Fifth Independent Review Committee on Non-natural deaths in custody that occurred between April 1st, 2017 to March 31st, 2019

Acknowledgements

There were many people who were helpful to us as we conducted our review. First and foremost were the staff of the Incident Investigations Branch (IIB). Tara Tremblay, as the CSC coordinator for the IRC, helped us navigate through the CSC system and access the documents we needed, and we are very grateful for her assistance. Mark Nafekh, Mel Sater, and Isabelle Bastien made themselves readily available to the committee throughout the process. We had a lot of questions about their investigative policies and operations, and they unfailingly answered them. We thank them for their engagement and support of our work.

We conducted interviews with numerous CSC staff in various operational areas. We appreciate their giving us their time, particularly those who were facing additional work pressures related to the pandemic. Similarly, we thank those we interviewed outside of CSC: a Community Board Member, a retired National Investigator, officials at Public Safety, the Correctional Investigator and staff from his Office, the former Correctional Investigator, chaplains, representatives of community-based organizations who provide support to inmates and their families, and family members of inmates who had died while in CSC custody. We are grateful to Professor Fannie Lafontaine for sharing with us her expertise and experience in the role of independent observer of a major investigation. Our work benefitted from the information and observations they provided.

Finally, we wish to recognize the contribution of officials (whom we interviewed) from agencies with responsibilities for investigating deaths in prisons in other countries, the Deputy Ombudsman at the UK Prisons and Probation Ombudsman; the Prisoner Ombudsman for Northern Ireland and a senior staff member; the Chief Inspector of Prisons, Office of the Inspectorate, Department of Corrections, New Zealand; and senior officials from the Justice and Review Office at the Department of Justice, Victoria, Australia as well as the New South Wales Ombudsman's Office. They were generous and accommodating. We learned a great deal from these inquiries, and this is reflected in our report and its recommendations.

Introduction

There is a significant history surrounding non-natural deaths in Correctional Service Canada (CSC) institutions and CSC's responses to the recommendations of investigations into the circumstances of these deaths. The Correctional Investigator in his 2005-06 Annual Report raised concerns regarding the timely completion of meaningful investigation reports and action plans developed by CSC to address the underlying issues identified in these reports. As a first step, the Office of the Correctional Investigator (OCI) engaged an academic to undertake a study of non-natural deaths in Correctional Service Canada (CSC) institutions (Gabor, 2007). The report on this study provided some preliminary findings regarding recurring factors contributing to deaths in custody (e.g. lack of information sharing, security lapses, post-incident emergency care, mental health programming, and illicit drugs) and identified some areas for further study. Following this report, CSC has commissioned a series of Independent Review Committees (IRCs) with a mandate to review its Board of Investigation (BOI) reports; policies and procedures related to the investigative process; corrective measures and action plans in response to BOI recommendations; strategies to prevent non-natural deaths in custody; and best practices in other jurisdictions. Reviews have been completed by four IRCs and this is the report of the 5th IRC.

In addition, CSC has constructed and maintained a database on deaths in custody, and analyses of these data are presented in its Annual Report on Deaths in Custody. The most recent Annual Report analysed data over an 8-year period (FY 2009-10 to FY 2016-17), and included analyses on the cause of death, regional distribution, recommendations from BOI reports, profiles of offenders who have died in custody, and compliance issues related to the death.

Although the general mandate (as described above) of the IRCs has been common across the five IRCs, the specific Terms of Reference in the Convening Orders for the committees has varied as the policies and processes related to investigations have evolved, partly in response to the recommendations of earlier IRCs. The Terms of Reference (ToR) for the 5th IRC were as follows:

• analyzes the impact of the changes CSC has made to its investigative process in response to the recommendations of the 4th IRC report;
• analyzes the independence and impartiality of CSC's investigative process;
• provides suggestions on how CSC can share its investigative reports with staff to ensure broader awareness of "lessons learned" and make them more accessible to the public;
• assesses whether recommendations made by Boards of Investigation appropriately address the underlying causes that led to the incident(s), as well as any need for systemic improvements to policy and procedures;
• assesses CSC's actions and responses to incident investigation reports, including the appropriateness and adequacy of the corrective measures and action plans initiated by CSC in order to address the investigation recommendations and underlying issues;
• examines international standards and practices in other jurisdictions with respect to their investigative process is in general, and specifically in relation to deaths, and how the standards and practices could further inform CSC's investigative process.

In addition to these Terms of Reference, the committee was asked by IIB to review a section of the 2017/18 Annual Report of the Office of the Correctional Investigator, titled Special Focus: Investigation into the Riot at Saskatchewan Penitentiary. This OCI investigation concluded with significant findings on the investigation process and a recommendation that the Minister of Public Safety conduct an independent review of Section 19 of the Corrections and Conditional Release Act, the section that requires, and authorizes, CSC to conduct an investigation into incidents that result in the death or serious bodily injury of an inmate. These matters were taken up by this Committee and added to our work, particularly the analysis of the independence and impartiality of CSC's investigative process, which accounts, in part, for the length of the chapter on ToR #2.

In the course of our review we interviewed, at length, senior management at IIB. We interviewed National Investigators and a Community Member, all selected at random by us. We also interviewed a retired CSC investigator who had sat as a member on several BOIs.

At our orientation session in February 2020, we received presentations from several CSC staff on a number of topics, including Structured Intervention Units, the Health Services Clinical Framework and the impact of ATIP on IIB's work. We were also given a detailed briefing on the Saskatchewan Penitentiary Riot, and were informed that this IRC was referred to in the Minister's response to a recommendation from the OCI following his critique of the BOI report on the riot in a segment of his 2017/18 Annual Report. We spoke to senior CSC members based at Institutions who had been involved in IIB death investigations as a witness.

We interviewed three family members of inmates who had died while in CSC custody, as well as representatives from five other organizations who deal with inmate families. We also spoke to three Chaplains who work in the CSC system.

We met with Office of the Correctional Investigator staff and interviewed the Correctional Investigator and OCI Legal Counsel. We also spoke to the previous Correctional Investigator, Howard Sapers and reviewed material on deaths in custody from his 2017 Independent Review of Ontario Corrections report.

We interviewed other CSC officials, including senior staff at Values, Integrity and Conflict Management as well as Strategic Policy, Performance Management and Research. We also interviewed officials in the Corrections and Criminal Justice Directorate at Public Safety.

We contacted the Union of Canadian Corrections Officers (UCCO) but received no response.

We spoke with senior staff members from the Office of the Chief Coroner of Ontario and the Government of Alberta, Department of Corrections.

We also spoke to agencies that investigate deaths in custody of the state, including the Province of Ontario Special Investigations Unit (SIU) and the Office of the Ontario Ombudsman.

We interviewed Professor Fannie Lafontaine, who is the government of Quebec appointed Independent Observer of investigations being conducted into allegations of offences by police officers against Indigenous people.

We contacted a number of agencies across the world who conduct investigations into deaths in prison custody. We interviewed, at length, the Deputy Ombudsman at the UK Prisons and Probation Ombudsman; the Chief Inspector of Prisons, Office of the Inspectorate, Department of Corrections, New Zealand; the Prisoner Ombudsman for Northern Ireland; senior staff from the Justice and Review Office at the Department of Justice and Community Safety, Victoria, Australia and a senior official at the New South Wales Ombudsman's Office. We also obtained a significant quantity of documents from these organizations. We reviewed information on their websites, including systemic and individual reports into deaths in custody in their jurisdictions.

IIB asked the Committee to review the OCI's critique of CSC's BOI of the Saskatchewan Penitentiary riot and death on December 14, 2016, included in a segment of the OCI's 2017/18 Annual Report. The Minister of Public Safety responded to a key recommendation emanating from that segment, which made reference to the 5th IRC. We were provided with CSC's BOI report of the Saskatchewan Penitentiary riot as part of that process.

We studied a large number of other documents in the course of our review, including the 25 Board of Investigation reports into non-natural deaths that fell within our mandate, along with any corresponding National Investigation Meeting (NIM) reports. We also reviewed the BOI on the assaults against the Protected Status inmates at Edmonton Institution from 2018. While there were no deaths in this review, it was a report of interest in terms of CSC's culture and an example of the newer style of IIB report writing.

We also obtained and reviewed CSC/IIB policies, Commissioner's Directives and procedures relating to the investigation of deaths in custody, family engagement, victim engagement, chaplaincy, facilitated disclosure, and discipline. IIB shared its investigator job description, responsibilities, training materials, conflict of interest guidelines and workflow chart. We were provided with information on the backgrounds of Community Members.

We reviewed previous IRC reports, in particular the 4th IRC and CSC's response to it, the OCI Annual Report 2017/18, as well as CSC's Annual Report on Deaths in Custody 2016/2017.

We studied CSC's material concerning media relations, information sharing, notification of death training package, and reviewed several types of communication platforms and broadcasting mechanisms available to the service.

We also reviewed CSC's research on the Summary of Findings of the 1995 CSC National Inmate Survey and the methodology and surveys behind the more recent Multi Overdose interviews from IIB Enhanced Investigation Analysis.

We reviewed the research literature on Measuring the Quality of Prison Life (MQPL).

As we note on several occasions in this report, we were unable to cast our investigative net as broadly as we would have wished, due to time constraints imposed by CSC when engaging the members of this Committee to conduct the review.

For the same reason, a lack of time, we were unable to dig as deeply as we would have wished into how each of the BOIs was conducted. We simply did not have the time available to obtain and review the entire investigative product of each, or indeed any, BOI. This material could have included investigation plans, interview plans, transcripts of interviews, copies of documents, BOI members notes, emails to, from and between BOI members and so on. Review of this material would have certainly given additional insight into the investigative quality of that particular BOI.

To be clear, we are not in any way suggesting that there were any barriers to obtaining this information, had we requested it.

In summary, a lack of time somewhat fettered our ability to forensically determine the level of independence, impartiality and thoroughness of any given BOI. Ultimately, we relied on the contents of a given BOI report itself to make our assessment. While we are comfortable with our conclusions, they are not based on an exhaustive review of all potentially relevant information.

All that said, we found no prima facie evidence that, had we been able to dig deeper, we would have found information that contradicted our ultimate conclusions. We are confident that our findings and recommendations are based on a sufficiently sound evidential foundation, as we explain in the report itself.

The organization of this report follows the ToRs as listed above. Accordingly, the first chapter addresses the impact of the changes CSC has made in its investigative process in response to the recommendations from the 4th IRC. Subsequent chapters cover the issues of the independence and impartiality of CSC's investigative process; how CSC can better share its investigative reports and "lessons learned" with staff and make them more accessible to the public; whether recommendations made by BOIs appropriately address underlying causes; and the appropriateness and adequacy of the corrective measures and action plans initiated by CSC. The 6th ToR, which required the committee to examine international standards in other jurisdictions with respect to their investigative process, is not the subject of a separate chapter but rather is incorporated in other chapters, as appropriate, in particular in Chapter 2. While the committee conducted its work collaboratively, the lead for each ToR was assigned to a committee member who authored the chapter of the report on that ToR. The assignments were as follows: Robert Cormier, ToR 1 and ToR 5; Gareth Jones, ToR 2; and Louise Leonardi, ToR 3 and ToR 4.

Recommendations are made throughout the report, and are summarized in Annex A.

Chapter 1

• Analyzes the impact of the changes CSC has made to its investigative process in response to the recommendations of the 4th IRC report

In order to address this element of its mandate, the committee began by reviewing the recommendations of 4th IRC report, and determined that five of the 17 recommendations in the report addressed issues related to the investigative process. We received a document (Corrective Measures and Management Action Plan) prepared by the Incident Investigations Branch (IIB) which provided CSC's response to each of the recommendations of the 4th IRC. In addition, we were informed that, although the 4th IRC report was not finalized for publication until November 2018, action on the recommendations began when the report was received by CSC officials in July 2018.

The task for the committee was to examine each of the five relevant recommendations and the CSC response to the recommendation; identify the changes that followed from these responses; and assess the impacts of these changes. In a significant number of the cases reviewed by the committee the Board of Investigation had been convened prior to July 2018. For the purposes of this task, we examined the reports of those BOIs that were convened in July 2018 or later. Ten of the 25 BOI reports fell into this category. With this caveat, our review and analysis for each of the recommendations is presented below.

To explore, in its investigation terms of reference, the inclusion of i) CSC's core values of dignity and respect for inmates, and ii) international standards such as the United Nations standard minimum rules for the treatment of prisoners, as criteria relevant to CSC incident investigations for suicides that take place in segregation.

This recommendation was made by the 4th IRC following an examination of a case where the inmate committed suicide in a condition of despair after a prolonged period of segregation, and the committee concluded that the decisions taken by CSC staff failed to attend appropriately to CSC's Core Values (Respect; Fairness; Professionalism; Inclusiveness; Accountability, see CD 001) and the internationally recognized standards for the treatment of prisoners. In its response to this recommendation, CSC indicated that Board members were instructed to consider CSC's mission and core values as well as international standards as criteria relevant to their investigations.

The committee did not find any examples where the BOI explicitly linked their findings and supporting facts to CSC's core values or international standards. However, there were a few cases where Board Members were attending to the core values in their investigation even if they were not explicitly mentioned. For example, the BOI of a suicide that occurred in segregation prior to July 2018, which is described in a later section (ToR #5) in this report, does not make any specific reference to CSC's Core Values or international standards for the treatment of prisoners, and yet was remarkable in demonstrating respect for the offender by carefully examining the events from the perspective of his experiences and his needs. This committee is in agreement with the 4th IRC regarding the importance that should be accorded to Values and Ethics in the investigative process. Accordingly, we make the following recommendation.

The Values and Ethics of Correctional Service Canada should be referred to in the Board of Investigation Terms of Reference, be a focus of the investigative process where relevant, and included in findings and recommendations, as applicable.

The committee noted that, according to the Annual Report on Deaths in Custody 2016/17, approximately 39% of the suicides over an 8-year timeframe occurred in segregation. Since the writing of the 4th IRC report, the Corrections and Conditional Release Act has been amended such that the segregation provisions have been abolished and structured intervention units (SIUs) introduced. Given that the SIUs provide a level of engagement with offenders that was not the case with segregation, there may be fewer suicides in the SIUs than there were in segregation. It will be important to continue the collection of data on the location where deaths occur in order to see whether this is borne out. CSC indicated in its response to the 4th IRC recommendation that BOIs would continue to take into consideration the core values and international standards in the new context of the SIUs.

CSC incident investigations examine all four pillars of addressing problematic substance use to inform prevention, treatment, harm reduction and enforcement strategies.

This recommendation by the 4th IRC was prompted by the committee's observation that the investigation of suicides focused almost exclusively on surveillance and enforcement strategies with little inquiry into the other three pillars. In response to this recommendation, CSC indicated that the IIB instituted an Enhanced Investigation Analysis approach to investigating fatal and non-fatal overdose incidents using focus groups and interviews/questionnaires centred around the four pillars. The information gathered from this work has been used to report on the development of prevention and harm reduction strategies. We consider this initiative to be a valuable way for CSC to broaden its understanding of problematic substance use in penitentiaries and improving strategies to prevent overdoses, and in our view, it should be continued.

In addition, the CSC response indicated that, "As part of their work, incident investigators routinely review the care, treatment and monitoring received by an offender and are required to make appropriate recommendations when necessary." The committee examined the relevant subset of BOI reports for evidence of this. The Committee reviewed four cases involving overdose where the investigation was convened after July 2018. In one case, the BOI contained detailed information on his prior drug use as well as his engagement in treatment and harm reduction programs. In another case there was a description of his drug use and the link to his criminal record as well as a reference to his recent participation in a general correctional program. A third case was unusual in that he was not known to have used substances in the past although he had convictions for possession and trafficking in controlled substances, and this was documented in the BOI. The fourth case included the Enhanced Investigation Analysis approach with pre-Board interviews and an advance survey prior to the arrival of the BOI.

