, third line in, "within eight hours, most likely between half an hour and six hours"? A. Yes. Q. And then you refer to that being NovoMix 30 and Mixtard 30/70 and the like, correct? A. Correct. Q. And today you've talked about the possibility of at lunch and at breakfast - A. Yes. Q. -- and the like. Are you saying, just so we understand, is it within or at are You're talking about? A. I certainly cannot say at, and I did mention that I can't exclude various other times, and perhaps I should emphasise that it's certainly within, when we're talking about, I think, Mrs Fowler and, well the two people who were found hypoglycaemic in the middle of the afternoon, it is likely that the injections, injection or injections were given around breakfast time or sometime after that, and when I suggested that it is unlikely to be short acting insulin given at breakfast, but it could be short acting insulin given at lunch, perhaps I should've expressed it differently, saying it could've been short acting insulin given just before lunch, in view of the timeframe of action of short acting insulins, and if it were just a long acting insulin, I did mention they could've been given at 4am or something like that, I can't exclude that. Certainly can't exclude a mixed insulin being given in the middle of the morning, 10am, 11am, mixed insulin having a significant component of, or a proportion of short acting insulin. We could just concentrate on the timeframe of action of the short acting insulins. If we're thinking purely of short acting versus a mixed insulin, the length of, which we haven't discussed so far, the length of, from memory, the hypoglycaemia that was induced with the blood sugars being low once they were in hospital, the length of hypoglycaemia makes it unlikely that the injections were, irrespective of what time they were given, were just purely short acting insulin, almost certainly they have to be at least a component of long acting insulin. But I would conclude what I said in my report, that it's likely that the injections were given within eight hours, and I certainly cannot say at breakfast or at lunch. So my apologies, your Honour, if it appeared I was changing my view." (T.701.32-702.17, AB 2016-2017)
Furthermore, the other expert who gave evidence of the time period when Ms Fowler was injected, Professor Dennis Yue, opined that Ms Fowler was "almost certainly" injected after breakfast that morning (T.577). Professor Yue's evidence was that if Ms Fowler's irritability during lunch was out of character, that would indicate that she was probably already "mildly hypoglycaemic" at that time and it was consistent with the progressive onset of symptoms. In contrast, the state she was found in during the afternoon (around 4pm) was consistent with very typical symptoms of "severe hypoglycaemia".
The evidence of Professor Yue was: "Q. Now, can I ask you this? If her BSL that morning taken pre-breakfast was 5.8, that's something you did not know when you wrote your report, isn't it? A. No. Q. All of those factors, does that impact at all upon your timeframe of potential injection to Mrs Fowler? A. Yeah, I think all these factors will add up the fact that the timeframe of the injection counting back from the time that she was found to be hypoglycaemic is likely to be less than 24 hours. 24 hours a very general statement because I don't know what type of insulin that the - and her general condition, but knowing that the fasting glucose that morning 5.8 was normal, she was able to walk and converse and had lunch that day, I would guess that the time that she had been given an injection of insulin, right, is considerably less than 24 hours and almost certainly after breakfast that morning. I would put it down to several hours. Q. When you - A. Sorry, and the fact that she was irritable at lunch time would suggest that she was starting to become hypoglycaemic. You'll remember that I said people can become irritable and behave in an irrational way and so I think that she probably was already mildly hypoglycaemic round about lunch time and if she had been given short acting insulin, then that will be - timing would be just before lunch. Q. Can I ask you this, sir, just so I understand? You are opining for us, professor, that she was most likely injected after she ate on that breakfast morning? A. That's correct, yeah. Q. And you mentioned to me earlier that some of the symptoms heading towards - I'm sorry, I withdraw that. The description that her daughter gives and the nurse Zubcakova give, that is - would it be fair to describe that as, what, profoundly hypoglycaemic at that stage; unconscious, sweating, hypothermia, those things? A. Yes, yeah, the sweating, which I did not - wasn't aware of before, was a very typical symptoms of severe hypoglycaemia. Q. Now, but the irritability or potential irritability at lunch time, that's that progressive thing that you and I spoke about earlier? A. Consistent with that." (T.577.6-.45, AB 1892) "Q. Having those factors now given to you professor, does that assist you in any way of giving us a time frame with respect to when Mrs Manuel could have been injected with the insulin? A. Knowing the onset of hypoglycaemia, it's a bit imprecise because depend on when you want to call it, but let's say we go on the time of 11.30 when she was clearly have evidence of hypoglycaemia okay, knowing her body weight being very low, knowing her age now and knowing that she did have breakfast I think you said yeah, I would guess that the time of insulin injection would be six hours, eight hours, no more than that before she became clearly hypoglycaemic at 11.30." (T.582.43-583.2, AB 1897-1898)
In addition to the opinion of Professors Carter and Yue, it needs to be kept in mind that when Ms Fowler became frustrated and attempted to bite AIN Matthews at lunchtime, AIN Matthews was untroubled and said to the applicant that this would sometimes happen as Ms Fowler would occasionally become frustrated when being encouraged to eat and drink. That observation is important because AIN Matthews was a person who was quite familiar with Ms Fowler on a day to day basis and certainly more familiar with her than her general practitioner, Dr Gay. It is also significant that AIN Avery observed that Ms Fowler appeared her usual self throughout the rest of the afternoon following lunch.
It follows from that evidence that the applicant's contention that Exhibit D was unreliable and was not reasonably open to be acted upon by his Honour, because it failed to factor in the possibility that Ms Fowler had been injected as early as 4am, is not sustainable. It was clearly open to his Honour to conclude on the basis of the expert opinions and the lay evidence that each of the three victims, including Ms Fowler, were injected within eight hours of them displaying symptoms of significant hypoglycaemia (judgment, [105]).
In a further written submission received after oral argument in the appeal, the applicant submitted that by reference to Exhibit D, if the timeframe of eight hours in relation to Ms Fowler were to extend backwards from 12 midday, there were would be five other employees of SummitCare who could have administered an injection of insulin to Ms Fowler. They were RN Blaston, AIN Agha, AIN Raylene Collins, AIN Debbie Wilson-Jones and AIN Long. The applicant submitted that those persons were present on both days and had not been excluded by the Crown's process of elimination from being a possible perpetrator of the offences. They were persons whose shift ended at approximately 7am on Friday, 18 October 2013. The applicant submitted that the information available did not identify with any precision where those employees were working over the two days. The applicant submitted that this was a further indication of the inadequacy of Exhibit D and the problems with the arbitrary eight hour limit found by his Honour to be appropriate for the likely administration of the insulin injection.
Despite the additional detail, this submission goes no further than to again challenge his Honour's finding that the same person was responsible for all three injections and to challenge his Honour's finding, based on the expert evidence, that the appropriate timeframe for the injection of insulin in respect of each victim was within eight hours of them developing significant hypoglycaemic symptoms. The reasons why these findings were well open to his Honour and why these further submissions take the matter no further have been set out above.

