Coroner's Finding: COSTI John Steve

CORONERS ACT, 2003 SOUTH AUSTRALIA FINDING OF INQUEST An Inquest taken on behalf of our Sovereign Lady the Queen at Adelaide in the State of South Australia, on the 6th and 7th days of April 2017, the 8th day of May 2017 and the 17th day of January 2018, by the Coroner’s Court of the said State, constituted of Anthony Ernest Schapel, Deputy State Coroner, into the death of John Steve Costi.

The said Court finds that John Steve Costi aged 39 years, late of the Adelaide Remand Centre, 208 Currie Street, Adelaide, South Australia died at the Adelaide Remand Centre, Adelaide, South Australia on the 27th day of November 2013 as a result of neck compression (due to hanging) and plastic bag asphyxia. The said Court finds that the circumstances of his death were as follows:

1. Introduction and reason for Inquest 1.1. John Steve Costi was 39 years of age when he died on Wednesday 27 November 2013.

At the time of his death he was an inmate at the Adelaide Remand Centre (the ARC).

On the morning of that day he was found dead in the prison cell that he occupied alone.

The time was approximately 6:07am. A ligature made from strips of bedding was around Mr Costi’s neck. This ligature had been secured to an unusual hanging point that in the interests of prisoner cell safety I do not intend to describe here. There was also a pillow case and two plastic bin liners over his head. Further material was secured to both wrists which were tied behind his back. This scenario gave rise to a suggestion that some other person may have played a role in Mr Costi’s death which, as revealed at a post-mortem examination, had been caused by the ligature around his neck and

possibly contributed to by the plastic bags over his head. For reasons that will become apparent, including the fact that it has been established beyond doubt that no other person could have been in Mr Costi’s cell from the time it was locked down with a demonstrably alive Mr Costi in it until the time the cell was unlocked and his body was discovered, I have dismissed any suggestion that Mr Costi’s death was caused or assisted by some other person. I have found that Mr Costi’s death was due solely to his own actions. The method of his death is consistent with nothing other than an intent to take his own life. I so find. The only document that might be construed as one written with his imminent death in mind was a note found within documentation that Mr Costi had placed in envelopes addressed to his female partner but which were still in the possession of the ARC prior to it being dispatched. The note, the content of which smacks of finality, urged Mr Costi’s partner to look after his dog. The other documentation consisted of letters that his partner had written to him while in custody.

All of this material had been placed in the prisoner mail the day before Mr Costi’s death.

Mr Costi’s act of attempting to return this correspondence is consistent with the recent formation of an intent to finalise his affairs and therefore with his suicide having been thought out in advance. So is his apparent disposal of legal papers and correspondence that he had accumulated but which were no longer in his cell.

1.2. The evidence suggests, and I so find, that Mr Costi had last been seen alive at approximately 4am on the morning in question. He was seen during a routine observation patrol conducted by correctional officers at that time. On that occasion he was alone in his cell and apparently asleep on his bed. As would be routine at that time of the night his cell door was locked as it had been since the day before. It was locked in a manner that I will describe in due course.

1.3. Mr Costi’s death was a death in custody and for that reason an Inquest into the cause and circumstances of his death was mandatory. These are the findings of that Inquest.

2. Background 2.1. It is necessary to describe the circumstances in which Mr Costi came to be at the ARC as an inmate.

2.2. Mr Costi was no stranger to the criminal justice system, to correctional custody or for that matter to the ARC. From 2002 Mr Costi developed a criminal history that included assault occasioning actual bodily harm, the imposition of domestic violence restraining orders, firearm and drug offences and failure to comply with bail conditions. The

statement of the South Australia Police (SAPOL) investigating officer, Detective Brevet Sergeant Drew Bynoe of the Police Corrections section1, records that in 2009 Mr Costi spent 269 days in custody at the ARC as well as 5 days at the Yatala Labour Prison and that in 2012 and 2013 Mr Costi spent 351 uninterrupted days in custody at the ARC. This last period of custody was that immediately preceding Mr Costi’s death.

This period represented a remand in custody following Mr Costi’s arrest on 10 December 2012 for alleged offences of robbery, attempted robbery and rape committed between 26 July 2012 and 10 December 2012. The underlying allegations included the robberies of female escorts involving use of a handgun or replica handgun and the rape of an escort. At the time of his death Mr Costi was awaiting trial in relation to these matters.

2.3. At the time Mr Costi was charged with these offences a SAPOL prisoner screening form created and subsequently maintained within the SAPOL PIMS warning system indicated that Mr Costi was at risk of self-harm. A related SAPOL ancillary report detailed an incident on 6 January 2009 in which Mr Costi, said to have been grossly affected by alcohol and drugs, held police at bay while armed with a firearm. Mr Costi is recorded as having emerged from a premises and then to have cut himself with a Stanley knife thereby severely injuring himself. Mr Costi allegedly made threats to kill police, a former girlfriend and himself. It was following this incident that Mr Costi was held in custody in the ARC during the period in 2009 referred to above. Due to his perceived risk of self-harm, for a time during this period of remand Mr Costi was under a High Risk Assessment Team (HRAT) regime and was required to share a cell. He was removed from the HRAT list on 19 February 2009 and he was ultimately released in October 2009. I am not certain of the outcome of the proceedings underlying this remand in custody. It is not material to the issues under discussion here.

2.4. The SAPOL paperwork regarding the PIMS warnings to which I have referred accompanied Mr Costi when, after his arrest in December 2012, he was transferred from SAPOL custody to Department for Correctional Services (DCS) custody and it remained within his DCS management file. There was no record of any behavioural matter while in SAPOL custody in December 2012 that would signify any issue in respect of Mr Costi’s welfare. There is no suggestion other than that Mr Costi was calm and compliant during SAPOL custody prior to his transfer.

1 Exhibit C35

2.5. Mr Costi was maintained in SAPOL custody until the day following his arrest on which day he was transferred to DCS custody and ultimately accommodated in the ARC. The SAPOL Custody Transfer Form was endorsed with a warning of self-harm or suicide likelihood in an affirmative manner. The documentation was endorsed to the effect that Mr Costi had once slashed his wrists and his throat in 2009. This documentation accompanied Mr Costi to the ARC where he was admitted on 12 December 2012.