Overall, the Committee concluded that the inclusion of information concerning the inmate's engagement in programs and strategies to address substance use provided a fuller picture of the circumstances leading up to the incident, and a broader framework for identifying issues for the report. CSC has made important strides in its investigations of drug overdoses by attending to the four pillars. Building on the 4th IRC recommendation, we make the following recommendation.

"To examine all four pillars" be included in the Terms of Reference in cases involving overdoses.

CSC policy for incident investigations encourage incident investigators to go beyond assessing whether or not specific policies are adhered to and, in a dedicated section of their reports, highlight any findings and recommendations regarding improper practice, policy gaps and underlying issues.

This recommendation was made by the 4th IRC following its review of a case where the BOI report contained a finding of a regular institutional practice that failed to meet the tests of transparency and proper decision-making, without a recommendation that it be corrected. The CSC response to this 4th IRC recommendation indicated that the BOI report format was modified to include a section titled "Key Issues of Non-Compliance and Underlying Issues." This section is designed to pinpoint the key policy gaps and underlying issues that directly support the recommendations, so that similar incidents can be prevented from occurring in the future. The committee found that this section is being routinely used, and represents a significant improvement to the BOI reports. We applaud CSC for this initiative and encourage continuation of this practice.

That the terms of reference for an investigation into a death in custody require, in those cases where the investigation finds multiple, serious failures to comply with policies, that the investigators examine factors related to the environment and operations at the site. These factors would include policies, plans and procedures that impact a healthy and respectful workplace, any workplace reviews or staff surveys, complaints and grievances by offenders, or any other warning signs that may have foreshadowed the incident.

This recommendation of the 4th IRC flowed from its review of the Hines case, which was an "extremely troubling" (Nadeau, 2018) incident where serious violations of policy, including repeated inappropriate deployments of inflammatory agent against a mentally ill offender who was not aggressive, contributed to his death. The 4th IRC committee was struck by such a gross departure from the standards of practice expected of CSC, taking into consideration its Mission and Core Values. The committee concluded: "Given the constellation and magnitude of the breaches that led to this death in custody, it is difficult to imagine that there were no antecedent interactions that would have foreshadowed the extreme, security-focused response with such little regard for the well-being of the individual. Were there warning signs in the months leading up to the incident?" (p. 55)

The CSC response to this recommendation was twofold. The first was that the IIB would emphasize the importance of developing terms of reference that are specific to the individual circumstances of an incident. The committee examined the BOI reports from this perspective and found many instances where specific ToRs were added to the standard ones (i.e., existence of pre-incident indicators, security classification/placement, staff presence in the area, staff response to the incident). Specific ToRs included "Mental Health Review completed by the psychologist", "Inmates overall Mental Health Assessment and Treatment Plan", "Measures in place for prevention of drug incidents and review of BOIs for lessons learned", "Drug Interdiction Strategy and Search Plan compliance", and "Frequency and quality of the searching completed in #3D range", and "Management of Security Threat Groups (STGs) and Aboriginal affiliation with STGs." Our general finding is that CSC has done well in tailoring its terms of reference to the specific circumstances surrounding the incident, with the result that investigations are better focused to target the areas that are most relevant to the incident.

The second response to Recommendation #16 of the 4th IRC was that BOI reports would include a "context" section that would describe and analyse factors related to the environment and operations at the site. Although not prescriptive, the Board of Investigation Report guidelines set out a long list of factors that Board members may consider for inclusion in their reports, such as, environmental scan research information; population management/ demographics; issues from previous investigations; trends and analyses of incidents at the operational unit, regionally and across the country; relevant findings from other BOI reports or the OCI; inmate complaints and grievances prior to and at the time of the incident; and incidents involving use of force prior to the incident. Currently, the context section is only used in certain high-profile cases specifically asked for or approved by management. The committee examined BOI reports to determine whether there were cases that included a context section. We found several examples of BOIs that contained a context section. These sections covered a variety of topics related to the incident under investigation, including the opioid crisis and related research, inmate population and double-bunking, inmate culture and gang affiliation, STGs and Aboriginal affiliation with STGs, and the role of a community residential centre in the supervision of an inmate in the community and their relationship with CSC. In our view, the inclusion of the context section was very helpful in understanding the circumstances in which the incident occurred and has resulted in more comprehensive reports.

A context section be included in every Board of Investigation report.

The issue underlying Recommendation #16 of 4th IRC Report is captured in the following paragraph which preceded their recommendation.

Although we did not find, among the cases we reviewed, any that were as troubling as the Hines case, there were a few cases where there were multiple failures that contributed to the incident (which are described in the chapter on ToR #4 below), raising for this committee questions about the quality of the institutional performance more generally and its impact on the inmates in the institution at that time. The standards of practice for CSC are clearly set out in CD 001 which, in the context of its mandate and mission, provides a strong foundation of Values and Ethics (V&E) for the organization. To bolster V&E, there are multiple platforms for training and support at all levels of the organization (i.e., National Employee Orientation Program, "What if, then What …", Correctional Training Program for new recruits, Operational Senior Managers Training Program, Ethical Leadership Training Program, Ethics Training for CSC trainers, Organizational Integrity and Ethics session for EXCOM, and hoc V&E sessions to meet team specific needs/requests). A National Advisory Committee on Ethics chaired by the Commissioner has been established to ensure that V&E are embedded throughout the organization. In addition, CSC has launched an ethical risk assessment initiative to assist in identifying areas of ethical risk in the workplace and developing strategies to address them.

In the course our review, we became aware of a program of research conducted in the UK by a research team at Cambridge University headed by Professor Alison Liebling on the quality of prison life (Liebling, 2004; Liebling, 2014; https://www.prc.crim.cam.ac.uk/directory/research-themes/mqpl; Auty & Liebling, 2020). This research, Measuring the Quality of Prison Life (MQPL), focusses on dimensions that are central to the quality of prison experience (e.g., respect, relationships, trust, support, fairness, safety, organization, consistency, etc.), and incorporates surveys of inmates and staff. Although not within the parameters of the ToR, given its high relevance to the issue underlying the 4th IRC recommendation, and its potential value to CSC, we decided to examine this topic further. This is a complex line of research that has been ongoing for two decades. It has required significant effort and resources, but it has produced dividends. Applying the methodology in particular prisons has enabled senior managers to identify weaknesses and take corrective action to address them. In addition, the measures have been shown in the UK research to be related to higher well-being, lower suicide rates, institutional order, and reduced reoffending. MPQL has been formally adopted by the National Offender Management Service (NOMS).

From our review of the current research and performance measurement in CSC, we did not find anything comparable to MQPL. CSC has conducted research related to culture (i.e., the norms and expected behaviours of individuals in a particular group and setting) and how these are shaped in institutions, but this is not synonymous with the experience of the quality of life in an institution. From a performance measurement perspective, the committee was impressed with the Trends, Analyses and Performance (TAP) tool. Essentially, TAP provides a searchable database on key topics including population management, offender profiles, security, incidents and grievances. It provides institutional profiles as well as changes in key indicators over a rolling 13 months, and allows comparisons across institutions along these dimensions. The committee was struck by the wealth of information contained in this database, and the accessibility of this information through its searchable function. Indeed, it would be a valuable tool for Incident Investigations. For example, an analyst in IIB could use TAP to provide the BOI with a picture of the institution where the incident occurred as context for their investigation. Notwithstanding that TAP is a useful tool, and would be an excellent resource for BOIs, it does not include measures that are captured in the MQPL.

We were informed that an unsolicited proposal to conduct research on MQPL was submitted to CSC some years ago but it was not pursued at that time. We are not aware of the basis for this decision. Nevertheless, there is clearly merit in conducting this kind of research, and this was the consensus of those we interviewed on this subject. Indeed, research on the quality of prison life could assist CSC to better understand the conditions in institutions as experienced by inmates and staff; improve performance for the general well being of inmates and staff; and contribute to the reduction of incidents of non-natural deaths.

That Correctional Service Canada embark on a program of research on the Quality of Prison Life in Correctional Service Canada institutions along the lines of the research undertaken by Professor Alison Liebling and her associates.

That the Trends, Analyses and Performance database (TAP) be made available to the Board of Investigation teams to assist them in understanding the institutional context of the incident.

That CSC conduct a research study on a model that incorporates recent enhancements to CSC's policies and practices regarding engagement with families of offenders who die in custody, with a view to establishing best practice in this area.

This recommendation was made by the 4th IRC in order to further the work that CSC had undertaken in response to the recommendations of a report prepared by the OCI, titled, "In the Dark: An Investigation of Death in Custody Information Sharing and Disclosure Practices in Federal Corrections." The collection of initiatives designed to improve engagement with families in cases of a death in custody were listed in the 4th IRC report as follows:

• Develop and implement a facilitated disclosure process.
• Establish a guideline outlining procedures regarding notifications to the family in circumstances of serious medical emergencies.
• Establish CSC points of contact with families (i.e., Family Liaison Officers) from notification through to the completion of the investigative process.
• Provide suitable training for staff to assist them in communicating with families in these circumstances.
• Send a letter of condolence to the family.
• Prepare a guide for families to explain the policies and processes following a death in custody and key contacts and community services that may be helpful to them.
• Modify the approach to vetting and releasing information by establishing a dedicated team of Access to Information and Privacy experts to work closely with the family members and other partners and stakeholders to ensure that information is shared appropriately and consistently. (p. 61)

The CSC response to this 4th IRC recommendation indicated that the Research Branch was unable to find an external partner to undertake this research. Instead, they were conducting a review of practices in other jurisdictions and would prepare a summary of the results. When the committee interviewed staff in the Research Branch, we were informed that the review yielded very little documentation on practices in other jurisdictions and would not be particularly helpful going forward. We also learned that the main obstacle to finding an external research partner was that researchers expressed ethical concerns surrounding contact with families in the circumstances of trauma following a death. This concern is understandable considering the sensitive nature of conducting interviews in these conditions. We noted in our discussion that one way of addressing this issue is for the research team to work with a partner from the voluntary sector that has experience and expertise engaging in sensitive matters such as this, and one that has established trust with inmates' families. Canadian Families in Corrections Network (CFCN) is precisely such an organization, and could be engaged in the project to provide the initial contact with the families to gauge their interest and willingness to participate in the research.^{Footnote 1} The research staff agreed that this would be an option for overcoming the identified obstacle so that the research could be undertaken. Before undertaking an evaluation of the model, there would need to be a comprehensive document describing the components of the model, and the policies and protocols established to ensure that it is operational across the country. Based on interviews and review of CSC documentation, our conclusion is that such a document has not yet been produced.

The CSC policy for notifying Next of Kin (NoK) in cases of the death of an inmate is established in Commissioner's Directive (CD) 530 Death of an Inmate: Notifications and Funeral Arrangements. This CD sets out the responsibilities of the relevant parties and the procedures regarding notification and funeral arrangements. The Regional Administrator, Communications and Executive Services (RACES) is designated as the Family Liaison Coordinator (FLC), and in that capacity, is the key point of contact for sharing information with the family following a death in custody, after the initial notification has been done at the site level. The policy also provides for this responsibility to be delegated to the site personnel, i.e., the Family Liaison Officer (FLO), who would perform these duties under the oversight of the RACES. The procedures require prompt notification at the site level of the person's emergency contact or NoK by telephone where possible, and a letter of condolence to the inmate's emergency contact or NoK that includes the coordinates for the RACES. The guide, titled "Death of a person in the care and custody of CSC: A guide for family and friends", will be shared by the RACES (or designate) with the NoK. In addition, the RACES (or designate) will collect and share factual information with the NoK within the parameters of the Privacy Act and the Access to Information Act, unless it would interfere with an investigation that could lead to a criminal charge.

In our review of the BOI reports, we noted that the notification was most often done by the chaplain who contacted the NoK by telephone. The NoK Training Guide instructs the staff who are notifying the family to offer their contact information, and that of the FLO in the institution, within 24 hours. The Training Guide also specifies that the FLO will follow up with another call the day after the death (never the same day and within 48 hours of the death), and tell the families that they can contact them.

The committee conducted several interviews with families and staff involved in the notification process. This is a very unsettling time for families who are anxious to know what happened in the institution that led to the death of their family member. It is also a significant challenge for CSC to help families at this time, and making these notification calls to a family is a difficult job. The investigation has not been done, so it is difficult to provide to families what they most want, information on what happened. One chaplain we interviewed had to call several numbers to reach the family member, who when told of the death said nothing and hung up, and he was unable to reach her again. It is a difficult conversation, particularly when the staff member making the call has little or no information about the circumstances of the death.

From our review of the BOIs, we noted that there was considerable variation in the time it took to notify the family, i.e., from one hour to three days. Every effort must be made to do this as soon as possible, using all available contact information listed for the inmate. In the case of an imminent death, the urgency for contacting NoK is even greater so that they may have an opportunity to travel to the hospital to be with the inmate at the end of their life or to participate in end-of-life decisions. In our view, there is a policy gap regarding notification in cases of imminent death, and we are making a recommendation to address it (see below).

Regarding letters of condolence, we found from the review of BOIs and our interviews that a letter was usually sent out by the warden, and another by the RACES or delegate, in addition to the letter sent from the Incident Investigations Branch if an investigation was pending. What we learned from families is that they felt that receiving three letters from CSC officials was excessive, and led to confusion regarding who to contact for further information. We will be making a recommendation (below) that is intended to bring greater consistency and streamline the process of initial communication with the family.

The feedback that we received from families regarding the CSC developed Guide ("Death of a person in the care and custody of CSC: A guide for family and friends", 2017)^{Footnote 2} was very positive. The information is clearly set out for the reader; contact information is included; and Annex A contains a very helpful checklist to assist the family in following the steps in the process. The Guide is also useful for community agency staff who may be supporting the family following an inmate's death. Internally, the Guide is useful for CSC staff to know what steps are to be taken, how information is shared and who is responsible for liaison with the family.

The disclosure of information regarding a death is addressed in more than one area of CSC policy. As noted above, it is covered in CD 530, which stipulates "that the RACES (or designate) will collect and share factual information with the NoK within the parameters of the Privacy Act and the Access to Information Act, unless it would interfere with an investigation that could lead to a criminal charge. In addition, disclosure provisions are contained in the following CDs:

• CD 022 Media Relations states that "In all cases of an inmate's death, CSC will inform the Next of Kin that a news release will be issued within 24 to 48 hours of the inmate's death. If more time is required to inform families, adjustments will be made for the release of the information to the media." (Para. 33)
• CD 041 Incident Investigations specifies that "The Director General, Incident Investigations Branch, will, in the case of the non-natural death of an inmate, notify the Next of Kin, or other designated person, that an investigation has been convened and that they may request a copy of the report from the Access to Information and Privacy Division of CSC." (Para. 14, d)
• CD 784 Victim Engagement states that "victims of inmates who are notified, normally within five days following the death and only after Next of Kin has been notified" and if victim is also the Next of Kin, Victim Services or institutional staff will notify them of the death. (Annex C)

In short, there are disclosure policies in place and, with the caveat of the recommendations that follow, they are generally suitable to support disclosure of information to NoK. In practice, we found some variations across regions/sites, and some features of the facilitated disclosure process that are still being refined. For example, the role of the RACES team is not always clear and regional variations exist regarding sharing information in writing with NoK. Most sites don't disclose the circumstances of death or provide referrals to address the trauma for the family, but rather attend to issues such as body transportation and personal effects. The Access to Information and Privacy team at CSC is working closely with the Office of the Privacy Commissioner (OPC), and provides the OPC with a checklist reflecting the type of information that CSC intends to share proactively with the family in a timely fashion.