(ii) Other submissions on expert evidence

The applicant's submissions were directed to the evidence which was led at trial from Dr Song about testing that had been carried out on particular types of insulin. The applicant submitted that overall Dr Song's evidence was of limited relevance.
It is clear from his Honour's judgment that he treated the evidence of Dr Song as of very little, if any, significance. At [91] of his judgment, his Honour said: "91 Dr Xiaomin Song, an analytical biochemist gave evidence about testing samples from each of the victims at the Australian Proteome Analysis Facility. As an attempt to identify the type of insulin injected it was far from definitive; for example, no attempt was made to test for any long-acting only types of insulin, only short-acting and mixed types (and only some of those)." His Honour later observed, when addressing the similarities that the Crown had attempted to rely upon, that Dr Song's testing was unreliable and that he would "put this further asserted similarity aside" (judgment [184]-[186]). Accordingly, this submission on behalf of the applicant goes nowhere. His Honour effectively ignored the opinion of Dr Song.
In this part of his submissions, the applicant also sought to challenge his Honour's observation that defence counsel at trial did not take issue in his closing address with the proposition that the injections must have occurred within a timeframe of eight hours before the display of significant hypoglycaemic symptoms.
In fact the concession identified by his Honour was made by defence counsel in the following terms: "HIS HONOUR: Going back to my question, is it correct to say that, having regard to the inability to be precise, because one doesn't know exactly the type of insulin that was injected, and also that there are a range of variables in terms of different people, their ages and their weight and all sorts of things, that accommodating all of the variables, the opinions are generally to the effect that it would be eight hours or less. So all those things are factored in to the timeframe of eight hours or less. WATSON: I think that's right and I think that that was what I'll call a general consensus with the experts. We say that that's not optimal situation that the best testing, the complete testing would have been of great assistance, particularly to determine the source of insulin itself. For example, there was and coming back to the insulin at the facility, that there was no stocktake of that even when it was able to be counted and catalogued." (T.841.46 832.9, AB 2156-2157).
The applicant further contended under this heading that the evidence at trial regarding the type of insulin which was used did not provide a sound basis for inferring that the insulin came from any of the treatment rooms at the facility. The applicant submitted that therefore his Honour's finding that he had access to insulin as one of the pieces of circumstantial evidence, was not available on the evidence. The applicant further submitted that there was no evidence to suggest that the insulin came from any of the treatment rooms in the facility so that the fact that the applicant could access those rooms was "largely irrelevant".
The applicant's submissions on this issue are misconceived. His Honour accepted that the exact type of insulin that was injected in each instance could not be identified. The experts could only express a general opinion about whether it was short acting insulin, long acting insulin or a combination of two. The overall effect of the expert opinions was that each of the victims had been injected with either a long acting insulin or a combination of long acting and short acting insulin. The key issue that his Honour had to be satisfied of, based on the expert evidence, was the relevant timeframe within which each of the injections occurred (judgment [105]).
There was evidence that insulin of varying types was kept in a refrigerator in the treatment room in each of the wards. As a team leader, the applicant had a key to each of these rooms. No records were kept by the facility for insulin stocks and usage. It was therefore open to his Honour, even though the precise type of insulin could not be identified, to take into account as a piece of circumstantial evidence that the applicant had access to insulin and that given his possession of a key, the applicant unlike some other staff members, could access it at any time without raising suspicion (judgment [245]). This is so given that it was an integral part of the applicant's job to obtain oral medications from the treatment room to give to residents in the ward at breakfast and at lunchtime (judgment [246]).
None of the issues raised under this heading has been made out.