2.6. It is of relevance to the issues ventilated in this Inquest that from 12 December 2012 until the day of Mr Costi’s death in November of 2013 Mr Costi was not serving a sentence of imprisonment as a convicted prisoner. Rather, he was remanded in custody pending the resolution of the charges that had been laid against him. It is evident that Mr Costi intended to plead not guilty to the various charges which meant that he would be required to undergo a trial in the District Court. Notwithstanding the fact that a remand prisoner undergoes a period of custody not as a sentenced prisoner but as an accused person with the presumption of innocence, a remand in custody at the ARC involves what any reasonable person would regard as a punitive regime. For instance, evidence was given that on a ‘standard day’ a prisoner is locked in his cell for about 16.5 hours out of every 24. Of those 16.5 hours they are locked continuously for 15 hours and 40 minutes between 4:40pm and 8:20am2. It goes without saying that the imposition of such a regime attracts a very high duty of care towards prisoners in terms of scrutiny and safe cell design. There is nothing novel about such a suggestion.

2.7. Mr Costi’s court proceedings are also of relevance. He first appeared in the Adelaide Magistrates Court on 11 December 2012 on which date he was remanded in custody.

From time to time thereafter he was required to appear either by way of audio-visual link from the ARC or by way of personal attendance firstly at the Magistrates Court and then at the District Court following committal for trial. In respect of his court proceedings Mr Costi was represented by a solicitor, Ms Yasmin McMahon, and a barrister, Mr Anthony Allen. Mr Allen appears to have come into the matter in October

2013. Mr Allen was the counsel engaged to represent Mr Costi at his trial. On several occasions Ms McMahon attended at the ARC taking instructions from her client.

Ultimately a trial date of 9 December 2013 was listed in the District Court. The allocated prosecutor from the Office of the Director of Public Prosecutions (the ODPP) 2 Transcript, page 153

was Mr Will Ellis. Mr Ellis provided a statement to police3. His estimate had been that the trial would take approximately 10 sitting days. It was apparent that the earliest trial date was 9 December 2013 and that the trial was to be before a judge and a jury. The charges were ultimately trimmed, but it is clear from court transcripts that the trial would involve all counts being tried simultaneously and would require no less than five individual complainants to give oral evidence. As Mr Costi pleaded not guilty to all of those charges there was a requirement for a trial. The calendar for the year 2013 reveals that 9 December was a Monday. A 10 day trial would have finished on Friday 20 December. Christmas Day that year was on Wednesday 25 December and, as is customary, the District Court would thereafter not sit again until early January. A jury trial not completed before Tuesday 24 December 2013 would have needed to be adjourned from that day until early January 2014, with the members of the jury remaining empanelled but being free to go about their normal daily affairs during the period of the adjournment. Such a scenario is considered to be undesirable and one to be avoided. I note that the first anniversary of Mr Costi’s arrest in respect of the offences of which he had been remanded in custody would have occurred shortly after the commencement of the trial, had it taken place as scheduled. As will be seen it did not take place as scheduled or at all.

2.8. In the statement of Mr Ellis of the ODPP he explains that continuing investigations undertaken during the course of Mr Costi’s remand period, and even after the trial date had been set at a directions hearing in July 2013, resulted in an expansion of the amount of evidence to be adduced at the trial. This gave rise to a concern that the trial would run longer than the original estimate and that this, for the reason I have identified in the preceding paragraph, would be complicated by the proximity of Christmas. In his witness statement4 Mr Anthony Allen, who was Mr Costi’s barrister, and who not been party to the original scheduling of the trial, indicates that shortly after he was briefed the ODPP sought to have the matter listed in court for a directions hearing, the purpose being to bring to the District Court’s attention a potential need to relist the trial. Mr Allen’s statement indicates that any application to postpone the trial was not of his or his client’s making. At the directions hearing which took place on 24 October 2014 unsuccessful attempts were made by counsel either to bring the trial commencement date forward a week or to postpone its commencement for only a short period to January 3 Exhibit C26 4 Exhibit C25

of 2014. However, the Court, citing the District Court’s reduction in the number of judges available to hear trials, indicated that a new trial commencement date earlier than late April or early May of 2014 could not be accommodated. Accordingly, the trial was removed from the trial list and a fresh trial date of 5 May 2014 was set. It was set down for 15 days. Thus it was that the inconvenience that may have been occasioned by the adjournment of a trial for just over a week between Christmas and New Year took precedence over the need to deliver timely justice to a prisoner who already would have been in custody for a year by the time of his original trial date.

Having regard to the nature and complexity of the prosecution case as described by the prosecutor at the directions hearing, it is hard to disagree with the observation made at that hearing that it was regrettable that the trial had been listed to commence on the date that it was.

2.9. Following the postponement of Mr Costi’s trial to May of 2014 he decided to make an application for release on bail. Mr Costi was refused bail, unsurprisingly so having regard to the nature and seriousness of his alleged offending and to his criminal history.

On 21 November 2013 the bail application came on before a judge of the District Court.

According to Mr Allen’s statement, Mr Costi’s family members attended court to provide support and if necessary to sign guarantees for Mr Costi’s release on bail. It was proposed that Mr Costi would reside with Ms Lauren Foster with whom Mr Costi had a relationship. The bail application was opposed by counsel for the DPP. In support of the application Mr Costi’s barrister, Mr Allen, made reference to the fact that by the time his client’s postponed trial would occur he would have been remanded in custody for almost exactly 18 months. In opposing bail counsel for the DPP pointed out to the Court that in the past Mr Costi had received suspended prison sentences, one of which would have been breached by the current alleged offending, the implication being that if Mr Costi was found guilty of the offences for which he was to be tried in May 2014 an immediate custodial sentence would be inevitable. In refusing bail the judge referred to the seriousness of the charges.

2.10. Thus it would remain the case that Mr Costi, but for his death, would have spent a significant period in custody even before his trial in May 2014. I would add here that if Mr Costi were to have been found guilty of the offences with which he was charged, and in that event were to have received what probably would have been an immediate custodial sentence, the period that he had spent in custody pending his trial and

sentencing would almost certainly have been taken into account by backdating the commencement of any immediate custodial sentence to the date on which Mr Costi was first taken into custody. On the other hand, if Mr Costi had been acquitted, and there were to be no other convictions to which the time that he had spent on remand in custody could be attributed, that time would have been dead time for which he could not have been given any credit. It is naturally beyond the scope of this inquiry to speculate upon the outcome of Mr Costi’s trial.