Regarding possible family input in the investigation, an invitation to do so is provided in the letter from the IIB. Although it is rare that families wish to be involve in the investigation, they can contact the IIB to give input into the Convening Order and for investigative updates. The BOI may interview family members if the Board's inquiry indicates that they may have information relevant to the investigation, although there is no specific mention of NoK in the IIB training manual. The DG, IIB is open to meeting with families to review the findings of the investigation, and has occasionally done so.

From our review, one aspect of the model that needs more attention is training. As the 4th IRC noted, engagement with families in these circumstances is a very complex matter. While the event itself is traumatic, there may be layers of trauma based on events from the past that intersect with the immediate trauma. Staff who engage with families following the death of an inmate, from the initial notification to ongoing information sharing throughout the process, require skills to engage with sensitivity and must have an awareness of the grief and trauma that families are experiencing. We did not find evidence that the staff who are called on to perform these duties have received this training. We see a continuing need to raise awareness of CSC's "Death of a Person in the Care and Custody of CSC" Guide (2017) and their "Notification of Death" Training Package produced in the last few years to existing and new staff, and those involved in the disclosure process. In most of the BOIs that we reviewed, the chaplain of an institution was tasked with notifying NOK of the death but in our interviews, some chaplains were not aware of the Notification Training Package.

Returning again to the 4th IRC report, the impetus for the OCI study on information sharing and disclosure practices was that some families of offenders were contacting the OCI seeking advice and assistance to access information from CSC following the death of a family member. These families reported that they were having difficulty getting information, particularly regarding the events that led to the death and the circumstances surrounding it. We followed up with the OCI to see whether families were still reaching out to their office for support. They noted a "steep decline" in communications from families following CSC's implementation of the recommendations in their report (In the Dark: An Investigation of Death in Custody Information Sharing and Disclosure Practices in Federal Corrections). It was their understanding that CSC's initiatives have led to significant improvements in the responsiveness of CSC staff to families' requests and, although they have not assessed the implementation of the model, the sense is that the communication lines with families are now operating reasonably well.

That Correctional Service Canada contact Next of Kin in the event of an imminent death in hospital in order to provide the opportunity for the family to travel to be with the inmate at the end of their life or to participate in end-of-life decisions.

That there are only two possible letters sent to families. That the Warden, notify the family, offer condolences and indicate that a letter will be sent by the Family Liaison Coordinator who will be their main point of contact for information about the circumstances of death, making funeral arrangements and further disclosure. If there is a subsequent investigation, the Incident Investigations Branch sends a second letter outlining that the family may participate in the Board of Investigation and can contact them for investigation information.

Chapter 2

• Analyzes the independence and impartiality of CSC's investigative process.

Given the scope of this ToR, we have divided this chapter into two parts. The first deals with independence. The second deals with impartiality. We appreciate that they are very much interconnected.

Independence, in the context of investigations, generally means the degree of separation between the investigators and whomever they are investigating. The more independent the investigative agency is, or is perceived to be, the more credible the investigation will likely be, not only to the parties involved, but also to anyone else who may have an interest.

The less independent the investigators are, or are perceived to be, the more potential for allegations that the investigation was tainted and any findings flawed. It is an easy allegation to make and can be difficult to disprove.

In the public sector, governments across the world have created oversight offices, such as statutory Ombudsman, Auditor and Inspectors General, and specialized Commissioners. They operate at arm's length from those they oversee. In the strongest models, these watchdogs report directly to the Legislature, as opposed to the government. The investigators who work for these agencies therefore have a robust degree of independence (and by extension, credibility) not available to investigative agencies that report internally within the organization they are investigating.

There is a growing trend in Canada, as well as the UK, the US, Australia and New Zealand to create oversight frameworks that have a significant degree of independence from whomever it is they are investigating.

This trend is particularly apparent in the investigation of deaths that occur in the custody of the state. Virtually every Canadian province has created an independent civilian or civilian-led agency to investigate deaths and serious injuries involving police. The focus of their investigation is whether or not there are grounds to believe that an involved officer committed a criminal offence in relation to the death. The involved police service plays no role in these investigations.

The federal government and several provinces have also created independent civilian agencies that deal with complaints against police that do not involve a death or serious injury. These agencies generally have a mandate to identify and investigate alleged or apparent systemic issues relating to policing in general.^{Footnote 3}

Provincial Ombudsman, who report to Legislatures, oversee provincial corrections systems. They deal with individual complaints and conduct investigations into systemic issues, such as segregation and use of force by Corrections Officers. The Office of the Correctional Investigator (OCI), which reports to the Minister of Public Safety, performs a similar function at the federal level.

Of course, different types of investigation require different degrees of independence in order to be credible. The more serious the investigation, the greater need for real and perceived independence. The death of an individual while in the custody of the state is, by definition, an extremely serious matter.

As mentioned in the introduction to this report, one of our terms of reference was to examine international standards and practices in respect to their investigative processes in general, and specifically in relation to deaths. We were asked to determine if there are any lessons to be learned about how to conduct investigations, both generally and in relation to deaths, from what other jurisdictions do.^{Footnote 4}

Here is a brief summary of how the agencies we contacted operate, with some background about their mandate and what degree of independence they possess. All of these agencies have a very similar mandate to IIB when a death in custody occurs. The focus of their investigations is to learn lessons from what happened and, where necessary, recommend changes to minimize the chances of something similar reoccurring. It is up to the prison system whether and precisely how these recommendations are implemented. None conduct criminal or disciplinary investigations.

The lessons that CSC could learn from some of the investigative practices of these agencies are incorporated throughout this chapter. Additionally, these practices are closely intertwined with investigative independence and impartiality, which is why we have combined Terms of Reference 2 and 6 (international standards and practices) in this chapter, as noted in the Introduction.

Prisons and Probation Ombudsman (PPO) England and Wales

PPO investigates all natural and non-natural deaths that occur in England and Wales in a prison, youth detention centre, immigration detention facility, court cells or being transported to/from any of the above. Their mandate also includes halfway houses and some children secure units.^{Footnote 5} They also have a similar role to the OCI, including investigating complaints from inmates.^{Footnote 6}

The PPO is 26 years old. It was given the authority to investigate deaths 16 years ago. It has investigated hundreds of deaths since then.

In 2018/19 they launched investigations into 334 deaths, of which 180 were due to natural causes. The remainder included self inflicted deaths (91) and homicides (4). In total, the PPO made 723 recommendations in death in custody cases, of which 138 related to healthcare, 117 related to emergency response and 80 related to suicide/self harm prevention

The purpose of their death investigations is to "understand what happened and identify how the organization whose actions we oversee can improve their work in the future." The Majesty's Prison and Probation Service (HMPPS - the UK equivalent of CSC) does not conduct an investigation into the death, other than into potential disciplinary issues.^{Footnote 7}

The PPO reports to the Minister of Justice, who also sets the budget. In an interview with an IRC member, the Deputy PPO stated that she feels that this compromises their independence. The optics are not good. They would prefer to report to a Parliamentary Committee.

She also commented on how important real and perceived independence is to their work. Firstly, from a legal perspective, by virtue of Article 2 of the European Human Rights Act when someone dies in the custody of the state, there must be an independent investigation. An internal investigation by HMPPS would not meet this requirement, in her view.

Secondly, independence adds an extra level of assurance to stakeholders and it is absolutely critical to their credibility. Even with their arm's length status, they still occasionally have trouble convincing families they are not apologists for HMPPS.

She advised that there has never been any interference in any PPO investigation to her knowledge.

The Prisoner Ombudsman for Northern Ireland (PONI) was created in 2005. It has jurisdiction over four prisons in Northern Ireland. There are roughly 1,400 prisoners in custody at any one time, with about 50% on remand.

The Prisoner Ombudsman him or herself is an Independent Public Appointment who is appointed by the Minister of Justice. He or she is completely independent of the Northern Ireland Prison Service (NIPS), reporting to the Northern Ireland Assembly via the Minister.

PONI has investigated natural and non-natural deaths in custody since late 2005. It has completed investigations into 33 non-natural deaths in custody since then, all of which were self-inflicted. It also investigates incidents non-fatal serious self-harm and the deaths of people which occur within 14 days of their release from custody.

The Office has 14 staff, three of whom are involved in conducting death investigations, and an annual budget of approximately $1.1 million CAD

Chief Inspector of Prisons of New Zealand: Office of the Inspectorate (OOTI)

The OOTI oversees the work of the Department of Corrections, which includes the prison network and Community Corrections (the equivalent of the probation service). The Chief Inspector of Prisons is in charge of the office. It deals with complaints, conducts inspections and investigates all deaths in custody, including natural ones. In effect, it functions as a hybrid Inspectorate, last line internal complaint handler and an investigator of serious incidents, such as deaths in custody.

OOTI can also conduct systemic investigations on its own motion, based on issues that may not involve a complaint or an incident. In August 2020, it released its thematic investigation in to the treatment of older prisoners.

OOTI oversees New Zealand's 18 prisons, with a population of approximately 9,200 inmates.

OOTI is based in Wellington, with staff in Christchurch and Auckland. Its head office is situated in the Department of Corrections National Office in Wellington.

In regards to independence, the OOTI is a part of the Department of Corrections. The Chief Inspector reports outside of the chain of command directly to the Chief Executive, who is the most senior civil servant in the Department. She does not report to any senior Corrections executive, including the National Commissioner of Corrections, who is the equivalent of the CSC Commissioner.

She meets with the Minister of Corrections regularly, to discuss issues and challenges.

In respect of some other indicia of independence, she selects her own staff, who may come from within or without the Department of Corrections. She has her own dedicated legal counsel, which she describes as being very important, particularly in providing a degree of separation from the Department itself.

In 2019/20 OOTI conducted 27 death-in-custody investigations. eight were suicides, one was a homicide and the remainder were natural deaths.

In respect of what it can choose to investigate, OOTI has complete latitude to determine what issues will be investigated. While it does not necessarily solicit input from any other quarter, including families or the Department, if anyone made representations to them, they would assess and decide whether to pursue them, on their merits. OOTI has the final call on what is and what is not investigated when a death occurs.

OOTI also has the latitude to unilaterally amend the focus of an investigation as it progresses, if circumstances warrant.

Their recommendations are provided to the National Commissioner (and the Deputy Chief Executive of Health, if relevant). OOTI requests a response within five weeks. We are advised that 97% of OOTIs recommendations are accepted.

There is no Federal prison system in Australia. Each State is responsible for all prisoners within its jurisdiction. Corrections Victoria (CV) has approximately 7000 inmates in custody.

The Justice and Assurance Review Office (JARO) conduct an investigation when a non-natural death in custody occurs. JARO is a unit of the Department of Justice and Community Safety (DJCS), as is CV. However, it is not in the same chain of command as CV. JARO reports directly to the Deputy Secretary (Regulation, Legal and Integrity) and then to the Secretary, who is the most senior public servant at DJCS. CV reports to a different Deputy Secretary within Justice.

As well as inmates, JARO also oversees the Youth Justice system and people under both adult and youth supervision in the community

JARO investigates all non-natural deaths. The focus of the investigation is on potential systemic issues - with a view to recommending improvements system wide, if required. It investigated eight such deaths between November 2018 and October 2019

JARO describes itself as the 'third line of defence' in relation to the investigation/review of non-natural inmate deaths. Normally the institution where a death occurs will conduct an internal investigation, which will then be reviewed by CV senior management.

In respect of potential Terms of Reference, JARO has complete autonomy over what it chooses to investigate. There is no external influence on how the issues to be investigated are framed, including from CV, though there may be occasions they consult with CV as potential issues are being identified in the review planning process.

In many cases, there will be a Coroners Inquest after an inmate death. JARO provides its report to the Coroner. We are advised that the Coroner often refers to JARO recommendations in their findings, including whether CV has accepted them.^{Footnote 8}

In respect of JARO's relationship with CV, it is, in general, collaborative. Cooperation overall is very good, especially at senior levels. As is the case with every other oversight agency we are aware of, they advised that there is some tension between the overseen and the overseers but - in our view - that is only natural. It would, again in our view, be a concern if there were not.

The State of New South Wales (NSW), Australia

We spoke to a senior official at the NSW Ombudsman's Office who deals with issues relating to corrections. The NSW Ombudsman oversees the corrections system in the State. She is very knowledgeable about how the death in custody investigation system works. We also reviewed Coroner's Reports and other information available on the Corrective Services NSW (CSNSW) and other websites.

There are roughly 14,000 inmates in the NSW system. In 2019 there were 44 deaths in custody. When a death occurs, Corrective Service NSW Investigations Branch and the NSW Police Force conduct investigations. There is also a mandatory Coroner's Inquest. The Coronial process is apparently very robust.

There are other mechanisms that may have some involvement in the investigation of a death in custody. They include the Inspectorate of Custodial Services and the CSNSW Management of Deaths in Custody Committee, which has a mandate to ensure recommendations from Coroners Inquest are implemented.^{Footnote 9} The Committee is chaired by a CSNSW Assistant Commissioner. Members include senior CSNSW officials, the Chief Executive of the NSW Justice Heath and Forensic Mental Health Network, the NSW Police Force and the NSW State Coroner's Court.

There are a number of factors that indicate how much autonomy an investigative agency actually has. They are listed below, followed by our assessment of whether or to what extent IIB meets each criterion, along with findings and recommendations where appropriate.^{Footnote 10}

• To whom does the agency report?
• Does it get to decide what it investigates?
• Does it have the authority to get the evidence it needs?
• How is it staffed?
• How does it deal with possible conflicts of interest?
• How transparent is it?
• How is it funded?
• Where does it get its legal advice from?

We have already discussed the various models, reporting to a legislature, directly to a Minister, internally to various levels of the organization and so on. The PPO and PONI are examples of arm's length agencies that are completely independent of the corrections system itself.

In cases where an investigative agency is embedded within the organization it investigates, to whom does the head of the agency report? For example, the Department of National Defence / Canadian Armed Forces (DND/CAF) Ombudsman reports directly to the Minister, outside of the military and civilian chain of commend. The OCI, although not part of CSC, also reports to the Minister.

Currently, CSC's Director General of IIB reports to the Senior Deputy Commissioner (SDC), not directly to the Commissioner. The SDC has significant responsibilities in a death in custody investigation, including keeping the CSC Executive Committee informed of ongoing Tier 1 investigations and approving any extensions to 'established timeframes'.

We recommend below that the DG report directly to the Commissioner.

Prior to explaining why, we want to emphasise that we heard no complaints from anyone at CSC that this reporting structure was an issue. There is no evidence we are aware of that the death investigation process is being adversely impacted by the fact that the DG reports to the SDC. Further, this recommendation is in no way a criticism of the SDC or his commitment to thorough and objective death investigations. In our meeting with the SDC and the Commissioner at the beginning of our review, they both made it very clear the very high level of importance they attached to how CSC investigates deaths.

We also recognize that the Commissioner has a multitude of responsibilities that cannot possibly be dealt with efficiently and effectively without delegating to her senior management team. Adding another direct report obviously increases an already very significant burden. We understand that.

But that, in our view, is not the point. As mentioned throughout our report, perceptions of independence when investigating deaths in CSC custody are absolutely crucial, both internally and externally. Maximizing confidence in the independence of the process is particularly vital when the investigation is being conducted internally. Reporting directly to the highest level in the organization can help build that confidence. The OOTO in New Zealand is an example. While her office is embedded in the Department of Corrections, she reports outside the Correctional system chain of command to the equivalent of a Deputy Minister and has regular access to the Minister.