(iii) Conversation with Mr Kelly and the nomination of a woman

The applicant submitted that the conversation by police with Mr Kelly on 22 October 2013 while he was in hospital where he purportedly nominated a "woman", who was not a doctor or nurse, as administering the injection should have been given some (and arguably significant) weight by his Honour. The applicant submitted that this evidence gave rise to a reasonable possibility that Mr Kelly had been injected by a woman and that this should have created a real doubt about the applicant being responsible for injecting him. The applicant's submission is that Mr Kelly's recorded responses "were responsive to the nature of what he was asked".
Mr Kelly had been admitted to hospital on 19 October 2013 and remained there until he died on 29 October. His granddaughter, who visited him regularly from the time he entered the hospital, found that he was completely unresponsive on most days. Even when he regained some form of consciousness, she described him having difficulties responding to conversation. His granddaughter's evidence was summarised by his Honour at [83] of the judgment.
One of the police officers, who was present taking a note of the conversation, Detective Stephen, gave evidence at trial. He described Mr Kelly as not lucid because his eyes were rolling from side to side. He was not looking at the officers, he was mumbling, he began crying, he could close his eyes and appear to be asleep and then appear to come back again. Overall, Detective Stephen's description of Mr Kelly was that he did not appear to understand where he was or what was happening (T.778-782, AB 2093-2097).
The full conversation is set out at [118] hereof. As defence counsel at trial acknowledged some of the answers given by Mr Kelly were clearly not responsive. An obvious example of this is Mr Kelly's response of "radio" to the question of whether the person had black or white skin. Similarly, his answer "I think about tid [sic] morning" to the question "Do you know the person who injected it?". Further, Mr Kelly's response to the initial question of who gave him the injection was "He pulled it out". This is inconsistent with his later description of it being a woman who was responsible for the injection.
It was clearly open to his Honour in the light of all of the evidence to conclude that the "conversation" did not give rise to a reasonable possibility that the person who injected Mr Kelly was not a doctor or a nurse but was a woman.

(iv) Factors said to distinguish the applicant Alleged tendency type reasoning in relation to accused administering insulin to residents

The trial judge had regard to evidence that the applicant continued to give insulin injections to residents, even after the policy change in the facility in early 2013 which prohibited team leaders from giving such injections and required such injections to be administered by a registered nurse. Contrary to the applicant's submissions, this evidence was not relied upon by the Crown as tendency evidence.
His Honour included this fact in his judgment as the basis for one of the pieces of circumstantial evidence that there was no doubt that "the accused had training, skill and experience in the administration of insulin" (judgment [240]). Such an approach did not involve his Honour engaging in tendency type reasoning to the effect that the applicant had a tendency to inject residents when he was not authorised to do so. Rather, it was recorded as a factual finding.
There was also evidence that the applicant had given residents in the Mountview ward, insulin injections that were medically required on the morning of 17 October 2013 (one resident) and the morning of 18 October 2013 (two residents). The injection of Ms Fowler occurred on 18 October 2013 while the injections of Ms Manuel and Mr Kelly happened on 19 October 2013.
His Honour found that the applicant "continued to demonstrate his preparedness to inject residents with insulin right up to the time of the injections given to the three victims". His Honour went on to specify the relevance of this circumstance and it did not involve invoking a tendency purpose. At [241] of his judgment, his Honour stated "In part, this and the previously mentioned circumstance [i.e. the applicant being skilled and experienced in the injection of insulin] mean that the [applicant] taking insulin and syringes from the treatment room on 18 or 19 October 2013 might be perceived as unremarkable".
As a piece of circumstantial evidence, it was clearly open to his Honour to make that finding and have regard to it.
The applicant submitted that his Honour erred by regarding the fact of the applicant continuing to give insulin injections as something which distinguished him from other team leaders. That was because there was evidence that other team leaders had also continued to inject insulin after the policy change.
The applicant's proposition might have had some force if it were looked at in isolation. His Honour's conclusion, however, was not so expressed. The list of 20 staff members, including the applicant, who were present at the SummitCare facility within the timeframes for the injection of each of the victims (at judgment [222]) only incorporates one other team leader from another ward (Amanda Nanstou). There was no suggestion in the evidence that she administered insulin injections after the change in policy. Other than two registered nurses, the other staff on that list were all assistants in nursing, kitchen staff and laundry staff who had never been responsible in their roles for carrying out insulin injections or distributing oral medication.
If follows that there was no error in his Honour taking that evidence into account.