2.11. Mr Costi’s death occurred six days after the District Court’s refusal of his bail. Mr Allen’s statement indicates that immediately following the refusal of bail he spoke to Mr Costi in the cells. Mr Costi appeared disappointed with the result, although it must be said that his expectations of being granted bail could not reasonably have been very high. He understood that he would remain in custody until his trial date the following year. Mr Allen states: 'Notwithstanding his disappointment, there was nothing about the demeanour of Mr Costi that led me to believe that he was in such a state so as to cause harm to himself.' 5

2.12. In her witness statement6 Mr Costi’s solicitor, Ms Yasmin McMahon, points out that at the time the defendant’s trial date was vacated he had already been in custody for a considerable period of time. She emphasises that he was awaiting trial for charges that he was defending and in respect of which there was a presumption of innocence. She says: 'In my view it was inappropriate for concerns of Christmas holidays to interfere with the trial date which had been listed at the agreement of prosecution many months earlier and they had been fully aware of the number of witnesses who would be required to give evidence.' Coming as they do from an experienced criminal practitioner, these observations are not readily dismissed. In this regard it is to be acknowledged that there is a sharp distinction to be drawn between on the one hand avoiding setting down a trial where it might clash with and be interrupted by the Christmas and New Year period and on the other removing a scheduled trial from the trial list due to the possibility that it might overrun and clash with that period, especially where, as in this case, the accused will 5 Exhibit C25 6 Exhibit C27

have to remain in custody for several more months before his trial can be commenced.

Even allowing for the difficulty that might be occasioned in empanelling twelve jurors who are all comfortable with the prospect that their Christmas / New Year break might be affected by being part heard in a trial in which they are participating, one would have thought that a decision to remove a trial from the list for the reasons identified in this case is one not to be taken lightly.

2.13. Ms McMahon did not speak with Mr Costi from the time of his unsuccessful bail application on 21 November 2013 to the day of his death. However, she states as follows: 'Unfortunately I anticipate that John had been prepared for a trial date in December 2013 and could not cope with that trial date being vacated, the new trial date being set so far in advance, and his home detention bail application again being refused.'

2.14. There is a striking temporal association between the postponement of Mr Costi’s trial, his unsuccessful bail application made in the light of that postponement and Mr Costi taking his own life. In the absence of, say, a suicide note from Mr Costi to the effect that all of that affected his frame of mind, a conclusion that there is a demonstrable connection between the circumstances that surrounded his court matters and his decision to end his own life cannot be made with complete certainty.

2.15. In another section of these findings I will deal with the circumstances surrounding Mr Costi’s induction into the ARC and in respect of relevant matters during his period of remand.

3. Mr Costi’s post-mortem examination 3.1. I have referred in general terms to the ligature that had been created and of its position relative to Mr Costi’s neck. This was an unusual method of suicide particularly having regard to the nature and location of the hanging point. There was also the question of Mr Costi’s hands being tied behind his back, as well as the fact that his head was covered with two plastic bags and a pillowcase. Naturally when Mr Costi was found all ligatures and ties were cut and removed as were the items on his head. For that reason to my mind the precise nature of the ligatures and how they were tied and fastened cannot be reproduced with complete accuracy.

3.2. A forensic pathologist, Dr Cheryl Charlwood of Forensic Science South Australia (FSSA), attended the ARC and examined the deceased at the scene. Dr Charlwood attended the scene at approximately 11:10am on the day of Mr Costi’s death. She noticed that there were areas of intradermal and speckled bruising to the right periorbital and outer upper eyebrow region and forehead, some with linear components to the right forehead and around the right eye with areas of bruising and swelling to the eyebrow.

There was also minor superficial ‘peri-mortem’ abrasion to the right knee and shin. No major injuries were identified. There was no hand, lip, mouth or other obvious facial injury noted at the scene. Rigor mortis was established and the body showed some central remaining warmth.

3.3. Dr Charlwood’s report7 states that at the post-mortem examination which she later conducted at FSSA, she observed that there was a faint ligature mark identified to the anterior right side of the neck which was rising. There were asphyxial changes in the form of multiple petechiae to the skin of the face and eyelids with suffusion of the facial and head tissue. There was minor bruising to the right periorbital, eyebrow and temple regions, which likely relates to minor terminal blunt force trauma. Dr Charlwood expresses the view that this may be related to terminal hypoxic seizure activity and had likely occurred due to impact with surrounding objects in the cell. She also states that the overlying pillowcase and bag materials may have produced the linear impressions on his face. There was also minor abrasion to the front of the right knee and shin consistent with terminal injury. There were no broken bones within Mr Costi’s skeletal system.

3.4. Dr Charlwood states: 'There was no injury to overtly suggest a third party or other significant injury. The injuries are not considered to be a factor in the cause of death.' 8

3.5. As far as toxicological analysis is concerned, alcohol, amphetamines, cannabinoids, cocaine, morphine, benzodiazepines and other common drugs were not detected in Mr Costi’s blood.

3.6. Dr Charlwood further states that the features of the post mortem examination overall 7 Exhibit C2a 8 Exhibit C2a, page 2

were in keeping with the nature of the hanging point. She adds: 'The presence of the pillow case and plastic bags also being present over the head of the deceased are difficult to interpret as to their involvement in the death, but these may have further contributed to hypoxia and the mechanism of death.'

3.7. Taking all of the circumstances and findings into consideration Dr Charlwood opines that the cause of death was neck compression (due to hanging) and plastic bag asphyxia.

I accept Dr Charlwood’s evidence. I find that to have been the cause of Mr Costi’s death.

3.8. Produced with Mr Costi’s body at post-mortem was an amount of material that either been had been tied to or otherwise associated with Mr Costi’s person at the time of his death. This included a sheet which had been cut along its edge and material that had been tied to the hanging point. There was also a piece of sheet that was likely to have been around the neck of the deceased. There were also two thin white plastic bags and a blue pillowcase which had patchy apparent blood staining to its upper aspect. There was also a blanket. In addition there was what Dr Charlwood has described as ‘ligature material from around the right wrist’. Also from the right wrist was string like material.