We are recommending that the DG report directly to the Commissioner for that reason. It sends a very strong message to everyone, CSC, families, the public at large, that CSC has no greater responsibility than to do everything it possibly can to safeguard the lives of everyone within its custody. When someone does die a non-natural death in custody, CSC will do everything within its power, at the highest level, to find out what happened and, where necessary, take steps to minimize the possibility of a future death in similar circumstances. Further, the Commissioner has determined that this is such an important priority for CSC, that the person who is in charge of conducting such investigations will report directly to him or her.

We anticipate that there will likely be pushback to this recommendation. Critics may characterize it as cosmetic, operationally unnecessary, that it creates an unwelcome precedent and/or will be unduly burdensome.

We would respond that, on occasion, perception trumps all of those things. This is such an instance.

Having the DG IIB report directly to the Commissioner will enhance the credibility of the process. It will demonstrate the importance CSC attaches to the work that IIB does, in particular when investigating non-natural deaths in CSC custody.

The Director General of the Incident Investigation Branch should report directly to the Commissioner of Correctional Service Canada.

Can an investigative agency decide what it will investigate? If not, who decides? Who, a cynic may say, pulls the strings? Once it has begun, can it go where the evidence leads it?

Each BOI has a Convening Order (CO) that includes Investigative Areas / Terms of Reference (ToRs).^{Footnote 11} The ToRs are a list of issues to be analysed by the BOI.

We reviewed the CO and ToRs in each of these that came within our remit.

The purpose of each investigation was boiler-plated in the Convening Order, to make findings and recommendations 'which may contribute to the effective resolution and/or prevention of similar situations or occurrences in the future'.

The ToRs are fairly generic, varying little from incident to incident, but were appropriate and sufficiently broad. Depending on the apparent nature of the death, BOIs were directed to investigate areas that were clearly relevant, including, generally:

• any pre-incident indicators;
• the security classification of the inmate;
• care, treatment and monitoring;
• staff presence; and
• staff response.

As we noted in the chapter above on ToR 1, we saw examples in more recent BOIs of additional ToRs that were tailored to the individual circumstances of the incident which, we concluded, were helpful in focusing the investigation without restricting it. We found no instances where areas that obviously should be investigated were ignored. In addition, there was a degree of latitude built into the way some of the ToRs were written, for example the use of the phrase 'but not limited to' when discussing possible pre-incident indictors in the Saskatchewan Penitentiary BOI that is discussed in some depth later in this segment. The investigative areas set out in the CO into a homicide and serious injuries at Stony Mountain included 'the staff response and management of the incidents, including but not limited to:' …and then went on to list certain areas.

There is an obvious problem, however. BOI ToRs are created and approved by CSC. The Commissioner personally signs off on them in non-natural death in custody investigations. That leaves CSC open to criticism that it is those who are being investigated that are setting the parameters of an investigation. Limits could, in theory, be placed on a BOI from exploring investigative avenues that should, objectively, be explored.

In order to counter that criticism, we recommend building an additional safeguard into the existing system at this point in the BOI process. BOIs should have the authority to expand an investigation as it is underway, without obtaining prior approval.

Our rationale is this:

• Permitting a BOI to unilaterally add an area or areas of investigation without seeking prior approval from CSC greatly enhances the real and perceived degree of the autonomy of the BOI. It sends a very clear message that a BOI calls the investigative shots. The BOI has the blessing of CSC to go wherever the evidence leads it, unfettered.
• Investigations are dynamic - and the focus may change as it progresses. Key issues that may not have been apparent when ToRs were originally being formulated may emerge as the investigation is ongoing. Having the ability to extend the scope, pivot or refocus mid investigation, without having to seek authority from the agency they are investigating, not only adds to perceptions of independence, it also adds to investigative efficiency.
• Other agencies do it. The PPO advised that it is up to the investigator and her/his manager to determine what issues are apparent and appropriate investigative avenues. That may be refocused as the investigation progresses.^{Footnote 12} OOTO decides what it investigates and can pivot to where the evidence takes it without seeking approval from anyone, as can PONI.

We believe that adding a blanket 'but not limited to' in the investigative areas segment of all non-natural death in custody BOIs, will pay dividends. Further, an additional ToR should be added to these BOIs, along these lines:

The BOI may extend the investigation to any other issue that the BOI deems relevant to the investigation, as the investigation is in progress.

We anticipate that this authority will be used rarely, if at all. However, the fact that it exists creates a clear and transparent additional degree of separation between those who are both setting the limits of an investigation and also being investigated - and those conducting the investigation. This is particularly vital when an agency is investigating the organization that it is a part of, as is the case with IIB.

It may be that a BOI already has that authority in practice - in fact several individuals who are closely involved in the BOI process told us that they thought it had. If that is the case, why not set it out clearly in writing in every CO?

We anticipate that there may be concerns raised about a rogue BOI going off on irrelevant tangents without any kind of supervision. This, in our view, is highly unlikely. Regardless, the possibility does not outweigh the benefit provided by a BOI being able to pursue whatever issues it deems fit.

That does not mean that a BOI who chose to expand an investigation would be unaccountable. In the event that it does decide to investigate additional issues, it should of course advise the DG IIB that it intends to do so. The rationale for all such decisions must be documented fully, including in its final report. Timelines and resources should be amended accordingly.

As currently drafted, ToRs are reasonably broad.

Issues may arise during an investigation that may not be fully covered within the scope of the ToRs. In those likely very rare situations, BOIs should have the authority to unilaterally decide to follow up on those issues, as it deems fit.

An additional ToR should be added to each non-natural death in custody BOI Convening Order, along these lines:

The BOI may extend the investigation to any other issue that the BOI deems relevant to the investigation, as the investigation is in progress.

There must be accountability and transparency if a BOI avails itself of this authority. It should provide its reasons for doing so in its final report.

A Board of Investigation should have the authority to develop additional Terms of Reference as an investigation is underway, without obtaining prior approval from Correctional Service Canada. The rationale for any such decision must be documented.

BOI ToRs include a specific segment that sets out that the BOI can 'adopt such procedures and methods as may be deemed necessary for the proper conduct of this investigation.' The ToRs provide that a BOI has the authority to search any CSC building and seize items, has 'complete access' to CSC personnel and can 'communicate with any person outside of CSC', albeit at the discretion of the Chair. CSC employees are obliged to cooperate with BOI investigations and commit an infraction if they do not do so.^{Footnote 13}

Additionally, non-natural deaths in custody BOIs are afforded strong powers under the Inquiries Act.

We did not find evidence of a BOI being unable to gather the evidence needed to conduct its work. Nor did we find evidence of any significant obstruction of any BOI investigation that we reviewed. We did hear of a few cooperation issues but none appeared to have prevented a BOI obtaining the evidence that it required to conduct a thorough investigation.

In summary, BOIs have robust powers at their disposal to gather evidence. As is the case with many similar investigative agencies, the fact that they have these powers means that they rarely have to use them.

How an investigative agency is staffed may give some indication of how independent it is. Does the investigative agency have the ability to select its own staff? Does it have control over the hiring process? Can it select its own staff without input from the organization it has a mandate to investigate?

Another related question is what degree of reciprocity is there between the investigative agency and the entity it investigates? What is the process for staff transferring between the investigator and investigated sides of the house? Is it a temporary move or a permanent one?

There may be barriers. For example, the Ombudsman of Ontario is an officer of the Ontario Legislature. The Office oversees the Ontario Public Service (OPS). However, Ontario Ombudsman employees are not OPS members. They are not eligible to apply for positions in the OPS that are being filled through closed competitions. Nor are OPS staff given any preference for vacant Ontario Ombudsman competitions. The rationale is to prevent the appearance of potential conflict of interest. It provides an additional degree of separation between the investigated and the investigators.

IIB staff members are employees of CSC. The DG, Directors, Team Leaders, Analysts and support staff are permanent IIB staff. National Investigators (NIs) are CSC staff who are assigned to IIB for a two to three-year period, though that can be extended. NIs are normally the Chair of the Board. Their duties include planning and co-ordinating the investigation, supervising the work of Board Members and 'delivering a quality product' on time.^{Footnote 14}

There is also a complement of 'casual NIs.' These are former CSC Executives who are under casual contract with CSC for a specific BOI. They are retained for a maximum of 90 days a year.

Board Members are CSC staff who are released from their normal duties to participate in the evidence gathering phase of the BOI. They also participate in drafting the report.

IIB selects its own staff. IIB advise that there is significant competition to join, at any level, including participating as ad hoc Board Members.

Tier 1 BOIs must include a Community Board Member (CBM) who cannot be, or have been, a CSC employee. IIB is responsible for identifying and retaining CBMs.

We found no evidence that IIB staff were anything other than fully dedicated to their function, or that a tour in IIB was seen as a stepping stone to the next promotion. In fact, we were impressed by the clear commitment that the IIB staff we had contact with demonstrated. However, that does not deal with the perception that IIB, because it is an internal body, may be subject to real or perceived pressure from others within the organization.

We were advised, and we agree, that there are considerable advantages to having individuals who are intimately familiar with CSC conduct these investigations. They know the system, the culture, the dynamics, the acronyms, the policies and procedures and so on. They understand issues such as the principles of assessing risk levels and managing violent offenders. A very experienced CSC staff member advised that, in their view, a BOI chair 'could not function' without knowledge of the correction system and that 'outsiders do not know what they do not know'.

While these arguments have merit, they are not in our view compelling. Knowledge and understanding of the organization you are investigating is very important. But that does not necessarily mean that you must necessarily have walked a mile in their shoes before you can investigate them. You don't have to be, or have been, a cop to investigate cops, or a lawyer to investigate lawyers. The Military Ombudsman has relatively few ex-military personnel and no serving ones. The majority of full-time Ontario Special Investigations Unit (SIU) investigators, which conducts criminal investigations of police officers involved in a death or serious injury, have no police experience.

While we do not propose to make a finding or recommendation on staffing, we note that the broader the non-CSC base from which BOI members are drawn, the greater will be the degree of separation between the investigated and investigators. We discuss this further below, in the segment on Community Board Members.

A related issue is whether an organization can retain professional services without having to seek approval from the organization it is investigating. IIB has to go through the normal CSC staffing process and CSC Contract Review Board for external hiring, at least in most cases.

One measure of independence is whether the agency has a rigorous process for determining if anyone involved in an investigation has a potential conflict of interest.

The conflict of interest could be real, i.e. pretty obvious to anyone. Don't sit on a BOI if a family member is the Warden of the Institution under investigation, or if you have previously worked fairly closely with a Correctional Officer who appears to be involved in the incident.

Or it could be a perceived conflict. Perceived conflict of interest can be more nuanced and difficult to define. Any perception must be reasonable, i.e. that a neutral person would reasonably conclude that it could be a conflict.

IIB has a conflict of interest process for potential BOI members. Generally, a CSC member appointed to a board will not be from the region where the incident occurred. There is a duty and expectation that a potential BOI member will advise of any apparent conflict at the earliest possible point. This does occur now and again, we are advised that mostly because a BOI candidate has had previous dealings with an employee or inmate involved in an incident.

BOI members are canvassed for potential conflicts. We have been provided with a copy of the IIB Conflict of Interest Risk Assessment form used for National Investigators in any BOI and a similar form for Board Members who may be appointed to Tier 2 investigations.^{Footnote 15} Both appear reasonably comprehensive and targeted at identifying potential conflicts before a potential member is appointed to a Board. They include questions on whether the candidate has previously worked at the facility where the investigation will take place or had any previous involvement with anyone who is likely to be interviewed in the investigation, including CSC colleagues.

CBMs are required to certify to IIB that they have no conflict of interest, including that they are '…not connected in any way to potential witnesses, anyone linked to the incident or other members of the Board of Investigation'^{Footnote 16}

We have not seen forms of this type in other investigative agencies that conduct oversight investigations, though of course that does not mean they do not exist. Most often, at least in our experience, the onus is on the employee to identify any possible conflict. We applaud IIB for developing this proactive process.

Another indication of independence is transparency. What information does the agency make public? Does it issue annual and/or special reports? Are its investigation reports released publicly, as is the case with Ontario SIU?

In the case of internal investigative agencies, is any policy that impacts its work or any general direction given to it made public, such as the Ministerial Directives that govern the operations of the DND/CAF Ombudsman?

In IIB's case, Commissioner's Directive 41 (CD 041) sets out what IIB does and how it does it. It is a public document.

We discuss the public reporting of death in custody BOI investigations in our response to ToR 3, including what agencies in other jurisdictions do.

How an investigative is funded is a key component of its independence. Who controls the purse strings? What control does the agency have over its budget? Is it provided with sufficient funding in order to conduct its investigations in a thorough and timely fashion? Can it spend that money how its sees fit? Is that funding stable?

IIB has a budget of approximately $5 million. We heard no issues relating to inadequate funding, though as discussed later, there were issues raised about workload, which related more to positions not being filled rather than lack of funds to fill them. Nor did we hear any issues about how that money is spent. As we note in the segment on impartiality, IIB investigations are reasonably well resourced, as they should be, given the seriousness and importance of the work they undertake.

Does the agency have in house counsel? Can the agency, on its own volition, access external legal advice for discrete issues?

The reason this is potentially important is that there may, on occasion, be conflicts that impact independence if the agency shares counsel with or otherwise receives legal advice from the organization it is investigating.

IIB does not have in house legal counsel. It receives legal advice from the CSC Legal Services Unit.

Based on the criteria set out above, the IIB BOI system as it stands is clearly not independent, both in reality and in perception. Indeed, no-one within or without of CSC claims that it is. It is not arm's length, does not have its own legislation, nor does it report outside of the CSC chain of command. It shares legal counsel with those it investigates. It does not have the same degree of independence as the PPO, PONI, JARO or the OOTI.

That said, and as we will discuss in the impartiality segment of this chapter, not being independent does not necessarily mean that IIB cannot conduct reasonably fair and thorough investigations.

We make the following observation on several occasions in this report, including in the 'Important Caveat' segment of the 'What we did and who we spoke to' section of the report. It is particularly important that we repeat it here, in the context of this particular issue.

We were not given sufficient time to examine the entire investigative product for each BOI that we were tasked to review.^{Footnote 17} Nor were we given the time to conduct interviews if we deemed them necessary, in order to determine whether there was any evidence of efforts to compromise or unduly influence the independence of the process in a given BOI.

However, based on our review of the 25 BOI reports and the other investigative steps we took, including numerous interviews of individuals involved in the process and speaking to the OCI, we found no compelling evidence of systemic interference from internal or external parties in any of the BOI investigations we reviewed. There was no obvious evidence of any attempt to fetter or direct any of the BOIs we reviewed, as the BOI investigation process was underway. Nor were we provided with any such evidence, other than that referred to in the OCI segment on the Saskatchewan Penitentiary investigation in its 2017/18 Annual Report.

That said, we were told of an instance where a BOI chair had allegedly told another member that their input 'was not helpful as it may be damaging to the reputation of CSC'. However, this related to a BOI that occurred at some point prior to the period that we are tasked to review, so we did not explore this any further. We were advised that the member ignored the comment and that the chair no longer works at CSC. Another CSC staff member told us that, on one occasion, they were told that a BOI 'overstepped its mandate'. However, that staff member also advised that this was very much an anomaly and did not stop that BOI from doing its work as it saw fit.