Items found at the home of the applicant

When a search warrant was executed at the applicant's home on 5 December 2013 a quantity of needles, syringes as well as literature concerning insulin were discovered by police. The literature included times that particular types of insulin took to achieve a hypoglycaemic effect (Exhibit AC). The documents contained references to overdosing causing hypoglycaemia. In addition, there was a recitation that severe hypoglycaemia may lead to unconsciousness and might result in temporary or permanent impairment of brain function or even death. The documentation referred to a number of types of insulin that were on hand in the treatment room of the relevant ward as at the date of the injecting of the three victims. Evidence was given by the officer in charge that this documentation was found in a black bag in the applicant's room (T.795, AB 2110). The officer in charge also gave evidence that a number of syringes were located. Photographs of the syringes were tendered (Exhibit AD).
During the course of his closing address, defence counsel at trial submitted that the insulin documentation was a readily accessible document and that you would expect that a person in the position of the applicant who was studying "prudently would have that type of material". After this submission, his Honour asked counsel (T.835): "HIS HONOUR: There's no suggestion of him having any material other than the extract relating to insulin. DEFENCE COUNSEL: No, that's right, and what there isn't, there isn't, for example and I think this arises in some cases where people are assisted to finalise them lives suicide where there's - there's no material such as downloaded from the Net to say "well if you want to finish - terminate somebody's life, this is what you do. This is purely a document that gives descriptors of different types of insulins and it doesn't - the only detail that it gives is that it says that it is injected subcutaneously. So there is no other material and there's certainly nothing downloaded from the Net, for example, or any booklet or guides of how to dispatch a person with the injection of insulin." (T.835, AB 2150).
In his judgment, his Honour observed that Mr Watson had argued that the insulin literature was readily obtainable and that it was at the accused's home in connection with his studies in nursing. His Honour stated "In relation to that first, there is no evidence of it and secondly, Mr Watson accepted that there was no other literature found that might be relevant to such studies." (judgment [243]).
Given counsel's initial answer of "No that's right", to his Honour's question, it is understandable why his Honour viewed this as an acceptance by counsel. Even if his Honour and defence counsel were at cross purposes, the simple fact is that there was only that one reference in the interview to "medical journals". There was no elaboration on what they were and even whether they were strictly connected with the applicant's studies.
In any event, error has not been demonstrated in his Honour's approach to this question as set out in the judgment.

Opportunity - access to insulin and relationship with the victim

The applicant's submissions in respect of this factor overlap with some of the matters that have already been addressed. His Honour's finding that the applicant had the "opportunity" to give the injections was based on the timeframes that his Honour had identified according to the expert evidence. The applicant again challenged this part of his Honour's findings by reference to the possibility that Ms Fowler was injected very early in the morning before the applicant's arrival at SummitCare. As already indicated in relation to Exhibit D, it was open to his Honour to be satisfied that the relevant timeframe for Ms Fowler was from 8am to 4pm on Saturday, 19 October. Once that finding is accepted, there was no issue that the applicant was working in the Mountview ward within the timeframes for the injection of each victim. It was therefore open to his Honour to find that the applicant had ready access to each victim so as to administer an injection of insulin while acknowledging that he was not the only one who had such access (judgment [240]).
It was also not in issue that the applicant was involved in the administration of oral medications to all residents in the Mountview ward, including the three victims and also engaged in the unauthorised activity of giving insulin injections to some insulin dependent residents. This provided an adequate basis for his Honour's finding that the applicant had the opportunity, probably more so than anyone else, to administer the injections.
It is no answer to that finding by his Honour to submit (as the applicant does) that there were registered nurses on duty who would have had a greater opportunity to administer injections without coming under notice. That response does not take into account the list of 19 other staff members, who were present during the timeframes for the injection of each victim, only one of whom was a registered nurse (Stephen Zhao). Although Mr Zhao was on duty in the Mountview ward during the relevant timeframe, he gave unchallenged evidence that his responsibilities did not involve him at any time on 19 October 2013 having any reason to come in contact with either Mr Kelly or Ms Manuel. In contrast, the applicant's statements to the police acknowledged that he regularly administered medications to Ms Fowler, Ms Manuel and Mr Kelly including on 18 and 19 October (judgment [121]). His responsibilities clearly involved him coming in contact with each of the victims during the relevant timeframes on those two days.
This evidence was important and his Honour was entitled to rely upon the fact that the applicant had the opportunity to commit the offences. Put the other way, if the Crown had not been able to establish that the applicant had the opportunity to commit the three offences, then the Crown would have failed to make its case. Moreover, as has been often stated when considering a circumstantial case, all of the circumstances are to be considered and weighed (The Queen v Baden-Clay at [47]).
There was no issue that the injections were administered and there was no suggestion that any of the three victims remonstrated or protested about the injection. It was therefore open to his Honour to find that in those circumstances, particularly in the case of the two more ably functioning victims, that this tended to support the notion that the injections were given by someone with whom the victims were familiar. Accordingly, the nature of the relationship which the applicant had with the victims, including as their primary supplier of medication, was a piece of circumstantial evidence that it was open to his Honour to take into account.
The view of the trial judge in relation to these factors
The applicant challenged his Honour's finding in relation to the 19 other staff members who were present at all the relevant times that "there was nothing to suggest anyone of them in particular might have been responsible" (judgment, [252]).
As part of the analysis leading to the acceptance of Exhibit D, it was open to his Honour to conclude on the basis of all the evidence that there were only 20 staff members who could have administered all three injections and therefore that the perpetrator must be one of those persons.
Given that the Crown case was circumstantial, it was necessary for any other reasonable hypothesis to be excluded (The Queen v Baden-Clay at [46], [50]. It was therefore necessary for his Honour to assess whether or not the alternative scenario that the perpetrator was somebody else from within that group could be excluded. His Honour took that into account when he assessed the evidence of each of those persons (except for the one who was deceased).
The applicant submitted that his Honour had reversed the onus of proof by treating the absence of evidence in respect of other potential perpetrators to make up for the shortfall in the evidence which could be used to inculpate the applicant. This submission fails to have regard to what his Honour actually did. His Honour was very careful about not reversing the onus of proof. He expressly stated in relation to that group of employees that he had seen give evidence: "224 None of these members of staff were cross-examined in a way to raise doubt about their denials. I am not suggesting that was required on behalf of the accused; the onus of excluding these people is on the Crown. …" What his Honour did take into account was that there was no suggestion that the police investigation was deficient in scrutinising the possible culpability of any particular individual. His Honour relied upon the ability he had to observe each of those persons give evidence. He described them as impressing him as "candid and honest". While he acknowledged that a guilty person might obviously still be capable of portraying an image of innocence, he was not left with any feeling of unease that any of those persons might be a suspect (judgment [226]).
By following such an approach, his Honour was not utilising the absence of evidence to make up for a "shortfall" in the evidence which could be used by the Crown to inculpate the applicant.
The applicant submitted that the only circumstance which was "peculiar" to the applicant, as distinct from the other 20 employees, was the items recovered at his home as a result of the search warrant. This submission fails to recognise that a large proportion of the 20 staff members did not work in the Mountview ward. Some of them had no particular knowledge of or contact with the three victims. Many of them had no skill or experience in injecting insulin and only team leaders and registered nurses (of whom there were only three in that group) had keys to access treatment rooms where insulin was stored. It also ignores the evidence about the text messages sent by the applicant