From the left wrist was tightly applied blue bedding material and string-like material which had both been cut at the scene. As indicated earlier the precise manner in which all this material had been attached cannot now be accurately reproduced having regard to the urgency with which the material had been cut from his person and the understandable lack of attention to the precise manner in which the material had been attached. In the interests of prisoner cell safety I do not believe it would be appropriate to describe here the exact methodology that was utilised to carry out this hanging even if it was fully understood. I was satisfied that Mr Costi had applied all of the ties himself and had not been assisted by another person.

3.9. There was one other feature of Mr Costi’s autopsy that should be mentioned. Inside his stomach was a flat, circular electrical battery. It was approximately one centimetre in diameter. It was of a type commonly used to power the calculators that were made available for prisoners to use at the ARC. The presence of the battery in Mr Costi’s stomach did not contribute to his death. Why precisely Mr Costi would elect to ingest an object of this kind is less than clear, but the act would appear to be in keeping with a recent mindset of self-harm. How exactly Mr Costi was able to source this battery has not been established.

4. Mr Costi’s induction into the ARC and his period of remand 4.1. When Mr Costi was inducted into the ARC on 12 December 2012 a notice of concern (NOC) was raised due to the well documented 2009 episode of self-harm as described above. Seven-day observations were thereafter commenced. He was evaluated by the HRAT on 13 December 2012. During the course of that assessment there was discussion with Mr Costi about his previous act of self-harm. He stated that he had never considered self-harm or suicide since that event and in fact had seen it as a major wake-up call for him. His major concerns were said to be the fact that he missed his family and partner. He saw his future as being solely dependent on his court outcomes.

A number of protective factors were assessed to have existed in respect of Mr Costi at this time including a stable relationship with his partner who was said to be very supportive of him. He was also said to have a good insight into his emotions. He reported no thoughts of self-harm and did not show any signs to suggest that he was experiencing suicidal ideation. His estimated risk level was assessed as low.

4.2. Ms Sylvia Lowczak9 is a registered psychologist employed by DCS as a senior psychologist. Part of her duties include the assessment and treatment of prisoners who are identified as being at potential risk of self-harm while in custody. According to her statement Mr Costi was referred to her for assessment after the NOC was raised by custodial officers during his admission in the first instance to the SAPOL City Watch House on 11 December 2012. Ms Lowczak assessed Mr Costi at the ARC on 13 December 2012. She saw him again on 19 December 2012. In her statement Ms Lowczak indicates that she did not consider Mr Costi to be at significant risk of self-harm at the time of the assessment. She makes it plain that she knew of his previous self-harm. She recommended that he be removed from HRAT monitoring. The recommendation was supported by the HRAT Committee. Ms Lowczak had no further contact with Mr Costi during the remainder of his remand period.

4.3. A perusal of Mr Costi’s Offender Case Notes10 from December 2012 to the time of his death reveals nothing of particular concern. Ultimately he was placed in a cell on his own in Unit 4. It was recorded on 19 July 2013 that Mr Costi was very well settled and had no issues in Unit 4. It was also noted that he got along well with all in the Unit.

9 Exhibit C18 10 Exhibit C19d

4.4. I have examined the records relating to Mr Costi that were maintained by the South Australian Prison Health Service (SAPHS)11. He was seen at the City Watch House on 11 December and at the ARC on 12 December 2012. On both occasions he denied any thoughts of self-harm. There are other notations in December 2012 where nothing remarkable about Mr Costi’s frame of mind is noted. Thereafter there were further consultations with SAPHS regarding warts, gout and back pain. There is certainly no evidence of any complaint of a negative frame of mind on Mr Costi’s part.

4.5. I should also refer to the fact that statements taken from correctional services officers and comments made to investigating police by certain prisoners did not reveal any matter that ought to have been regarded with concern with Mr Costi’s safety in the ARC in mind. However, there was a suggestion that at one point during 2013 he started to isolate himself and that he would remain in his cell only to emerge for meals. This attitude of isolation would also be noticed by those close to him outside the institution.

However, on the whole nothing told to police by other prisoners suggests that there had been any cause for concern. According to the investigating police officer, Detective Brevet Sergeant Bynoe, prisoners in the Unit expressed shock and surprise at Mr Costi’s death. I should say here that the statements of prisoners concerning Mr Costi and of their own attitude towards his death perhaps need to be seen in the light of what is generally said to be a reluctance on the part of prisoners to report to DCS officers the behavioural or other issues connected with fellow prisoners. This phenomenon has been the subject of comment in coronial proceedings in the past. Be that as it may, in this case there is no evidence to suggest that any prisoner or DCS or SAPHS employee had detected anything in the behaviour of Mr Costi that suggested that he was at risk of committing suicide or that steps needed to be taken to ensure that Mr Costi’s risk of self-harm was mitigated by such measures as accommodating him in a double cell. The assessment of the investigating officer, Detective Bynoe, is that the behaviour exhibited by Mr Costi during his period of remand was that of a ‘model prisoner’12. It is not difficult to see how his behaviour could be so characterised.

4.6. I should say something about the statements of those persons who were close to Mr Costi during his period of remand, including his father, his partner and a cousin who 11 Exhibit C38 12 Exhibit C35c, page 14

was the last person to visit Mr Costi. Mr Costi’s father was Steve John Costi13. In his statement Mr Steve Costi indicates that although his son was never happy about being held at the ARC, he did cope with the day to day routine. Mr Steve Costi expresses surprise that his son had been in a cell alone for so long. Mr Steve Costi’s statement tends to confirm that his son had isolated himself by reading a lot. He last saw his son in the ARC on 19 November 2013. He would see him again on 21 November 2013 at the bail hearing which was the last time he saw his son. He describes the final ARC visit as a normal visit. He describes his son in these terms: 'He was the same, there was nothing different in his behaviour or attitude that would have made me worried.' Mr Steve Costi states that he knew that he could have drawn anything untoward about his son to the attention of the correctional services officers. He said: '… they would have acted. There are signs in the prison telling visitors to report things about the prisoners if they are worried.'14 Mr Steve Costi’s statement makes it reasonably plain that he did not entertain any concerns for his son’s safety at any time during the period of remand in custody. He states: 'Nothing ever gave me cause with John’s behaviour or what he said, that I had fears for his personal safety inside the prison.' 15 Mr Steve Costi confirms that at one point his son stopped making phone calls from the ARC. He attributed this perhaps to a desire on his son’s part to save money.