IIB staff and other individuals involved in the process that we spoke to told us that they have never been subject to undue pressure in the course of their work, including in instances 'where a BOI finding may not be popular at a senior level.' Nor had they heard of anyone involved in a BOI who had been. Staff also told us that they felt comfortable 'speaking truth to power' and had no pushback for doing so. Additionally, we were advised that IIB has been very supportive of BOI chairs and members.

We understand that there are sometimes requests from entities within CSC, including at the local level, to amend BOI findings or recommendations. It has not been alleged by anyone, nor have we been provided with any evidence, that these requests are made in bad faith, or that inappropriate pressure is exerted on the BOI members. On the contrary, we are advised, and agree, that it is a normal part of any administrative investigations process that those who may be impacted by recommendations are given an opportunity to comment on them, prior to them being finalized. Not to allow this process would be procedurally unfair. BOIs should, and do, consider that feedback as it often provides additional important information.

However, when the finding/recommendation is, in the Board's opinion, an accurate representation of the facts and the issue, the BOI should stand firm in its conclusions. Any input from any CSC entity that may be interpreted by a reasonable person as intended to persuade a BOI to amend its findings or recommendations should be fully documented. That documentation should be included in the entire BOI review process, up to the National Investigations Meeting's (NIM's) process.

Overall, we found that the BOIs were conducted without any obvious signs of fear or favour, and that key issues were generally identified and, with some exceptions that are dealt with elsewhere in this report, were appropriately followed up.

In the interests of transparency and sound investigative practice, any request to amend a BOI finding or recommendation at the local, regional or national level should be fully documented by both the requestor and the BOI.

All communications between any Correctional Service Canada party and Board of Investigation members concerning findings and/or recommendations that the Board of Investigation has made or is considering making should be fully documented. If changes are made to any finding or recommendation, the process by which that occurred and the rationale for any changes should be included as an addendum to the Board of Investigation report.

In its 2017/18 Annual Report, the OCI had a segment entitled 'Special Focus: Investigation into the Riot at Saskatchewan Penitentiary'. It was a detailed account of its investigation into what it describes as 'a major riot' and what CSC describes as 'a major disturbance' at the Saskatchewan Penitentiary on December 14, 2016. During the event, which lasted for several hours, an inmate was murdered and two others were seriously injured by fellow inmates. There was approximately $3.5 million worth of damage. Corrections Officers discharged six shotgun rounds, injuring six inmates.

CSC convened a BOI. The OCI also conducted an investigation.

In its Annual Report, the OCI was highly critical of the BOI investigation. It concluded that the BOI had ignored key underlying issues, including what the OCI believed to be a connection between the incident and the introduction of a national CSC programme that impacted food supply at the Institution and had been poorly received by inmates. It disagreed with CSC's assertion that the incident was spontaneous. Most seriously, from our perspective, the OCI found that '(t)he means, manner and method by which the Board of Investigation approached the Sask. Pen riot was not transparent or credible.'

OCI also opined that a public Case Summary released by CSC in March 2018 did not match findings made by the BOI and was, in that respect, misleading.

The OCI made five findings that went beyond the investigation of the incident itself. The findings were a very serious criticism about the BOI system in general, alleging that. '…..the omissions of this particular Board are reflective of the inherent limitations and deficiencies of the NBOI process:

The findings were as follows:

1. Findings, lessons learned and recommendations from National Boards of Investigation rarely match the seriousness of the incidents under review: major disturbances, assaults, riots, serious bodily injury and deaths in custody.
2. The National Board of Investigation process is not free, impartial or independent from the Correctional Service of Canada, in form, function or appearance.
3. Investigative standards, credibility, rigour, integrity, thoroughness, quality, are not consistently met from one Board of Investigation to the next.
4. There is no requirement for Board of Investigation reports to be disclosed publicly or even distributed internally.
5. The BOI focus on policy and procedural compliance often fails to address the underlying cause(s) of recurrent incidents, impeding learning and limiting improvement.

The OCI made a recommendation to the Minister, based on these findings, as follows:

10. I recommend that the Minister of Public Safety conduct an independent review of the National Board of Investigation Section 19 process to enhance transparency, credibility, integrity and accountability of investigations convened and conducted by the Correctional Service of Canada. This review would consider an option for the Minister to authorise appointing an external and independent investigator into major disturbances (riots) resulting in injury or death, suicides in segregation and use of force interventions leading to serious bodily injury or death.

We carefully considered the OCI's Recommendation 10 and the findings, which of course involve both independence and impartiality.

We asked OCI for any additional evidence they had to support the contention that the shortcomings that they had identified in the Saskatchewan Penitentiary case were reflected in other death in custody BOIs. We were advised that the report spoke for itself.

When we interviewed IIB staff, it was clear that they felt that OCI's criticisms of the BOI were unfair. They argued that OCI had an agenda that focused on removing CSC from the death investigation process, at least in some cases. They defended themselves against the five findings, arguing that IIB conducts impartial, evidence-based investigations, that it focuses on key issues (though there was acknowledgement that that may not have always been the case in the past) and that they strongly disagree that BOIs fail to address underlying causes. They also strenuously denied that pressure had been brought to bear by CSC on the BOI to ignore issues relating to food.

The only way that this Committee could have fairly assessed the merits the OCI's criticism or the IIB's defence, would have been to reinvestigate both of their investigations, and possibly the incident itself. That would have been a gargantuan task. It would have involved obtaining the entire investigative product from both organizations. That is a vast quantity of information. We would likely have requested interviews with investigators, support staff, decision makers and others at OCI and CSC who had any involvement with the case. Once we had reviewed everything, we would have had to analyse all the information we had gathered and, potentially, come to a conclusion about which point of view, if any, had sufficient, relevant, reliable evidence to support it. Or possibly, that there was insufficient evidence to reach a conclusion. As mentioned repeatedly in this report, we were simply not given the time or the resources to do this.

Additionally, we would have had to undertake a similar process in each of the 25 BOIs we were asked to review. This, too, was not possible for us to do. Accordingly, we were obliged to rely on information that is in the BOI themselves, along with information from other sources, such as interviews with various parties.

So here is what we concluded. We were not convinced that there is a pressing need, at this point in time, to appoint an external, independent investigator to conduct such investigations, for a number of reasons.

The first, and most important, is that the BOIs we reviewed were, on the face of it, free of bias and conducted reasonably thoroughly. We found no evidence of issues being systemically ignored or swept under the carpet. With some exceptions, and based on the facts set out in the report, relevant evidence was gathered.

Additionally, and these were very much secondary considerations, there likely would be significant costs in implementing Recommendation 10. External investigators, often retired judges, and the ancillary staff required would involve a very significant expense. We are likely talking hundreds of thousands of dollars to retain an external entity to conduct an investigation of the scope likely in the cases identified by the OCI. Further, there may be delays caused by identifying and retaining a suitable external investigator, the investigator and their team may have a considerable learning curve to become an instant expert about CSC context, structure and practice, and there is a possibility the process would become adversarial. At worst, there is a danger the process could evolve into a de facto mini public inquiry, with parties seeking standing and seemingly endless legal challenges.

But that does not mean that we should retain the status quo. While we did not find egregious examples of the flaws in the BOI investigation process to the degree stated in the OCI report, we did find areas of concern. They are discussed briefly in the impartiality segment of this ToR below, and in detail in later chapters of this report.

In addition, and as also mentioned repeatedly in this report, perceptions of the independence of investigations when a person dies in the custody of the state are extremely important, regardless of how thorough and objective any given investigation actually is.

For those reasons, we agree with the OCI that there should be a greater degree of independence in the system for investigating deaths in custody, including in the circumstances mentioned in Recommendation 10.

Greater independence should not be limited only to the circumstances set out by the OCI. It should also include cases that may not fall within those categories, but where it is in the public interest to introduce an enhanced degree of independence into the BOI process.

We looked at various ways for enhancing real and perceived independence in the BOI process. None were perfect. However, we believe that the appointment of an Independent Observer (IO) in high profile IIB investigations may be a solution.

An Independent Observer (IO) is an individual who is not associated with the entity conducting the investigation, who monitors the investigation as it unfolds to ensure that it is conducted impartially.

The role of an IO is set out in the Terms of Reference for a particular case. The common factor is that an IO is not an investigator; rather she or he is an observer. An IO does not play an active role in the investigation itself. He or she does not conduct interviews or analyse evidence. While ToRs vary from case-to-case, they focus is generally on ensuring that the investigation is carried out impartially.

We recommend that an IO be appointed in the instances identified by the OCI in Recommendation 10. They are:

• Major disturbances (riots) resulting in death or injury
• Suicides in segregation^{Footnote 18}
• Use of force interventions leading to serious bodily injury or death

In addition, we recommend that these parameters be expanded to include incidents that may not fall within those categories, but also in case where it would be in the public interest to introduce a greater degree of independence into a BOI process. The Commissioner should have the discretion to appoint an IO to monitor any IIB investigation. The Minister could also direct the appointment of an IO, in any circumstance he or she deems fit, though it is unclear to us if that may involve an amendment to the Inquiries Act.

IOs were originally introduced in death and serious injury investigations in 2007. The RCMP and the then Commission for Public Complaints against the RCMP (CPC) created a pilot project, in response to concerns raised about the impartiality of RCMP investigations into such incidents. The pilot was limited to BC. Since then IOs have been appointed to monitor RCMP serious incidents in other Provinces, including Manitoba, Alberta and Newfoundland.

The Government of Quebec has also appointed an Independent Observer in a very high-profile case. That case is discussed in some detail later in this segment.

The precise role of an IO has varied from investigation to investigation. In some cases, it is limited to ensuring that the investigators have no conflict of interest. In other cases, the IO has a much broader function, including determining the overall impartiality of the investigation. However, given the creation of civilian oversight agencies that investigate police involved deaths and serious injuries in virtually all provinces since 2007, the IO project had become largely redundant for these types of investigations. That said, the concept of independent observation of an investigation that involves a death in the custody of the state remains a valid and useful one.

IO's have also been appointed by the RCMP to monitor incidents that do not involve death or serious injury. In 2014, allegations were made about harassment and misconduct at the Canadian Police College Explosive Training Unit (CPC ETU). The incidents had allegedly taken place between 2012 and 2013. The RCMP conducted an internal investigation and a resolution was reached. However additional allegations surfaced which called into question the adequacy of the initial investigation and resolution process.

This led to concerns being expressed about not only the adequacy of the original investigation, but also whether the RCMP was capable of investigating itself thoroughly and objectively. The Commissioner of the RCMP ordered that the RCMP conduct a review of how the allegations were handled and the initial investigation into them. Further, the Commissioner directed that a new investigation be undertaken in respect of new allegations of misconduct at the ETU that had recently emerged.

The RCMP requested that the CPC act as an Independent Observer of this review, which was conducted internally and overseen by a Steering Committee consisting of senior RCMP personnel, including two Deputy Commissioners. The main ToRs for the IO in this case are reproduced below.

1. The Independent Observer will assess whether the work of the multi-disciplinary team and the Steering Committee is thorough, impartial and professional. Impartiality will be understood as an absence of prejudice or bias, actual or perceived, in the outcome of the investigation(s) and reviews that will be guided solely by the evidence.
2. Throughout the course of all activities, the Independent Observer is expected to make observations and recommendations on any issues relating to his or her mandate. These observations and recommendations may be made to the Steering Committee or, in the event the recommendation relates to the Steering Committee or at the preference of the Independent Observer, to the Commissioner. This includes identifying potential concerns and proposing solutions for resolution of such concerns.
3. The Independent Observer will provide a final report of his or her observations and conclusions as soon as feasible following the conclusion of the investigation and review processes. The format of such report shall be determined by the Independent Observer.
4. The Deputy Commissioner, Specialized Policing Services and his office will act as liaison to the Independent Observer to ensure that he or she has access to the information and documents he or she deems to be required. This will include reasonable access to employees for the purpose of interview/discussion.
5. The Independent Observer will be invited to all meetings of the Steering Committee.
6. The Independent Observer is expected to provide a conclusory presentation to government, media, and/or other identified stakeholders.^{Footnote 19}

On July 16, 2016, the IO issued his report, which was made public. He found that the investigation had been conducted impartially, that the investigators had suitable experience and expertise and that the investigation was appropriately resourced and that its recommendations would, if implemented, improve how the RCMP dealt with harassment in the workplace.^{Footnote 20}

The RCMP/CPC IO programme had mixed results. There has been criticism from some quarters that it is a fig leaf and is not robust enough, as the IO does not have an investigative function. On the other hand, the Val D'Or case, discussed below, is an excellent example of how the process can be extraordinarily effective and have credibility amongst a spectrum of stakeholders.

That is not always the case, unfortunately. The appointment of an IO by the RCMP in the fatal shooting of Donald Dunphy by a Royal Newfoundland Constabulary officer on Easter Sunday 2015 was not a success, according to the Commission of Inquiry into the incident.^{Footnote 21} This was, in part, because the IO retained by the RCMP, a retired judge, took a far more proactive role in the investigation than his ToRs permitted. This led to confusion and misunderstandings that played out in the public domain when the IO gave interviews to the media after the investigation concluded.^{Footnote 22}

The Commissioner, Justice Leo Barry of the Court of Appeal of Newfoundland and Labrador, found that the IO had exceeded his mandate and engaged in improper speculation in his report. He made the following recommendations about IOs in his report, with commentary explaining why:

An Independent Observer can be an effective method of increasing public confidence in police investigations of police. Even if a civilian oversight agency is established in this Province, there may still be a role for an Independent Observer for cases that fall outside of that agency's mandate.

Although the use of an Independent Observer did not meet its objective in the investigation of Mr. Dunphy's death, there is no need to throw the baby out with the bath water.

An Independent Observer's mandate and duties should be clearly defined before the Independent Observer commences work.

Commentary: The details of the procedures and protocols that the Independent Observer and the investigating agency are to follow should be clearly articulated and communicated to all parties involved.

The Independent Observer should follow a protocol that objectively measures the integrity and impartiality of the investigation.

An Independent Observer should have unrestricted access to members of the investigating force and the disclosure collected, but should not have direct contact with any potential witnesses.

Commentary: An Independent Observer is not an investigator and should not be permitted to take any investigative steps. If the Independent Observer has concerns during the investigation or suggestions that could improve the impartiality, fairness or transparency of the investigation, these should be communicated to the investigative team. The investigative team should maintain final decision-making authority over the direction of the investigation.

An Independent Observer should be subject to appropriate terms of confidentiality, but the ultimate findings of the Independent Observer should be made public, subject to any redactions or other modifications that may be needed to protect legitimate privilege or privacy interests.

Commentary: While it is not appropriate for an Independent Observer to make public statements or speak to the media other than with the advance approval of the investigative body, to maximize transparency and accountability of the process the Independent Observer's ultimate findings should be published.^{Footnote 23}

Should CSC accept our recommendation to appoint an IO in certain cases, we strongly suggest that they incorporate these recommendations into their process for doing so.

In 2014, a series of allegations were made against Sûreté du Québec (SQ) officers based in Val D'Or, Quebec, alleging widespread sexual assault of Indigenous females in the community. The Service de police de la Ville de Montréal (SPVM) was assigned to conduct an investigation.

Given concerns about police investigating police, in November 2014, the Quebec government appointed an independent civilian observer to monitor the SVPM investigation, Professor Fannie Lafontaine. Her mandate was to examine and evaluate the integrity and impartiality of the SVPM investigation. Professor Fontaine is a lawyer and Professor at the University of Laval, with a background in human rights law, including working for the UN High Commission for Human Rights.