The text messages

The applicant challenged his Honour's finding that the text messages supported the inference that the applicant had foreknowledge of the impending or prospective demise of Ms Manuel and Mr Kelly because he knew that they had both received injections of insulin. The relevant text message exchanges are set out at paragraph [149] and [161]-[164] of his Honour's judgment. The applicant's text at 11.27am on 19 October clearly predicted that the next resident to die was going to be "Bell or Kelly". There was no issue that "Bell" was a reference to another resident, Ms Ethel Cliff, and Ryan was a reference to Mr Ryan Kelly.
As at 11.27am, Mr Kelly had not demonstrated any symptoms of hypoglycaemia. Before being confronted with the text messages, the applicant had made a police statement asserting that on the relevant morning Mr Ryan seemed his normal self and generally fine. Aside from Mr Ryan's medical conditions he was otherwise in reasonable health (judgment [131]). In a further statement to police, the applicant asserted that "there was nothing that [he] saw that morning that gave him any indication that Ryan was unwell" (judgment [131]).
As his Honour found, even on his account, the applicant had no reason to think when he sent that text message that Mr Kelly would die. The applicant's description of Mr Kelly's well-being that morning was also consistent with what was described by others. It would reasonably be expected that if the applicant was going to predict someone's death, he would nominate people who were most ill amongst the residents in the ward rather than Mr Kelly who despite his medical conditions was relatively stable.
At 1.23pm on the same day, 19 October 2013, the applicant sent a text message to his colleague advising that Ms Fowler had died and that "Audrey" would be "no 3".
There was no issue that Audrey was a reference to Ms Manuel. A similar text was also sent to another colleague at 8pm nominating "Audrey" as the next that was going to die.
The applicant described in statements to police, again prior to being confronted with the text messages, that he had seen Ms Fowler "take a turn" that morning at about 11.30am. He described his initial assessment to be that Ms Manuel had suffered "a TIA which is a type of small stroke". He said that at that time it was his belief that despite being taken to hospital there was nothing that suggested to him that Ms Manuel would die. This was on the basis of his experience with other residents that had experienced TIAs.
His Honour therefore found that at the time of that text (1.23pm) Ms Manuel had been taken to hospital with a condition that the applicant claimed to police that he did not regard as life-threatening. Yet in private communications with two colleagues he was suggesting that she was the next resident in the ward who was going to die.
His Honour observed that it might be expected that if the applicant was going to predict someone's death, he would nominate people who were the most ill amongst the residents of Mountview ward, not Ms Manuel whom he thought to have taken a turn which was not anticipated by him to be catastrophic.
In the light of the above, the findings which his Honour made in respect of the text messages were well open to him and the applicant's contention that such findings were not reasonably open should be dismissed.