4.7. Mr Steve Costi speculates that the straw that broke the camel’s back was the postponement of the trial date and his son’s unsuccessful home detention bail application. Mr Steve Costi states: 'I think that the justice system has a lot to answer for in the way that people are remanded without facing a jury to say if they are innocent or guilty.' 16

4.8. Mr Steve Costi made an impassioned speech at the end of these proceedings. As well, he tendered a written submission. I indicate that I have taken everything that Mr Costi 13 Statement of Mr Steve Costi, Exhibit C20 14 Exhibit C20, page 5 15 Exhibit C20, page 12 16 Exhibit C20, page 13

has submitted into account. Mr Costi understandably lamented the disruption of his son’s court proceedings by the adjournment of his trial. He was also critical of the lack of screening of his son’s frame of mind during the course of his remand in custody. He also queried why his son had been accommodated singly and submitted that if he had been placed in a shared cell his death could have been prevented. He also criticised the predictability of prison officers’ routine checks thus enabling prisoners to carry out acts of self-harm without fear of detection, a matter that I will speak of later in these findings.

4.9. Mr John Costi’s partner was Ms Lauren Foster. Her statement17 goes into some detail about their relationship. She indicated that his arrest for alleged offences against women came as a real shock to her. She states that when he was remanded in custody he offered to allow her to cease the relationship, but she persevered with it notwithstanding. She confirms that while Mr Costi was on remand, to her knowledge he kept to himself and read books and watched television for the most part. She said: 'John always seemed fine to me when I spoke to him.' 18 She states that it did not appear that he was having a hard time of it. Ms Foster indicates that she had agreed to allow Mr Costi to reside at her premises should he be granted bail.

4.10. The last time Ms Foster saw Mr Costi was on 20 November 2013 at the ARC. This was the day before his bail application. Mr Costi was then in a light-hearted mood but held out only moderate hope of release. The following day Ms Foster attended court when bail was refused. This was the last time she saw Mr Costi.

4.11. In Ms Foster’s statement19 she surmises that Mr Costi’s suicide in custody was in some ways to be predicted because, as she puts it, ‘there was no way that he would spend a long time in jail’. However, she said that she did not share those feelings with any person. She does say that Mr Costi had never exhibited any behaviour that would have prompted her to tell the guards or anyone at DCS that she feared for his safety. She knew that she could have approached officers if she had entertained any such concern.

She believed that Mr Costi had been coping. She believed that the turning point for 17 Exhibit C21 18 Exhibit C21, page 8 19 Exhibit C21, page 13

Mr Costi was the refusal of bail. Ms Foster’s statement makes it reasonably plain that the note that had been written to her which urged her to take care of their dog, but which had yet to be dispatched by the ARC at the time of Mr Costi’s death, was highly consistent with an intent on Mr Costi’s part to take his own life. She states: 'This was exactly the type of note that I was expecting. I knew there would have to be something about the dog.' 20

4.12. Mr Peter Kyriacou was a cousin of Mr Costi. He gave a statement to the police which was tendered to the Inquest21. Mr Kyriacou states that when Mr Costi was taken into custody he was really angry that he was back in jail. However, as time went on Mr Costi became calmer and began to accept his situation22. Mr Costi never said anything that would have indicated that he might hurt himself. Mr Kyriacou did not think that his cousin was depressed about anything. The last couple of visits that Mr Kyriacou paid to Mr Costi at the ARC were positive. He said that Mr Costi had been very relaxed.

The last visit occurred on 24 November 2013. Mr Costi had been relaxed and easy going. There was nothing about Mr Costi that gave him any cause for concern.

Mr Kyriacou did not believe that at that time he was depressed or showing any signs of a change in behaviour.

4.13. As far as Mr Costi’s relevant medical history is concerned the police investigation revealed that in 2010 he consulted a psychologist, Paul Kassapidis, on three occasions in relation to unresolved grief reactions and associated difficulties in adjusting to the death of his wife that had occurred in 2000. This had manifested itself in drug use, broken relationships and a self-harm incident. In this regard Mr Costi had referred to an incident in which he had taken oxycontin and had consumed two bottles of Jack Daniels. He had conceded to Kassapidis that he had a gun and had wanted to kill himself. He referred to having placed the gun to his head but the chamber with the bullets had fallen out. Mr Costi had then spoken of the incident in October 2009 in which he had lacerated himself with a Stanley knife after which he had been arrested by police. This of course was a matter that in 2012 and 2013 was known to the authorities including the police and DCS. Mr Costi had indicated to Mr Kassapidis that he believed that drugs were the main cause of his problems but denied current use of drugs or alcohol. At a later consultation he mentioned to Mr Kassapidis that he had 20 Exhibit C21, page 14 21 Exhibit C22 22 Exhibit C22, page 2

reduced his drug intake which seemed inconsistent with his earlier assertions of abstinence.

4.14. I doubt that detailed knowledge of Mr Costi’s medical history had it been made known to correctional authorities would have made any difference to the outcome. As seen, the authorities were aware of historical matters that were germane to the issue of his perceived risk.

4.15. I have referred to Mr Costi’s initial screening when inducted into the ARC in December

2012. This was the only formal screening or formal risk assessment process that he underwent during his period of remand in custody. In the past this Court has had occasion to comment upon the lack of routine screening and risk assessment during a remand prisoner’s incarceration in the ARC. I refer here to what was said by this Court in its findings in relation to the death of Damian John Cook who in 2003 had committed suicide by hanging in the same Unit at the ARC in which Mr Costi would hang himself.

In the Court’s findings the point was made that a prisoner’s initial stress screening on induction was essentially a snapshot in time and that circumstances of an inmate may change for the worse during his period of incarceration, a change that may not be detected by DCS personnel. Such changes might occur when a remand prisoner experiences an adverse court outcome, for example a failed bail application. However, I see nothing in Mr Costi’s Offender Case Notes from October and November of 2013 that gives any hint of a negative frame of mind following the postponement of his trial and failed bail application or anything that might signify an enhanced risk of self-harm.

I will return to this issue later.