In her November 2016 report, she drafted an Independent Observer Protocol, as a guide to what Independent Observers should consider as they conduct their work. It provides a detailed roadmap for future IOs, including do's and don'ts, along with a series of indicators that IO's can use to measure the impartiality of the investigation they are monitoring.^{Footnote 24} The Commissioner in the Donald Dunphy Inquiry, who reviewed the Protocol and embedded many of its principles in his recommendations, noted that it offers an approach that would ensure that the evaluation of an incident by an IO would be 'fair, transparent and impartial.'^{Footnote 25}

The Protocol is an excellent basis for determining exactly what the role of an IO would be in the IIB BOI process. We will now quote extensively from Professor Lafontaine's report, to explain why we came to that opinion.

In her report, Professor Lafontaine set out what the objectives of an IO are:

The objectives of the observation process are defined in my mandate as follows:

• Increase public confidence in the impartiality of police investigations;
• Increase the perception of the integrity and transparency of the process;
• Build confidence in the respect for victims' rights.

She also discussed what authority she was provided with, in order to fulfill her duties.

To carry out the mandate, it was agreed that I would have access to the documents, locations, and individuals required. More specifically, I can:

• Obtain any documents or information deemed useful from SPVM;
• Communicate with the supervisor of the investigators assigned to cases and obtain information relevant to my mandate;
• Meet with anyone able to provide information relevant to the assessment of the investigation's integrity or impartiality;
• Visit, as needed, certain locations related to the investigation (accompanied by investigators);
• Review various testimonies, whether in the form of transcripts or video recordings or by witnessing them in person in an adjacent room.

Professor Lafontaine noted that her mandate was restricted in the sense that she was not an investigator, emphasizing that her role was to monitor, not conduct, an investigation. She did not contact complainants, subject officers or other witnesses. Nor did she attend interviews in person though, as noted above, she did review video or audio recordings of them.

In the introduction to her report, she set out conditions necessary for an IO to do their job effectively, as follows:

• To be effective and credible, the model for independent civilian observation of a police investigation of other police officers is based on a series of key conditions. They include:
  • Full access to the evidence at every step of the investigation;
  • Unrestricted access to the entire investigation team of the police force under observation and the police force's full cooperation;
  • The opportunity to meet with anyone able to provide observations and information on the investigation (subject to restrictions on direct contact with victims, witnesses, police officers involved, and witness police officers);
  • The transparency of the process and of the observation results;
  • The appropriate resources to carry out the task. The scope of my mandate and the manner in which I interpreted and applied it respect these conditions.^{Footnote 26}

The Protocol covers such areas as objectives, definitions and guiding principles. It includes what information the IO should be entitled to. It has a detailed segment on what criteria the IO will use to determine the integrity and impartiality of the investigation. They include:

• Consistent application of a rigorous established investigation process at every step of the investigation,
• Consideration of the Indigenous context and the sexual nature of the allegations at every step of the investigation,
• No real or apparent conflict of interest between the members of the SPVM investigation team and the police officers involved, witness police officers, other witnesses, members of the management team at the station under investigation, or the victims.

Professor Lafontaine expands on what she means in each of the above categories in the Protocol. For example, she defines a 'rigorous established investigation process' as follows:

1. Timeliness of the investigations;
2. Courteous and respectful behaviour at all times toward victims, witnesses, and police officers involved;
3. Presence of highly qualified investigators who have the training and experience required to carry out investigations;
4. Appropriate intervention commensurate with the gravity of the incidents under investigation;
5. Investigation methods and approaches similar to those used for crimes of the same gravity committed by civilians;
6. For current incidents, measures taken by SPVM to isolate the police officers involved or witness police officers and to restrict communications between them after an incident until their interview with SPVM investigators;
7. For past incidents, verification by SPVM of the measures that were taken by SQ^{Footnote 27} or another police force to isolate the police officers involved or witness police officers and to restrict communications between them after an incident until their interview with SPVM or SQ investigators, as the case may be;
8. Rank of the SPVM investigators who conduct interrogations with respect to the rank of police officers involved or witness police officers;
9. Respect for everyone's fundamental human rights, particularly those guaranteed by the Canadian Charter of Rights and Freedoms and Québec's Charter of Human Rights and Freedoms;
10. Seriousness and thoroughness of investigations, particularly through the exploration of all reasonable investigative leads to determine whether a criminal act was committed and identify those responsible, and by providing appropriate follow-up to further investigation requested by the Director of Criminal and Penal Prosecutions.

Clearly, in the Val D'Or investigation, the role of the IO went well beyond examining the investigative independence basics, such as whether the investigation was appropriately resourced, that there were no conflicts of interest and that the investigators had appropriate skills. Rather the IO's role in this case extended to the thoroughness of the investigation. It also evolved into providing advice, not instruction, when requested. In an interview with an IRC member, Professor Lafontaine described her role not just as an observer but also as, on occasion, a mediator and a sounding board for investigators.

This does not mean that the IO is a de facto investigator, or attempts to direct the investigation. The IO is there, in part, to ensure that all reasonable investigative stones are turned over by the investigators.

We are very confident that Professor Lafontaine's methodology would work well as a template for an IO in a CSC context. All of the criteria she sets out are adaptable to a non-natural death in custody BOI.

The IO should not be involved in the crafting of recommendations. However, if, in the opinion of the IO, a BOI egregiously and wantonly ignores or avoids making findings and recommendations that are, in the IO's opinion, clearly warranted, without a reasonable explanation, the IO should comment accordingly in his or her report. He or she should set out detailed reasons for doing so.

We believe that an IO's mandate should include monitoring the process of accepting or not accepting any recommendations made by the BOI. The focus of the IO at this stage would remain the same, was the review process impartial? It would not be to advocate for specific recommendations or to fetter senior management from making executive decisions. Rather the sole purpose of IO involvement at this point would be to ensure that no undue influence was brought to bear on BOI members or other CSC staff to amend or alter their findings or recommendations.

The IO should remain in place at least until the final report is complete and any corrective action are in progress. Although we have not made a formal recommendation on this particular aspect, CSC may want to consider having the IO monitor the implementation process until all recommendations are completed, in order to provide tangible oversight throughout the entire BOI process.

The Commissioner in the Dunphy Inquiry wrote that:

The IO should make any findings and observations public at the conclusion of the process. Those findings and observations should be limited to areas covered by his or her mandate.

In the event that the IO identifies an issue that falls within his or her mandate as the investigative process is in progress, and is unable to resolve it in good faith with CSC to his or her satisfaction, then the IO should have the authority to make a public report about his or her concerns, as soon as he or she deems fit. The rationale for this is that if an IO has concluded that an investigation is going off the rails as it is underway, every reasonable measure should be taken to get it back on track before it is concluded.

The selection of an IO is key. In RCMP cases, the RCMP selected the IO, with input from the CPC in some instances, as far as we understand it. In the Val D'Or case, the IO was appointed by the Minister for Public Safety, not the SPVM.

There may be value in consulting outside of the investigative agency when appointing an IO, in certain circumstances. In Manitoba, in 2008 the RCMP signed a protocol with the Assembly of Manitoba Chiefs, the Southern Chiefs Organization and the Manitoba Keewatinowi Okimakanak that permits civilians:

The selection process of an IO must focus on identifying individuals who are clearly independent from CSC and have no real or perceived vested interest in the outcome of an investigation.

The selection of an IO should be based on these principles:

• The IO should be appointed as soon as is practical after an incident occurs.
• The IO should not be a current or former provincial or Federal correctional service employee. Given that the BOI will have CSC members, and that it is possible that the CBM will have non-CSC corrections experience, knowledge of the correctional system is not necessary for an IO candidate.
• The IO's skills and background should include extensive experience in conducting or reviewing investigations, in particular investigations that have a systemic component.
• Where possible, an IO should have direct knowledge and/or experience of any racial or cultural issues that may be relevant to the incident under investigation.

We appreciate that there will inevitably be logistical problems in the selection process, due to contracting and other possible bureaucratic impediments. Some of the challenges we discussed above about the appointment of an external investigator - cost, delay and so on, may also apply to appointing an IO. However, those challenges will be much less acute, given the far more limited duties of an IO.

If an IO is appointed to monitor an investigation, it does not obviate the need for full CBM participation in a BOI, including a CBM acting as the Chair of a BOI, when circumstances warrant.

In her November 2016 report, Professor Lafontaine concluded that the SVPM investigation was conducted thoroughly and fairly. She set out her reasons why in great detail. While there was some initial criticism that the IO was not of Indigenous origin and that a public inquiry should have been called instead, overall, her work was very well received. Appointing an IO in this instance dealt very effectively with legitimate concerns that internal investigations - in this case police investigating police - are inherently biased. By any objective standard, it was a successful process.

We believe that appointing an IO who has a role similar to that set out in Professor Lafontaine's Protocol has every chance of being successful in CSC death in custody investigations, provided an IO is properly resourced and enjoys unequivocal support from CSC at all levels. That said, and as mentioned above, the IO system has proven far from perfect. We suggest that IOs be used for a reasonable trial period, to be determined by CSC. If unsuccessful, CSC should then consider supporting the approach set out in Recommendation 10.

• Based on our assessment of the BOIs we reviewed, we did not find compelling empirical evidence to support the totality of the OCIs findings about flaws in the BOI process in general. Accordingly, we do not fully support the OCI's Annual Report 2017/18 Recommendation 10.
• That said, there are clear areas of concern evident in some of the BOIs we reviewed. They are discussed in the impartiality segment of this ToR and in the response to other ToRs.
• We do not believe that, at this point in time, it is necessary to remove CSC from the investigative process entirely by appointing an external, independent investigator into the type of incidents identified in Recommendation 10.
• We agree with the OCI that there should be a greater degree of independence in the system for investigating deaths in custody, including in the circumstances mentioned in Recommendation 10.
• In order to provide an enhanced degree of independence, CSC should consider appointing an Independent Observer (IO) for each BOI that falls within the criteria set out above, for the duration of the investigation, including the NIM review process.
• In order to enhance public confidence in the integrity of the BOI process, the IO should issue a public report at the conclusion of his or her work.
• Enhanced independence in the BOI investigative process should not be limited only to the incidents identified by the OCI. It should also include the investigation of incidents that may not fall squarely within those categories, but where it is in the public interest to introduce a greater degree of independence into the BOI process. An example of such an incident would be the current investigation into the death of a sex worker in Quebec. In such circumstances, the Commissioner should have the discretion to appoint an IO to monitor the investigation. The Minister should also have the authority to direct the appointment of an IO, in cases where she or he deems it in the public interest to do so.

We recognize that IOs have only previously been used, to the best of our knowledge, in cases involving police organizations. However, we are very confident that the approach can easily be adapted for IIB BOI investigations that fall within the categories identified above.

• The IO should have a similar mandate and authority to that given to Professor Lafontaine in the Val D'Or investigation. They should not have a direct investigative role. Rather they should ensure that the BOI process is impartial, that all reasonable investigative leads are pursued to an appropriate degree, that there is no interference into the investigation from any party, that the BOI report reflects the evidence gathered and that no inappropriate influence was brought to bear to change BOI findings and recommendations. If changes are made to findings and recommendations, then the reasons should be made public in the IO's final report.

An Independent Observer should be appointed to monitor a Board of Investigation in the instances defined by the Office of the Correctional Investigator in Recommendation 10 of its 2017/18 Annual Report, as well as any other death in custody where the Commissioner or the Minister of Public Safety determines it would be in the public interest to do so.

All non-natural death in custody BOIs must include at least one Community Board Member (CBM). Additionally, BOIs that are convened under Section 20 and Section 154 (4), which relates to persons on parole, must be chaired by a CBM.

We understand that the Community Board Member programme was created about 25 years ago, as a result of a recommendation by the OCI.

CBMs are individuals who have never been employees of CSC who are retained by CSC to sit on BOIs as, in the words of CSC, 'fully fledged' members. They participate in the evidence-gathering phase of the investigation, including conducting interviews with fellow BOI members, normally at the institution. This process normally lasts for two to three weeks. According to IIB, while the CBM does not write the report, they are expected to participate in briefings and the formulation of findings and recommendations. They are also given an opportunity to comment on the final draft of the report.

According to IIB, CBMs are selected for their specific expertise and knowledge in any given case. For example, in a BOI involving an apparent suicide, the CBM may have a background in psychology with a specialization in suicide and self-harm related issues.

CD 041 stipulates that BOIs into incidents involving Healing Lodge issues 'will normally include an Aboriginal community member.'^{Footnote 30}

We reviewed short biographies of the CBMs who were appointed to the BOIs we were tasked to review. Several had a provincial corrections background; others came from academia or the policing world. Some were the CBM on several BOIs during this period, including two who were CBMs on four separate BOIs.

The duties of a CBM are set out in the letter that IIB sends to all Board Members at the beginning of an investigation. They include:

• Certifying the integrity of the process
• Contribute to all the steps in the investigation, including conducting research and taking part in interviews and debriefings.
• Respond to questions about the independence and objectivity of the investigative process

A CBM has access to all information that other Board members do. However, there may be security and information technology issues that can impede their work. We heard for example that CBMs may not always have direct access to CSC information systems and may have to make requests for such information via fellow Board Members who are CSC employees. It can be difficult obtaining documents due to security requirements. We understand that CBMs are not usually issued CSC laptops or other devices. We believe that they need the same tools that are available to other Board members, to do the job.

We encourage IIB to ensure that CBMs fully participate in the post investigation process, including local debriefings. It is unclear whether this is currently happening. It should be, as the independent perspective that a CBM brings can be equally as valuable at this stage of the process as it is in those that precede it.

CBMs are not involved in the NIM process, nor are NIs. We were advised by IIB that this was because at this point all BOI members, including CBMs, had 'agreed upon the last version of the report presented.'^{Footnote 31}

We also identified issues re training, both for CBM members in investigative techniques and for CSC employees about the role of a CBM. We will discuss these later in this chapter.

One of the primary benefits of appointing CBMs to BOIs is, of course, that it provides an enhanced element of independence in the BOI process. CBMs are not, and never have been, employees of CSC; few, if any, have any direct connections within CSC (at least as far as we are aware), they are not subject to any pressure from any quarter about future career considerations. They can, at least in theory, provide a fully objective opinion on what they are investigating, and just as importantly how it is being investigated, without fear or favour.

The CBM therefore acts not just as an investigator but also as a barometer of the independence and impartiality of the BOI. In an interview with us, the Correctional Investigator advised that, in his view, a CBM should play a 'challenge role' in the BOI process, where warranted. We agree. If a CBM does not agree with investigative steps (or lack of them), findings and recommendations, they should be actively encouraged to include their views and the reasons for them in the final report itself. That this is one of their key duties this should be incorporated into the CBM induction and training process, as well as clearly stated in the letter they receive that sets out their responsibilities prior to a BOI.

Another significant benefit is that CBM participation provides a magnificent opportunity to ensure that potentially relevant cultural and race issues are considered throughout the investigative process, particularly in cases where those issues may be a factor in a death. For example, we were advised that IIB has retained experts from the Indigenous community to be CBMs in cases involving Indigenous offenders, in cases not involving healing lodges where, as mentioned above, it is mandatory to do so.

CSC should expand this initiative to proactively identify and retain qualified individuals from communities who are disproportionately represented in the federal corrections system, and appoint them as CBMs to BOIs in appropriate cases.

Another consideration when selecting CBMs may be identifying individuals who volunteer and/or are involved in organizations whose work involves the correctional system, including victim and inmate advocates, as a possible source of CBM candidates. The key principle should always be the real and perceived objectivity of any such potential appointee. Knowledge of the prison system may be desirable - but it should not necessarily be a prerequisite.