(v) Shortcomings in the Crown Case: lack of motive and inadequacy of investigation

As was acknowledged in the applicant's submissions, the Crown was not required to prove motive. All that can be said is that the presence of a motive would have made the Crown case stronger. Its absence was not fatal to the Crown case, if the Crown case was otherwise made out.
In relation to the asserted inadequacy of the police investigation, the circumstances in which it was carried out need to be kept in mind. The extent of and nature of the police investigation was covered in detail by the Crown in the evidence in chief of the Officer in Charge. He was not cross examined on those issues. The extent of the delay between the administration of the injections and the police being notified should also not be disregarded. The fact of the crime scene being substantially contaminated by that time was also significant.
Contrary to the applicant's submission, the impression clearly gained by his Honour was that of a long, drawn out and painstakingly detailed investigation. His Honour observed that the time and effort put into the analysis of 37 hours of footage from each of five CCTV cameras in order to identify, or attempt to identify, the 317 people who attended SummitCare, some once and some on multiple occasions, must have been considerable. It was also necessary to take statements from those persons (judgment [235]).

Conclusion

His Honour's conclusion that one person was responsible for all three offences and his reliance upon the eight hour period specified by the experts to narrow the pool of potential perpetrators who were present during all three relevant timeframes for the injections did not involve impermissible backwards or reverse reasoning. At all times, his Honour was conscious of the onus carried by the Crown. As part of that onus, it was necessary for the Crown to eliminate alternate hypotheses and persons as possible perpetrators. That did not involve reverse reasoning but rather was a legitimate method of eliminating possible alternatives to the applicant as perpetrator of these offences. After giving due consideration to the circumstantial case presented by the Crown as a whole, and not by a piecemeal approach to each particular circumstance, it was well open to his Honour to be satisfied beyond reasonable doubt as to the guilt of the applicant. Certainly, as a result of my review of the whole of the evidence at the trial, both as to its sufficiency and quality, I have not been left with a reasonable doubt as to the correctness of the verdict.
The orders which I propose are: 1. Leave to appeal against conviction is granted. 2. The appeal is dismissed.
HARRISON J: I have had the benefit of reading the judgment in draft of Hoeben CJ at CL as well as the additional remarks of Schmidt J. I agree with their Honours' respective reasoning and the conclusions that they have reached.
I have also undertaken the not inconsiderable task of reviewing the transcript of the proceedings below and the evidence that was before his Honour. I am satisfied that it was well open to his Honour to be satisfied beyond reasonable doubt that the applicant was guilty of the offences with which he was charged. In my view, the collocation of events that went to make up the Crown's circumstantial case was overwhelming.
More particularly, with respect to ground 2 of the appeal, I am completely satisfied that the applicant has not established that the verdicts were unreasonable or that they cannot be supported by the evidence.
In precisely the same way that juries are urged and required to apply their collective wisdom and common sense to their deliberations, so too is a judge sitting alone in a criminal trial expected to do the same. The tragic and extraordinary events that gave rise to these proceedings included the unprescribed administration of insulin in either lethal or potentially lethal doses to three elderly patients in the same ward within a time frame of 48 hours. Applying one's common sense, and without drawing any inferences or conclusions about who was or might have been responsible, the likelihood of these events occurring accidentally or by chance must be very small indeed. In the same way, the likelihood that more than one person, rather than a single individual acting alone, could have been responsible for the relevant acts must be equally small if not almost non-existent. In making that observation, I do not intend to suggest that juries and judges can draw conclusions or make decisions without reference to the particular evidence in the trial. However, by the same token, consideration of that evidence ought not to take place in a common sense vacuum or in a state of suspended reality.
It is therefore unsurprising that trial counsel accepted that the acts causing death or injury in this case were performed or committed by a single perpetrator. It is equally unsurprising that his Honour felt able to come to the same view to the criminal standard of proof based upon his detailed and critical analysis of the evidence and independently of any concession about it made at trial.
SCHMIDT J: I agree with the conclusions which Hoeben CJ at CL has reached in relation to each ground of appeal, having myself considered both the circumstantial case which the Crown advanced, as well as the case advanced for Mr Davis, on the evidence which I consider was correctly admitted by R A Hulme J.
I also have been left with no reasonable doubt as to the correctness of the verdict which R A Hulme J reached and thus also agree with the orders which his Honour proposes.
For myself I wish to add the following about the arguments advanced in relation to Ground 1, which concerns ss 98 and 101 of the Evidence Act, the coincidence evidence and coincidence reasoning.
The Crown's case against Mr Davis was circumstantial. It had to be approached in the way discussed in Shepherd at 579. The ultimate issue was whether the Crown had proven beyond reasonable doubt that it was Mr Davis who had injected each of the three victims with insulin, thereby committing each of the offences charged.
The Crown had given a coincidence notice under s 98(1) of the Evidence Act which provides: "98 The coincidence rule (1) Evidence that 2 or more events occurred is not admissible to prove that a person did a particular act or had a particular state of mind on the basis that, having regard to any similarities in the events or the circumstances in which they occurred, or any similarities in both the events and the circumstances in which they occurred, it is improbable that the events occurred coincidentally unless: (a) the party seeking to adduce the evidence gave reasonable notice in writing to each other party of the party's intention to adduce the evidence, and (b) the court thinks that the evidence will, either by itself or having regard to other evidence adduced or to be adduced by the party seeking to adduce the evidence, have significant probative value. Note. One of the events referred to in subsection (1) may be an event the occurrence of which is a fact in issue in the proceeding."
Section 101 also applied to the evidence the subject of the notice, s 101(2) providing that such evidence "cannot be used against the defendant unless the probative value of the evidence substantially outweighs any prejudicial effect it may have on the defendant". "Probative value" is defined in the Dictionary to mean "the extent to which the evidence could rationally affect the assessment of the probability of the existence of a fact in issue."
The notice provided: "Notice is given that the Prosecution presently intends to adduce "coincidence" evidence pursuant to the coincidence rule in sub-section 98(1) of the Evidence Act 1995, ie evidence that two or more events occurred to prove that, because of the improbability of the events occurring coincidentally, a person did a particular act or had a particular state of mind. The "person" referred to in the preceding paragraph is Garry Steven DAVIS. The two or more events which are the subject of the proposed evidence are: the death of Gwen Fowler, the death of Ryan Kelly and the hospitalisation due to insulin overdose of Audrey Manuel. The purpose of the evidence is to rely upon similarities in each of the three events themselves and the circumstances in which each of the three events occurred to prove that the accused did the acts as alleged by the Crown, that is injected each of the three persons named in the indictment with insulin and with the relevant intent by reasoning that it was improbable that the events and the circumstances surrounding the events occurred coincidentally and are the acts of different persons. The substance of the evidence of the occurrence of the events is contained within the following documents which previously have been served upon you. DOCUMENT DATE PARAGRAPH