4.16. When the matter of Mr Costi is examined as a whole there is nothing to suggest that a suicidal frame of mind was or ought to have been detected in November 2013. It is true that there was evidence to suggest that Mr Costi had become less gregarious and had started to keep to himself, but this in itself would not reasonably have led DCS personnel to suspect that he was experiencing suicidal ideation. There was in reality nothing to suggest that in November 2013 Mr Costi should have been kept in dual accommodation. In any event I suspect that by then Mr Costi was savvy to the matters that would have ensured his continued single cell accommodation, such as refraining from sharing anything about himself or his frame of mind with anybody who might have the power to alter that arrangement.

5. The night of Mr Costi’s death at the ARC 5.1. As indicated above, Mr Costi occupied single cell accommodation in Unit 4 of the ARC. Unit 4 was one of a number of self-contained accommodation units within the ARC. At night it would only be occupied by prisoners. Correctional staff were situated in another location. During lockdown each individual cell would be locked. As well, the door to the Unit itself would be locked. During the shift in question, only patrolling prison officers would have access to the Unit. CCTV covered the Unit but not the individual cells. The CCTV was monitored by staff at a remote location in the ARC.

5.2. Lockdown occurred at 4:40pm daily at which time each individual cell within the Unit was locked with a key that was retained by DCS staff on duty. Prisoners would remain locked in their cells until about 8:20 to 8:30 the following morning. In the normal course of events at about 6pm each evening, every individual cell would be locked with a master key. Thus the cell door would be doubly locked, firstly with the keys available to correctional officers and secondly with the master key. The master key was kept at a central location under the control of the shift supervisor. The situation of double locking would prevail until 6:00am the following morning when officers would withdraw the master key and unmaster each cell within a Unit. However, the cell would still be locked with the general lock until 8:20 to 8:30am at which time prisoners would be permitted to leave their cells.

5.3. Patrols of each Unit, involving the shining of a torch into each cell to see that all was in order, were conducted on a two-hourly basis during the shift in question.

5.4. A fresh shift of correctional officers commenced at midnight and ended at 8am. It was during this shift that Mr Costi undoubtedly died.

5.5. It will be accepted that if patrols were conducted strictly in accordance with the two hour interval, and if the cell had an available hanging point together with material from which a ligature could be fashioned, this interval would have instilled in a singly accommodated prisoner enough confidence to carry out an act of self-hanging without detection.

5.6. I was satisfied by the evidence of Mr Aaron Lee, a DCS officer on the midnight to 8am shift and who conducted the 4am patrol with another DCS officer, Ms Scott, that at about 4am he checked the upper cells within Unit 4 and found nothing untoward in

Mr Costi’s cell. In particular I accepted his evidence that if Mr Costi had been in the position and location in which he would be found dead approximately two hours later, and/or with bags or a pillow over his head, he would have been able to view that with the use of his torch. I was also satisfied that if he had seen any scenario of that kind he would have raised the alarm straight away. Mr Lee convinced me that no such scenario existed at the time he inspected the cell at approximately 4am on the morning in question. I find that Mr Costi was alive at that time.

5.7. The 6am patrol was conducted by Mr Lee and Ms Sharne Scott. On this patrol it was Ms Scott who went to Mr Costi’s cell which was on the upper floor of Unit 4. Mr Lee examined the cells on the lower level of the unit. This was the reverse of the situation that had prevailed at the 4am patrol. It was at the 6am patrol that the cells would be unmastered. Both Ms Scott and Mr Lee had taken possession of master keys for the purpose of unmastering the cells within that Unit. In Ms Scott’s witness statement23 she states that she arrived at the door of Mr Costi’s cell at 6:07am and shone her torch into the cell. She first noticed that Mr Costi was not on his bed. She then saw that he was in the position in which he had hung himself and was wrapped in a blanket. She called for other officers to attend. The door was unlocked and opened. Mr Costi had a white bin liner around his head. A Hoffman tool was used to cut fabric around Mr Costi’s neck and wrists. Medical assistance was called by radio and CPR was commenced. Medical staff arrived at the location at 6:09am. Ambulance officers arrived at 6:19am. When found, Mr Costi was already beyond medical intervention.

5.8. I was satisfied that the door to Mr Costi’s cell was locked at the time Ms Scott approached the door. It was locked with both the cell key as well as the master key.

There was no other person in Mr Costi’s cell at the time the cell door was opened.

5.9. The three correctional officers who were present when the cell was opened were called to give oral evidence. They were Mr Lee, Ms Scott and Mr Peter Crompton who was the shift supervisor that night. I was satisfied from their witness statements and their oral evidence that Mr Costi’s cell was locked and mastered prior to it being opened upon the discovery that he was deceased. Furthermore, I was satisfied that no person was in the cell with Mr Costi at the time it was entered by correctional officers. Mr Crompton gave oral evidence from which it can be concluded that no person had access 23 Exhibit C43

to the master key from the time Mr Costi’s cell was mastered on the evening of the day before and when his cell was opened upon the his discovery just after 6am on the morning in question. Furthermore, Mr Crompton’s evidence persuaded me that CCTV monitoring of the Unit would have revealed the unexpected and unauthorised entry of a person into the Unit during the course of that night and that had that occurred, steps would have been taken to investigate any such entry. There had been no such entry detected. Accordingly, I was satisfied that: a) Mr Costi’s cell had been locked with him as the sole occupant from approximately 4:40pm on the afternoon of 26 November 2013 until just after 6am the following morning; b) The cell had also been mastered from approximately 6pm on the evening of 26 November 2013 until just after 6am the following morning when it was opened; c) No person entered Mr Costi’s cell at any stage between 4:40pm on 26 November 2013 and just after 6am the following morning when he was located deceased in his cell; d) There was no unauthorised entry by any person into Unit 4 during the night in question; e) There was no unauthorised removal of the master key during the night in question; f) Mr Costi was alive at the time of the 4am patrol.

g) Mr Costi was solely responsible for his hanging and death.

6. Routine prisoner observations 6.1. The requirement was that a cell observation should not take place at intervals of greater than two hours. Adduced in evidence was the fact that only a few days prior to this incident there had been a direction that the two-hourly intervals be conducted randomly.

The amended Standard Operating Procedure 97 (SOP 97) stated that patrols should be conducted randomly and not always in the same sequence to avoid becoming predictable. This measure was intended to induce in the minds of prisoners a state of uncertainty as to when the next prisoner observation would be made; in other words, so that prisoners could not be confident that any misbehaviour within a cell could be carried out without detection in the period between observations.