While CBMs are appointed as Chairs in parolee related Tier 1 investigations - such as the one currently being conducted in the case of the homicide of a Quebec sex worker, allegedly by a parolee - none of the BOIs we reviewed had a CBM as chair. We believe that there are significant advantages to appointing a CBM as Chair in high profile cases. For example, it is clearly in the interests of both the public and CSC to appoint a CBM chair in all or some cases identified by the OCI in Recommendation 10. This is so even when an IO is involved. Given that a CBM is an investigator, and an IO is not, having both involved in a high-profile investigation enhances real and perceived independence of the process. Other Board members who are CSC staff, including NI's, would provide technical and investigative expertise to support the Chair and would of course be involved in reaching findings and crafting recommendations.

We appreciate that having a CBM as Chair may create additional logistical problems, particularly in cases where CSC will be also appointing an IO. Regardless, we believe that it will be worth the effort by providing a clear and transparent enhanced degree of independence to the entire process, particularly the investigative component.

Finally, CSC is to be congratulated for using CBMs for death in custody BOIs, as is OCI for recommending that they do. CBMs bring not just a measure of independence but also, as one very experienced IIB member said, a 'different life experience and a different lens' to an investigation. None of the other agencies we spoke to across the world have a similar approach. IIB is, to the best of our knowledge, unique in bringing individuals with significant relevant expertise from outside the agency to participate directly in the conduct of a death in custody investigation.

• The use of CBMs is commendable.
• Increased use of CBMs as BOI Chairs would strengthen perceptions of the independence of the BOI process, particularly in high profile cases.
• CBMs should be actively encouraged to challenge and question throughout the BOI process.
• Appointing appropriately qualified CBMs can ensure that potentially relevant cultural and race issues are considered during the investigative process, in cases where those issues may be a factor in a death.
• CBMs are not always provided with the same tools as Board Members who are CSC employees.

Correctional Service Canada should appoint Community Board Members as chairs in high profile Boards of Investigation, where it is reasonable and practical to do so.

Correctional Service Canada should work to appoint Community Board Members who have expertise and insight into issues facing disproportionately represented inmate populations, particularly in cases where race and/or cultural background may be an issue in a given investigation.

All Community Board Members should be given the same tools to do their job as those provided to Board of Investigation members who are Correctional Service Canada employees.

Just because an investigative agency is not independent of the entity it is investigating, as is the case with IIB, does not mean that its investigations cannot be impartial.

How do you measure impartiality? An effective test is to assess how thoroughly and objectively an investigation was done. In this segment of the report, we do our best to do so, based primarily on the information that was set out in the BOI reports themselves.

Which leads us to repeat, without apology, our important caveat. We did not review any of the investigative product or other related material created by or for any given BOI, such as emails, investigators notes, investigation plans and transcripts of interviews. Nor did we interview individuals who had participated in any capacity in a specific BOI. Such steps would certainly have provided greater insight into the process itself, how the investigation was conducted and potentially revealed any issues related to independence and/or impartiality.

As mentioned more than once in this report, the reason we did not do this is that we simply were not given the time to do so. This of course restricted our ability to conduct a full analysis of the thoroughness of any particular investigation and, by extension, its impartiality.

We want make it crystal clear, once again, that we were in no way obstructed or denied access to information that would have allowed us to conduct a more thorough analysis. In fact, CSC offered us access to any BOI related documentation, as well as access to personnel.

While not having access to this information was a serious handicap, we do not believe it was a fatal one. The BOI reports themselves, plus information we gathered via interviews and other sources, leave us confident that our findings and conclusions about the thoroughness and fairness of the BOIs we reviewed are based on a reasonably firm evidentiary foundation.

Impartiality in an investigative process boils down to those who are responsible for conducting, supervising or reviewing an investigation doing so with an open mind and without any real or reasonably perceived conflict of interest. It means not bending to pressure or tailoring an investigation to reach a predetermined or preferred outcome. It means identifying and exploring all reasonable investigative avenues to a reasonable degree, turning no blind eyes in the process. It means going where the evidence leads. It means having appropriately trained investigators. It means conducting thorough interviews, including asking difficult questions where necessary. It means making sure evidence is analysed objectively. It means, in short, conducting an investigation without fear or favour, where no reasonable investigative stone is left unturned.

In order to determine if an investigative process was impartial, it is helpful to go through the investigation with as fine a toothcomb as possible in the circumstances, using these principles:

• The investigators must be as independent as possible.
• Were the investigators appropriately trained and experienced?
• Were all potentially relevant issues identified and, where appropriate, pursued?
• Was the investigation sufficiently resourced?
• Did the investigators identify and gather all potentially relevant physical and digitally stored evidence? Was that evidence then appropriately reviewed?
• Was all relevant documentation secured and reviewed?
• Were all persons who may potentially have information about what happened identified and, where necessary, interviewed?
• Was the analysis of the evidence gathered during the investigation objective and based solely on the facts?

We were pleased to note that IIB refer to these as 'key investigative principles' in their own training manual.^{Footnote 32}, ^{Footnote 33} If these principles are not followed it is likely, based on our experience, that an investigation will not be impartial.

We reviewed the BOIs and other material using the criteria set out above. This is what we found.

The investigators must be as independent as possible.

We discussed this principle in Part 1 of this chapter.

Were the investigators adequately trained to conduct the investigation? Did they have sufficient experience in conducting investigations of this nature? Much will depend on the seriousness of the issues under investigation. Death investigations, particularly deaths that occur while in the custody of the state, are by definition extremely serious.

We understand that there is a formal training programme for incoming IIB staff, including National Investigators. We have been advised that steps are being taken to beef up training, including an increased use of case studies based on real incidents. We were advised that the principles of independence and impartiality are foundational to the training process, including information about bias and unconscious bias, and the difficulties of remaining neutral throughout the entirety of an investigation. There is also refresher training, including attending investigations courses.^{Footnote 34}

We reviewed the investigations training manual for IIB staff.^{Footnote 35} It covers the investigative basics, including planning the investigation, identifying issues, approaches to interviewing and so on.

One of the first things that we noticed was that the training material on who to interview is somewhat focused on investigating CSC staff. The 'sample interview questions' segment is almost exclusively geared toward questions for CSC employees.

We appreciate that the focus of BOI investigations is what CSC staff did and this forms a large part of the investigation. However, it should not be to the exclusion of other parties who may have useful information, depending on the circumstances of a particular case. When planning their investigations, BOIs should consider whether other potential sources of evidence, particularly in cases that may have a systemic component. Those sources may include inmates, community-based correctional organizations such as the John Howard or Elizabeth Fry Societies, Canadian Families and Corrections Network^{Footnote 36}, St. Leonard's Society, and victims, family members and others. We discuss this further in a later segment of this chapter.

In respect of training for CBMs, CD 041 provides that '….to the extent possible, each board of investigation should include individuals with appropriate expertise into the incident being investigated'. In many cases, the experience and background of the CBM will fulfill this requirement. Of course, that expertise should include knowledge of how to conduct investigations.

We were surprised that CBMs currently receive no training in how to conduct investigations, particularly as CBMs play such a vital role. The training manual provides that 'each Board member is expected to fully participate in the investigation.' If CBMs are to be 'fully fledged' members of BOIs, as IIB told us they were and is also stated in the letter they receive setting out their duties prior to a BOI beginning its work, then it is essential that they receive training in how to execute their core function, investigation.

When we asked IIB why no training was provided, we were told that CBMs used to receive training but it has not been offered to them 'in the past few years.' There were barriers to training, including costs and concerns about entering in an employer/employee relationship.

A new training programme is being prepared by IIB but no decision has been made yet as to whether it will be offered to CBMs, though the move to online delivery may make it easier to include them.

We think it is essential that all CBMs should receive training into how to plan and conduct investigations, in particular in conducting investigative interviews, as well as the challenges of systemic investigations. CBM training should also include a component about how CSC functions, its Mission and Core Values. Most importantly, any training should fully address the vital role of a CBM challenging and questioning if necessary, what a BOI is doing, finding or recommending, as the OCI suggested when we interviewed them.

We appreciate some CBMs may already be experienced investigators, including former police officers. That does not negate the requirement for training, either as a refresher or because the type of investigations that BOIs undertake have a different focus and process than criminal investigations.

We leave it to IIB to determine if CBM training can be coordinated with existing training programmes. We recommend that CBMs also receive refresher training prior to being appointed to a subsequent BOI, as required. As stated above, objectivity and thoroughness, and therefore impartiality, depends, in part, on board members being appropriately trained and CBM's are no exception.

On a related issue, we were also advised that CSC staff, including some NIs, may not be fully up to speed with the role of a CBM. If that is indeed the case, IIB should consider adding a segment on the role of CBMs to its training curriculum.^{Footnote 37}

The training material for BOIs on who to interview is very much focused on CSC staff. When planning their investigations, BOIs should consider whether other parties should be spoken to, particularly in cases that may have a systemic component.

CBMs do not receive training in how to conduct investigations. It is essential that they do.

The Incident Investigation Branch should develop an additional component to its existing investigation training programme for all Board of Investigation members that covers contacting parties from outside Correctional Service Canada, who may have relevant evidence, including inmates, their victims, and their families.

The Incident Investigation Branch should create a formal training programme for Community Board Members, including training on investigative interviewing, Correctional Service Canada Mission and Core Values and their vital role in questioning and challenging, if necessary, how the investigation is proceeding. This training should be delivered as soon as a Community Board Member is appointed, with refresher training as required.

One measure of thoroughness and impartiality is whether all potentially relevant issues were identified as the investigation began and progressed. What rationale was provided either to explore or not explore any such potential issue? Was each issue pursued to an appropriate degree?

This is particularly relevant, given the OCI's concerns about a BOI ignoring a key relevant issue in the Saskatchewan Penitentiary investigation.

Overall, we believe that BOI investigations are moving in the right direction in how they identify and pursue issues, though possibly not as quickly as some would like. IIB senior management argue that there is now more of a focus on the reasons why a death occurred, rather than whether policy was complied with. This change in emphasis was confirmed by front line staff and corroborated to a degree by the recently introduced report-writing format discussed elsewhere in this report.

As noted above in the segment on the OCIs 2017/18 Annual report, in the BOIs we reviewed, we did not find clear evidence of major issues being systemically ignored or otherwise swept under the carpet.

We did, however, identify several areas of concern in terms of issue identification and appropriate follow up. While they were not apparent in every BOI, they did appear often enough to warrant comment.

In particular, in several BOIs we noted areas relating to a death that were touched on, but were then not adequately dealt with in the report. In some cases, the issue was either flagged by the BOI, or clearly apparent in the narrative of what had happened, yet there was no follow up in the report itself. These issues included, but were not limited to:

• defective equipment,
• suspension points in cells
• delays in entering cells to provide first aid, and
• delays in notification of Next of Kin.

In fairness, there may be a reasonable explanation as to why these issues were not pursued. If that is the case, then the BOI should provide a clear explanation as to why in its report. It did not in the cases we reviewed.

Our concerns are discussed in more detail in a later chapter of this report.

Additionally, and as discussed in the previous chapter, we noted that CSC Mission and Core Values were not directly referred to in BOI ToRs, and made a recommendation in that regard.^{Footnote 38}

Planning an investigation is a crucial part of the process. If it is not done properly, issues may not be identified and followed up. A good investigation plan will assist a BOI develop a road map for the investigation, pinpoint sources of evidence, use resources effectively, identify potential roadblocks and set timelines and milestones.

The IIB training Manual has a segment on establishing a 'game plan' for an investigation. It is detailed and appears reasonably fit for purpose. We did not review the 'game plan' for any of the BOIs that came within our remit, due to lack of time.

We were told by a senior IIB staff member that IIB provides BOIs with 'all pertinent information to the BOI team for a specific case.' This includes information about similar incidents that occurred at that institution or potentially on a similar set of circumstances at different institutions, as part of the investigative planning process. This is discussed further in the chapter dealing with ToR 4.

IIB recently introduced an Enhanced Investigation Analysis (EIA) process. It was initially dubbed a 'pre-board investigation.'

An EIA is designed to obtain information about the overall context surrounding incidents that can then be used to plan and conduct a BOI, prior to the BOI beginning its work. The decision to use this approach is made on a case-by-case basis by IIB.

In both cases where the EIA has been used to date, the incidents to be investigated by the BOI involved multiple overdoses. One involved a fatality. As part of the EIA process, IIB staff attended the institution and conducted separate staff and inmate focus groups. The groups were open to anyone one who wanted to participate, not only individuals involved in a related incident. The purpose was to 'provide additional detail' for the upcoming BOI. The discussion focused on prevention, harm reduction, treatment and enforcement - the four pillars mentioned in Chapter 1.

The results of the EIA were made available to the BOI and were referred to in the context section of their reports.

We think that EIAs are a welcome development in identifying potential issues in deaths in custody. They can help in the initial evidence gathering process, uncover or flesh out potential issues, including systemic issues, and streamline the work of the BOI.

We believe there is an opportunity for IIB to expand the EIA process into other major investigations, particularly where there is prima facie evidence of a potential systemic issue or issues. It can help a BOI target key issues, hone in on sources of evidence and get investigations competed sooner. As an IIB staff member noted, it is '… a fairly targeted approach to address a specific area or concern of interest'. As such, it is a valuable tool that should be used as often as possible in appropriate circumstances.

We identified several areas of concern about how issues were dealt with in the BOIs that we reviewed.

The recently introduced EIA process is a valuable tool, particularly in identifying potential systemic issues early in an investigation. CSC should consider using it more broadly.

The Incident Investigation Branch should consider expanding the Enhanced Investigation Analysis process to more investigations, where appropriate.

To be done thoroughly, an investigation requires:

• enough people to get it done in a reasonable time. How many will depend on a number of considerations, including the scope of the investigation, the number of people who should be interviewed, the quantity of physical, digitally-stored and documentary evidence that has to be gathered and reviewed and whether it is necessary to consult with experts in a particular area.
• enough time to ensure that all relevant investigation avenues are explored to a reasonable degree.
• the authority to gather all relevant evidence.

BOIs into deaths in custody usually have a fairly broad scope, often involve a significant number of witnesses and large quantities of documentary and digitally stored evidence. They can therefore be resource intensive.

Generally, we found the BOI process to be well resourced, as defined above. The fact that there are usually three individuals on the BOI is a huge advantage from an investigative process perspective. It enhances thoroughness, at least in theory, and allows for a breadth of perspectives. It is an approach that, in our view, is superior to that of any of the other agencies across the world we reviewed. Those agencies tend to assign only one investigator to a death, regardless of the potential scope of the process, though there are some exceptions.

We heard that the workload can be onerous, particularly at the National Investigator level. Currently, there are 12 National Investigator positions at IIB, with about 10 'casual National Investigators' at the Executive level at CSC, who assist when the workload demands. Workload can be significant. In 2018-19, there were 30 Tier 1 and 33 Tier 2 national investigations, along with 43 local investigations. We understand that one NI has four BOIs on the go, at various stages of process. On the face of it, that is a considerable workload.

We heard no complaints about BOIs being unable to take the investigative steps they wished, due to travel or other fiscal restraints.^{Footnote 39}

The deadlines given to BOIs in the Convening Order appear reasonable on the face of it. BOI members who are not IIB staff are given the time to conduct the investigation. Commissioner's Directive 041 - Incident Investigations provides that any CSC staff member assigned to a BOI shall be immediately relieved of his/her regular duties while conducting the investigation and writing the report'.^{Footnote 40}

We briefly discussed the powers to gather evidence that BOIs have at their disposal in the segment on independence, including authorities under the Inquiries Act. They are fit for purpose.

Digitally stored evidence such as CCTV, records of entry and exit, recordings of radio communications, texts and emails, can obviously be vital evidence in a death in custody investigation. It must be identified, gathered, preserved, reviewed and when necessary forensically examined.