… So far as known to the Prosecution, particulars of the date, time, place and circumstances at or in which each of those related events referred to in paragraph 2 above occurred are contained within the documents referred to in paragraph 4 above. So far as known to the Prosecution, the names of each person who saw, heard or otherwise perceived each of those related events referred to in paragraph 2 above are contained within the documents referred to in paragraph 4 above."

The Crown contended that the coincidence evidence had significant probative value, because it established similarities in how the victims came to die or be seriously injured within a two day period, as the result of the overdose of insulin with which they were injected. That evidence, the Crown contended, showed that it was improbable that the events it relied on had occurred coincidentally, as the result of the acts of different people.
The Crown also relied on other circumstantial evidence to establish that the person who had injected each of the victims was Mr Davis.
Contrary to what was submitted for Mr Davis on appeal, the only fact in issue at trial was not whether it was he who had administered the injections: appeal submissions at [62].
Mr Davis then contended that the Crown's case was incapable of establishing his guilt, beyond reasonable doubt. He challenged the strength of the similarities on which the Crown relied: Judgment at [195]. He also challenged the legitimacy of the theory that there was only one perpetrator, contending that there were a range of possibilities as to the perpetrator being someone other than him, which had not been excluded: at [197]. Further, that the police investigation had been inadequate and had also not excluded the possibility that an unidentified person had been the perpetrator: at [200]; [233].
At trial evidence was thus led without objection about the circumstances in which two of the victims came to die and the third to be seriously injured, there being no issue that each of them had been injected with a toxic dose of insulin, which had not been prescribed. Evidence that Mr Davis was on duty on the day that each of the victims was so injected, was also admitted without objection, that being relevant to the question of whether he had the opportunity to commit all three offences. The expert evidence was directed to the timeframe within which the injections were administered, on the Crown case while Mr Davis was on duty.
All of this evidence was relevant under s 55 of the Evidence Act and thus admissible under s 56, for a non-coincidence reasoning purpose, it being evidence which "if it were accepted, could rationally affect (directly or indirectly) the assessment of the probability of the existence of a fact in issue in the proceeding", s 55. The facts in issue in a criminal proceeding being "those which establish the elements of the offence": Hughes v R at [16].
The case advanced for Mr Davis on appeal was that at trial the Crown had proceeded on an erroneous construction of s 98, "by failing to identify the particular issue which was sought to be established by use of the coincidence reasoning" (appeal submissions at [60]), contrary to the approach discussed by Simpson J in Duckworth. That argument cannot be accepted given the terms of the coincidence notice, the case which the Crown advanced at trial and what Simpson J said in Duckworth at [25]: "At its heart, s 98 is a provision concerning the drawing of inferences. The purpose sought to be achieved by the tender of coincidence evidence is to provide the foundation upon which the tribunal of fact could draw an inference. The inference is that a person did a particular act or had a particular state of mind. The process of reasoning from which that inference would be drawn is: two or more events occurred; and there were similarities in those events; or there were similarities in the circumstances in which those events occurred; or there were similarities in both the events and the circumstances in which they occurred; and having regard to those similarities, it is improbable that the two events occurred coincidentally; therefore the person in question did a particular act or had a particular state of mind."
At trial there was no issue that the three events on which the Crown relied had occurred. There were unarguably similarities between those events, which in the case of two of the victims had resulted in their deaths and in the other serious injury, they having each been injected with a high dose of insulin which had not been prescribed, while a resident of the same aged care facility.
There can thus be no question that the Crown was entitled to rely on the similarities between these events, which had occurred within a two day period, to advance its case that it was improbable that they had occurred coincidentally and that it was the one person who had administered all three injections, with an intent to kill or inflict grievous bodily harm.
This was in dispute and its resolution depended on an assessment of the combined effect of all of the relevant similarities: R v MR at [9]-[10]; [78]-[79].
To establish Mr Davis' guilt the Crown was also was entitled to rely, as it did, on the other circumstantial evidence received at trial, to prove that it was Mr Davis who was the offender in each case, that evidence putting into context as it did, the similarities between the coincidence evidence on which the Crown relied.