6.2. I need say very little about this except to say that it appears that the correctional officers involved in this matter were not aware of that initiative at the time of this event.

However, Mr Darren Hosking who is the General Manager of the ARC and who gave evidence in this Inquest, told the Court that having regard to the manner in which the entirety of the ARC units were subject to routine observation, and the manpower available to carry out that exercise, even with ‘random’ two-hourly observations a prisoner could have expected to be left alone for at least 90 minutes. Indeed, Mr Hosking conceded that in those circumstances Mr Costi would have expected to have been left alone for the 90 minutes following the patrol that had occurred at about 4am24.

Mr Hosking was asked this question by me: 'Q. He would have been reasonably confident about that, even with random patrols in the sense as you've described them.

A. Yes.' 25 Mr Hosking thus conceded that the randomness initiative that was contained within SOP 97 was not a substantial change at all26.

6.3. It seems to me that even if Mr Costi’s cell had been observed randomly in accordance with the new initiative, he had available, and could have expected, 90 minutes in which to carry out his suicide. To my mind this would have been ample time. Thus any element of randomness to the observations that were made of him during the course of that shift would not have made any difference to the outcome.

6.4. I make an observation that it is not the randomness of observations that is at issue here, but the infrequency of observations that is the main concern. I intend making a recommendation that prisoners be observed hourly and randomly during the shift in question.

7. The hanging point 7.1. As indicated earlier this was an unusual hanging point. For many years now this Court and other Coroners Courts have framed recommendations to the clear effect that 24 Transcript, page 148 25 Transcript, page 148 26 Transcript, page 149

hanging points should be eliminated from cell accommodation within prisons. Mr Hosking was naturally asked about how the hanging point in this case might be eliminated, as other hanging points in institutions have been over the course of the last 20 years or so. Mr Hosking acknowledged that he would obviously have to give some serious thought as to how this could be eliminated. He regarded the hanging point here as a ‘non-obvious ligature point’27. That is true.

7.2. I am not certain how the hanging point in this particular case could be eliminated in every cell that is subject of single occupation. Even dual occupation cells might contain this particular hanging point, and suicide in dual cell accommodation is not completely unheard of. All that needs to be said is that regardless of how obscure this hanging point might be to a prisoner intent on suicide, the fact is that it has now been used as such. The Department for Correctional Services thus has been and continues to be on notice that this item can be used as a hanging point and that there is a need to take this into consideration in safe-cell design.

8. Conclusions 8.1. The Court drew the following conclusions:

1. Mr Costi was solely responsible for his hanging between 4am and 6am during the morning of 27 November 2013. His cause of death was neck compression (due to hanging) and plastic bag asphyxia.

2. Mr Costi was able to carry out the act of hanging in his single cell accommodation by virtue of the fact that there was material within the cell from which he could fashion a ligature and there was an available hanging point that he could affix the ligature to. The hanging point was an unusual hanging point. It would have been very difficult for Department for Correctional Services’ authorities to predict that this item could be used as a hanging point.

3. Mr Costi had some history of self-harm dating back to 2009. The incident in 2009 had led to a period of custody within the ARC. Mr Costi had been treated as a prisoner at risk on this occasion.

4. On 12 December 2012 Mr Costi was again taken into custody in the ARC. He 27 Transcript, page 167

would remain in custody until his death on 27 November 2013. Mr Costi’s history of self-harm was well known to both SAPOL in whose custody Mr Costi had originally been in, to the Department for Correctional Services and to the SA Prison Health Service. In my opinion Mr Costi was properly assessed when he was first received into the ARC in December 2012. Ultimately he was considered to be at low risk of self-harm. He was accommodated in single cell accommodation in Unit 4 of the ARC. To my mind only in hindsight could it be said that it would have been better if, having regard to Mr Costi’s previous history of self-harm, he had been accommodated in double cell accommodation at all times, particularly having regard to the fact that observations of Mr Costi were only conducted on a twohourly basis. I am mindful of the fact that at no stage during his period of remand in custody in 2012 and 2013 did he exhibit any misbehaviour or any other behaviour that suggested that he was at risk of self-harm in single cell accommodation. It has to be borne in mind that by the time of his death he had been incarcerated for a period of 351 consecutive days without incident.

5. Mr Costi was remanded in custody in relation to matters that had been listed for trial in the District Court on 9 December 2013. Mr Costi’s trial was postponed because it was thought that it would not be completed by Christmas of that year.

On 21 November 2013 a bail application by Mr Costi was refused. Despite grave suspicion to the contrary, there is insufficient evidence to make a positive finding that the postponement of Mr Costi’s trial and the subsequent unsuccessful bail application contributed to his decision to take his own life.

6. No formal risk assessment or prisoner screening occurred in respect of Mr Costi from the time he was first inducted into the ARC in December 2012 until his death.

In particular, no formal risk assessment or screening was conducted either when his trial was postponed or when his bail application was refused.

7. During the shift in question, namely between midnight and 8am, Mr Costi was the subject of routine observations on a two-hourly basis. The observations were not conducted randomly. In fact they were conducted in a manner that was predictable and in circumstances whereby a prisoner could be confident that there would be sufficient time to carry out a complicated act of self-harm without fear of detection.

This in fact was the case with Mr Costi. Prior to the occasion in question a new Standard Operating Procedure had been promulgated to the effect that the twohourly observations should be conducted randomly and at intervals that could not

be predicted. However, I find that having regard to the manner in which the routine observations were carried out within the entirety of the ARC, a prisoner could nevertheless have been reasonably confident that there would be at least 90 minutes between routine observations even if the observations had been carried in a socalled random fashion. This also would have been enough time for Mr Costi to have carried out his intentions.

8. I find that in many senses there was a pointlessness to two-hourly observations whether conducted randomly or not given that two hours was sufficient for Mr Costi to carry out a complicated method of self-destruction. The two-hourly observations in many senses had an element of futility in terms of prisoner safety in single cell accommodation.

9. Recommendations 9.1. Pursuant to section 25(2) of the Coroner’s Act 2003 I am empowered to make recommendations that in the opinion of the Court might prevent, or reduce the likelihood of, a recurrence of an event similar to the event that was the subject of the Inquest.

9.2. There are a number of areas that the Court needs to consider in relation to the topic of recommendation. They are whether there is a need for regular formal prisoner screening and risk assessment of remand prisoners during the currency of their remand in custody, whether there is a need for the Department for Correctional Services to reconsider the frequency of routine observations of prisoners and whether there is a need to take into account the hanging point in this case in terms of prisoner cell safety.

9.3. Formal prisoner screening and risk assessment As indicated above this Court considered this issue many years ago in the Inquest into the death of Damian Cook. In that Inquest the Court made certain observations about the lack of regular screening and risk assessment of remand prisoners and made a recommendation in the following terms: 'That the DCS establish a panel to examine the feasibility of introducing a regime whereby inmates at the ARC are formally screened for risk of self-harm on a more regular basis.' In this Inquest I received into evidence the affidavit of Jane Farrin28 who is the Director 28 Exhibit C46

of the Sentence Management Unit, Offender Development Directorate at the Department for Correctional Services. Ms Farrin’s affidavit makes reference to the recommendation in the matter of Damian Cook. She describes the High Risk Assessment Team (HRAT) procedures as well as the Notification of Concern (NOC) procedures to which I have already referred. The affidavit indicates that the HRAT process was introduced in response to the death of Damian Cook in 2003. Ms Farrin also describes a process pursuant to Standard Operating Procedure 90 which was established as a whole of agency approach to the management of prisoners identified as being at risk of self-harm/suicide. The procedure was developed in response to a number of coronial recommendations including the recommendation made in Cook.

The process involves any prisoner who is identified as being at risk of suicide or selfharm having an NOC raised and an Initial Response Plan put in place. SOP 90 also requires that if a prisoner on a HRAT regime returns from court in a distressed state, the staff member observing the prisoner’s distress must immediately report their observations for further action. I would point out that Mr Costi was not on HRAT at the time of his court appearances in October and November 2013, nor at the time of his death.

9.4. Ms Farrin indicates that SOP 90 3.3.7 states: '.. when an at-risk prisoner returns from court (including Family Court and the use of Video Conferencing facilities), Police Escorts, Medical appointments, compassionate leave, or are transferred in from other locations, correctional officers are to check the prisoner and if there are any signs of distress or if the officer has any concerns they must immediately report their observations to the Responsible Officers for action.' 29 Again this appears to apply to HRAT prisoners, that is to say prisoners at risk. I am not certain why this procedure should be confined to HRAT prisoners.

9.5. Regarding regular screening and risk assessment Ms Farrin’s affidavit states as follows: 'The implications for ongoing risk screenings occurring for all prisoners at the ARC and other remand prisoners would be impracticable and resource intensive. This is because regularly screening would be required for all types of escorted leave including when returning from court (including the Family Court and the use of Video Conferencing 29 Exhibit C46

facilities), police escorts, medical appointments, compassionate leave or transfers. I am not aware of this practice occurring in any other correctional jurisdictions.' 30 However, the affidavit indicates that SOP 90 has recently been reviewed and that at the time of the swearing of the affidavit in May 2017 there was a draft amendment in existence which would mandate correctional officers to be vigilant for any signs of distress when prisoners ‘particularly HRAT prisoners’ return from court. If correctional officers identify any concerns, they are obliged to notify a responsible officer and complete an NOC. If the amendment has not yet been promulgated I would recommend that it is promulgated in principle, with a modification that it should apply to all remand prisoners in equal measure.

9.6. Frequency of observations I have already commented upon what some might say to be the futility of two-hourly observations conducted even randomly where, having regard to the exigencies of human resources, a prisoner during the shift in question could still reasonably expect at least a 90 minute gap from one routine observation to the next.

9.7. I received into evidence the affidavit of Michael John Reynolds31. Mr Reynolds is the Director of Operations of State-Wide Operations at the Department for Correctional Services. Mr Reynolds refers to the amended SOP 97 that was promulgated only a few days before Mr Costi’s death and which stated that the two-hourly patrols should be conducted randomly and not always in the same sequence to avoid becoming predictable. Mr Reynolds’ affidavit refers to discussion that took place during the course of this Inquest about the undesirability of two-hourly intervals and whether or not hourly patrols would be more appropriate. The affidavit states as follows: 'If the requirement was set for hourly patrols then this would require the officers to commence almost immediately after their last patrol and would restrict the ability for truly random patrols to occur. CSOs are also required to patrol the whole of the facility and check all locations and not just the cells. A requirement for hourly patrols could mean that these security checks are compromised due to time constraints.' 32

9.8. Clearly, therefore, the question of frequency of patrols involves a human resource issue.

The question of the undue disturbing of prisoners in the middle of the night would also need to be considered. However, it seems to me that based on the experience in this 30 Exhibit C46, paragraph 7.8 31 Exhibit C45 32 Exhibit C45, paragraph 8

particular case observations conducted truly randomly and no less frequently than onehourly would be highly appropriate and significantly more effective than two-hourly observations.

9.9. Hanging point I repeat the observation made above that the Department for Correctional Services has been and is now on notice that there is a hanging point in cells that needs to be addressed.

9.10. The Court makes the following recommendations directed to the Minister for Correctional Services and the Chief Executive Officer of the Department for Correctional Services:

1. That the Department for Correctional Services implement, if it has not already been implemented, the draft revised paragraph 3.3.9 of SOP 90 in the terms set out in paragraph 7.9 in the affidavit of Jane Farrin dated 4 May 2017 and that the requirement contained therein be stated to apply to all prisoners regardless of whether they are HRAT prisoners or not.

2. That the Department for Correctional Services amend SOP 97 to ensure that observations of prisoners are conducted at intervals no less frequently than onehourly and are conducted randomly and not always in the same sequence to avoid becoming predictable. I further recommend that the necessary human resources be made available to the Department for Correctional Services in order to implement this recommendation.

3. That the Department for Correctional Services ensure that safe-cell design takes into account the hanging point that was used by Mr Costi to hang himself.

Key Words: Death in Custody; Suicide; Prison; Hanging Points In witness whereof the said Coroner has hereunto set and subscribed his hand and Seal the 17th day of January, 2018.

Deputy State Coroner Inquest Number 02/2017 (1979/2013)