The other circumstantial evidence relied on included that the three victims each knew Mr Davis, because they were living in the ward where he worked; that he was responsible for the administration of their medication, when he was on duty there; that Mr Davis was on duty on the ward on the day that each victim was injected with an overdose of insulin; what the police investigation had revealed, including the information obtained as to who else had entered the facility and had access to the insulin kept there; the documents which dealt with insulin found at Mr Davis' home; the text messages which he sent co-workers about who would be next to die; and what the expert evidence established as to the timeframe within which the injections were administered to each victim.
As R A Hulme J found, much of that evidence was admissible for a non co-incidence reasoning purpose: Judgment at [194]. That explains why there was no objection to its receipt. His Honour also there noted that the case pressed for Mr Davis as to the coincidence evidence went to the strength of the Crown's case as to the similarities it relied on, rather than to its admissibility and perhaps to whether the evidence had significant probative value. His Honour found that it did: at [196].
R A Hulme J was correct. The conclusion that the probative value of the coincidence evidence significantly outweighed any prejudicial effect that it may have had on Mr Davis was inescapable, influential as it undoubtedly was in the fact finding exercise his Honour was required to undertake: IMM v The Queen at [148].
Given that this was a judge alone trial, correctly, there was then no suggestion made that there would be any prejudicial effect, flowing from the receipt of the coincidence evidence. That question had to be resolved by the coincidence evidence being considered together with all of the other evidence which the Crown had adduced: DSJ v The Queen at [9].
It follows that the case advanced for Mr Davis on appeal, that before R A Hulme J could deploy any coincidence reasoning, the Crown had to establish beyond reasonable doubt that he was guilty of at least one count, cannot be accepted. That is not a requirement imposed by s 98 of the Evidence Act and cannot be imported into this statutory scheme, in the way for which Mr Davis contended.
Indeed, Duckworth, the case on which Mr Davis relied on appeal, was itself largely a circumstantial case: at [7]. What was there observed at [28] was that the condition which has to be met when coincidence evidence is sought to be adduced is that "that the evidence, either by itself or having regard to other evidence adduced or to be adduced by the party tendering the evidence, will have significant probative value". What is not required is that such other evidence be either direct evidence of the alleged offender's involvement in an offence, or where more than one offence is alleged to have been committed, that the evidence relied on establishes guilt on at least one of the counts.
Further, contrary to the circumstances which arose for consideration in Gilham, on which reliance was also placed on appeal, in Mr Davis' case R A Hulme J did consider whether the coincidence evidence relied on had significant probative value, as the Crown contended. There was no error in his Honour's conclusion that it did.
Because there was no direct evidence linking Mr Davis to any of the three offences, the case which the Crown advanced had to be a circumstantial one. As was submitted for Mr Davis, determination of whether the Crown had met the onus falling upon it to prove his guilt beyond reasonable doubt by the evidence it had adduced thus required his Honour to consider, as he did, other possible explanations for the insulin administered to the three victims, which might have been consistent with his innocence. That exercise had to be undertaken, in order to resolve whether the inferences which the Crown sought to draw from the evidence, were available.
Contrary to the case Mr Davis advanced, however, his Honour's resolution of these issues, did not involve pursuit of a process of "reverse" or "backward" reasoning, he being found guilty only because he was "the last man standing", after other suspects were excluded. Rather, his Honour considered what was in issue, assessing the evidence on which the Crown relied and concluding for the reasons which he explained, that the coincidence evidence and that admitted for non-coincidence purposes as part of the circumstantial case which the Crown advanced, had proven beyond reasonable doubt that Mr Davis was guilty of all three offences.
There was no error in his Honour's approach.
I also agree with Hoeben CJ at CL's conclusions that his Honour did not fall into the various other errors for which Mr Davis contended and do not wish to add to the reasons which he has given for those conclusions.

DISCLAIMER - Every effort has been made to comply with suppression orders or statutory provisions prohibiting publication that may apply to this judgment or decision. The onus remains on any person using material in the judgment or decision to ensure that the intended use of that material does not breach any such order or provision. Further enquiries may be directed to the Registry of the Court or Tribunal in which it was generated. Decision last updated: 30 November 2018

At a glance

Court

Decision date

Before

Catchwords

Source

Catchwords

Judgment (36 paragraphs)

Parties

Legislation Cited (4)

Cases Cited (30)

Cited By (2)