In respect of the BOIs we reviewed, CCTV evidence was preserved as a matter of course.^{Footnote 41} Given the lack of detail about investigative steps taken that we comment on later in this segment, it was difficult to determine if all relevant evidence that falls within this category was, or was not, collected and reviewed. Many BOIs mentioned CCTV. We did note a reference in one BOI where members obtained and listened to Correctional Officer radio transmissions.

Generally, the best evidence is the freshest. Memories fade, documents get lost or tampered with and digitally stored evidence erased, intentionally or for other reasons. It is good investigative practice to secure evidence as quickly as possible. This may require significant resources, including having the capacity to gather evidence as soon as possible. Time also factors into the equation because if there are important lessons to be learned from a death, they should be identified and disseminated as quickly as possible.

BOIs take a considerable time to convene. This can result, potentially, in evidence being lost. By the time a BOI arrives at an institution, usually months after the incident, inmates may have been released or transferred. Staff may not be available. Memories will have inevitably faded. There is always a possibility that documentary and digitally stored evidence may no longer be available.

Accordingly, it is good investigative practice to secure evidence that may be relevant as soon as possible. This applies not only to digitally stored and physical evidence but also to documentary and witness evidence.

That is a practice followed by other investigative agencies when a death in custody occurs. Several of the agencies we contacted have a policy of dispatching a staff member to an institution immediately after a death in custody. PONI has a 24 hour on-call system. NIPS will inform PONI of a death immediately and an investigator will attend the institution that day, normally within four hours of notification. He or she will meet with the Governor (Warden), attend the place where the death occurred and canvass for witnesses, including leaving notices on the range. The purpose is, in part, to secure perishable evidence and make inmates and staff aware that an external investigation has been launched. PPO has a similar process. The PPO investigator who attends the Institution posts notices for information, asking staff and prisoners to contact PPO if they have any information.

Both agencies work collaboratively with other entities that may be gathering evidence simultaneously, or who have an interest in what happened, such as police, the Inspectorate and/or the Coroner.

While the OOTI does not normally attend the place where a death occurred immediately, we were advised that it may do so if it chooses, on a case-by-case basis.

In contrast, it appears that IIB relies on the institution itself to gather this evidence, including initial interviews of inmates and others who may have relevant evidence. We were advised that the Security Intelligence Officer (SIO) at an Institution is often responsible for doing this.

In our view, this is not a good practice, particularly from a real and perceived independence perspective. IIB should have direct control, to the extent possible, over the gathering of evidence that may be relevant to the issues a BOI will likely investigate, at all stages of the process. It also sends a clear message that an agency external to the Institution itself is taking charge of the investigation - or at least the aspects of it that fall within its bailiwick - from the very beginning.

IIB also noted that inmates may be reluctant to speak to IIB investigators. That may indeed be the case, but 1) how do you know unless you make the effort of asking them, and 2) it is difficult to imagine that they would be much more enthusiastic about speaking to an SIO

A potential collateral benefit is that it may speed up the BOI evidence gathering process by starting it immediately, thereby cutting down the time taken to conduct BOIs.

In our view, IIB staff should be dispatched to the place where a death occurred as soon as practical after being notified of the death. They should ensure that relevant evidence that may be relevant a future BOI into the death is identified, secured and gathered. They should also conduct preliminary interviews of parties, in particular those who may not be easily available for interview by a BOI at a future date, such as inmates and family members.

IIB should be involved in the initial evidence gathering process when a death occurs, including potentially interviewing inmates as soon as possible.

Incident Investigations Branch staff should be dispatched to the place where a death occurred as soon as possible. They should ensure that all evidence that may be relevant to the death is identified, secured and gathered. They should also conduct preliminary interviews of parties who may not be easily available for interview by a Board of Investigation at a future date, such as inmates and family members.

Documentation is the lifeblood of most government agencies and the prison system is no exception. Investigators should ensure that all documentation that may be relevant to a death in custody should be identified, gathered, read, understood and, just as importantly, explore the reasons for any gaps in the paper trail.

The IIB Training Manual contains a comprehensive list of documents that BOIs should consider obtaining.

All BOIs contained an appendix listing policies and guidelines that were reviewed but not other potentially relevant documentation that may have been part of the investigation, such as log books or personnel records. That said, in the BOIs we reviewed it appears that relevant documentation was gathered, with the caveat that in some cases, information about what the BOI had actually done was so sparse that it made it difficult to come to a conclusion. We discuss how BOIs should improve how they present information about collecting documentation in the analysis of the evidence segment, below.

Another test of the quality of an investigation involves what the investigators did to identify and interview anyone who may have information about what happened. In cases involving potential policy or underlying systemic issues, this extends to anyone who may have knowledge and expertise to assist the BOI in its work.

The test has three parts:

1. Were all persons who may potentially have information about what happened identified as soon as possible?
2. Was there then an evidence-based process in place to determine whether or not they should be interviewed? In some cases, there may be so many potential witnesses that it is simply not possible to interview everyone who may, just may, have information. In those situations, investigators should go through a triaging exercise, to determine who is most likely to have the most information. The key consideration is often proximity to the incident or issue.
3. If an interview - formal or informal - took place, was that interview thorough and fair? A good investigative interview boils down to asking the right person the right questions in the right way, in the right place and at the right time - and giving the interviewee an opportunity to respond fully - then actively listening to what he or she says. Was this done?

As mentioned above, IIB does not identify and interview potential witnesses immediately after an incident, or at some other point prior to the BOI visiting the institution, which is usually months after the death occurred. Arguably, the exception is if there is an Enhanced Investigation Analysis, though they normally occur a considerable period after an incident and have a limited mandate. They are not intended to establish the facts of a particular event, rather they are designed to provide context.

Prior to a BOI, IIB gives members a letter, setting out their responsibilities. The letter includes a short segment on interviewing 'victims' and 'victim's families'. IIB defines a victim as follows…

In the letter it notes that:

We deal with family members meeting with BOIs below.

It is clear that in many of the cases we reviewed inmates may potentially have useful information about the incident, and possibly its precursors and its aftermath. Good investigative practice dictates that such evidence should be collected as soon as possible, normally by interviewing the inmate. As mentioned in the segment on visits to the institution above, PPO and others actively seek out and interview inmates in the immediate aftermath of an incident. The PONI advised that her investigators often interview inmates.

As mentioned above, some interviews may be conducted by the SIO at the institution, in the immediate aftermath of an incident. This is not ideal, particularly from a perception of independence perspective, which is why we recommended above that interviews of persons, such as inmates, who may not easily be available when the BOI attends in the institution months later should be done as soon as possible, by a member of IIB.

It is not always clear what efforts were made to interview inmates. In one BOI report it was noted that, when discussing what investigative steps it had taken, 'The principal inmates involved in the incident were not interviewed' No explanation was given as to why.

We were advised by IIB that BOIs 'often' interview inmates. That does not appear to be always the case in the 25 BOIs we reviewed, at least based on the sparse information set out in many BOIs about what investigative steps where taken. (The lack of detail in BOI reports in general is dealt with later in this chapter). As noted previously, we were advised that many inmates decline to speak to BOIs. We also note that the OCI was critical of the BOI interviewing only one inmate in its Saskatchewan Penitentiary investigation, as well as not interviewing an inmate who appeared to have significant information.^{Footnote 42}

In the BOIs that we reviewed there were always police and Coroner/Medical examiner investigations. Except in cases involving a homicide, the evidence gathering phases of these investigations appear to have concluded fairly quickly and had no impact on the BOIs work.

The IIB training manual provides guidance when there is an ongoing police investigation, as follows:

This is common sense and, based on our review of other agencies, common international practice.

In summary, just because someone else is conducting a parallel investigation does not mean that a BOI has to remain in limbo until the conclusion of that investigation, even if it takes precedence. While any interview done immediately after an event must be done with the consent of any agency, such as police, that takes precedence over the BOI process. We see no reason IIB would not be able to do so as well. BOIs are, to the extent we are able to determine, dealing with concurrent investigations appropriately.

As noted in the previous chapter, an invitation to provide any relevant information is included in the initial letter sent from IIB to the Next of Kin of an inmate who dies in custody. As noted elsewhere in this report, PPO and PONI normally meet with family members, in part to establish if they have any information that may be relevant to the death.

We were where advised by IIB that a BOI may interview family members if their inquiry indicates that they may have information relevant to the investigation, although there is no specific mention of Next of Kin in the IIB training manual. Next of Kin have been interviewed by BOIs in the past, though less frequently more recently.

We found very few references to family members being interviewed in the BOIs we reviewed. In contrast, PONI always interview families, in part because they have a remit to include their queries in the investigation of the death.

There were at least two cases where they should have been, at least on the face of it, or the BOI should have explained why they chose not to do so. An inmate committed suicide on the day he was due to be released. The suicide was totally unexpected by staff and other inmates. Issues emerged over the level of support he would likely receive from family members, including inconsistencies between what he told his case managers on that topic and what was actually correct. He spoke by phone to family members three days before he killed himself. Yet the BOI, as it noted in its report, 'did not interview family members and had no knowledge of the content of the calls…' In fact, the BOI obtained information about the inmate's state of mind and that '….members of the family were shocked to learn of the death' from Chaplains at the institution.

Another case involved an inmate who tried to contact his mother, girlfriend and other parties 28 times by phone, just before committing suicide. While none of the calls got through, it may well have been worth contacting family members to ascertain what, if anything, they knew about the state of mind of the inmate immediately prior to his death, or the content of any voicemails, if they existed.

It is unclear why the BOI did not interview at least some of those family members. There may have been valid reasons why they did not. If so, those reasons should have been clearly articulated in the report itself.

While we fully agree that BOIs must be careful not to impinge on other investigations that rightly take precedence, they should go wherever the evidence takes them, at the appropriate point in the investigation.

In some cases, family members, inmates and others may have information that is relevant to the issues a BOI is investigating. The best investigative practice, as employed by some of the other agencies we contacted, is to proactively canvass families and others for any such information. This appears to be not always done.

Boards of Investigation should identify and, if warranted, interview family members, inmates, victims and anyone else, when it appears that they may have information that is relevant to the death.

Digital Voice Recording of interviews^{Footnote 43}

There is no IIB policy governing the digital voice recording (DVR) of interviews. BOIs do not always digitally voice record interviews. They should, as this is best investigative practice.

The BOI training manual does refer to recording during interviews, but only in the context of interviewees recording their own interviews. Interviewees are permitted to record and 'the person may keep the tape'. While this is not standard investigative practice generally, given concerns about collusion and control of information, it may be reasonable to extend latitude in investigations of this kind, which are focused on systemic issues, not on finding fault.

Digital Voice Recording of interviews is common investigative practice around the world, including for agencies that conduct investigations similar in scope to those that IIB is responsible for. For example, investigators at the PPO and the OOTI digitally voice record their interviews. Numerous investigative agencies have adopted the practice, many of them for decades, including several provincial and federal Ombudsman offices.

There are a number of very significant advantages to DVR of interviews, as follows:

Obviously, it is very important to have an accurate record of an interview. The DVR of an interview will show:

• Exactly who said exactly what, when and in what tone
• The overall tone and pace of the interview
• Asides, phrasing, emphasis, hesitations, pause and interruptions
• How long the interview took, in real time

A written statement or notes taken by a BOI investigator may not show any of these things.

We appreciate that the BOI process is not intended to be an adversarial one. However, it is not impossible that someone who has been interviewed may, at some point, challenge the BOIs record of that interview. The interviewee may claim, for example, that the investigator's record of the interview does not reflect exactly what they said or meant, that the interviewee was hectored or bullied, or that the investigator failed to include important information. Should that situation arise, it may boil down to a credibility contest between the investigator and the interviewee.

A DVR of the interview is usually sufficient to determine the merits of any such complaint, very quickly.

DVR is a cost effective and efficient method for recording investigative interviews. Investigators can conduct far more interviews in a given period if they digitally record them as opposed to writing a statement or taking notes.

This is because writing out a statement or taking comprehensive notes takes a lot more time than digitally recording the interview. The interviewers not only have to concentrate on questions and answers, they also have to assiduously record what is being said, often in significant detail, particularly if they do not know whether that piece of information is, or may become, relevant to the issues or not. It can be a very slow and tedious process.

Digitally voice recording an interview allows the interviewers to focus on the person being interviewed, not laboriously writing down everything said. That invariably enhances the quality of the interview.

From the perspective of demonstrating impartiality, DVR is a very effective accountability tool. It provides direct evidence of how thorough and objective an interview was, including how questions were asked, whether key issues were covered and, if so, in what depth. It may also provide evidence of whether undue pressure was brought to bear and the level of cooperation received.

We have heard arguments, though not from within CSC, that the presence of a DVR at an interview discourages interviewees from being candid - that they have a chilling effect. This is not the experience of any investigative agency that we are aware of, nor is it the author of this chapter's experience in the thousands of investigative interviews he has conducted or reviewed in the course of his career.

We also understand that there may be pushback from various parties, including unions. This is not at all unusual when DVR is introduced for the first time into investigations of this type. The solution is to educate and explain the advantages to all involved, including those being interviewed.

Finally, there may be concerns about confidentiality, particularly given provisions of the Access to Information Act and the Privacy Act.

CSC should find ways to resolve these potential barriers to digitally voice recording all interviews. Many other investigative bodies have faced similar challenges and have overcome them. Ultimately any real or perceived disadvantages of DVR are far outweighed by the advantages.

Not all interviews conducted by BOIs are digitally voice recorded. They should be. DVR is best investigative practice and an effective accountability tool.

All interviews conducted by a Board of Investigation during the course of its investigation should be digitally voice recorded.

An investigative agency may have gathered all relevant evidence thoroughly and fairly. It may have asked all the right questions of all the right people. It may have left no reasonable investigative stone unturned. However, if the analysis of that evidence is not objective, then it is all for nothing.

How the investigation report is written is key in this process. IIB has significantly revamped how it structures its reports in the last few years. It has adopted a 'top down' approach that focuses on facts rather than chronology or themes. This new approach to report writing is discussed in more detail in a later chapter.

We found that, for the most part, the BOI reports we reviewed were clear and focused. The feedback from staff we spoke to about the new format was generally positive. However, the OCI had some quite strong reservations, including what they perceived to be less information contained in the new style of report about sources of evidence. This makes it, in their view, more difficult to assess what any findings are based on. An OCI staff member told us that it will be interesting to note whether this new approach will lead to improvements by CSC decision makers in acknowledging and implementing recommendations.

As mentioned in the previous chapter, including information in a report about overall context can be very useful, hence the recommendation made in that chapter.

We reviewed the BOI reports through the following prism. Was there sufficient, reliable, relevant evidence for the BOI to reach the conclusions that it did? Was any evidence that did not support that conclusion addressed and did the BOI explain the reasons why they preferred other evidence? Were key issues dealt with in the report itself?

We think it is important to repeat at this - for the last time in this chapter - our caveat that we were not given sufficient time to review underlying investigative product or conduct in-depth interviews with BOI members about specific BOIs. Doing so would have likely given us much deeper insight in how evidence was assessed by a BOI.

Generally, we found that BOI findings were consistent with the evidence gathered and the rationale for reaching those findings was reasonably well explained. As noted above in the segment on issue identification, we found no evidence of deliberately sweeping issues under the carpet or intentionally pulling punches.

However, there were exceptions. We identified instances of where it was unclear whether compliance issues were being dealt with appropriately by BOIs, in some - though certainly not all - of the cases we reviewed. This was particularly so when multiple compliance issues were identified by a BOI. This may result in potential systemic issues - for example issues relating to training, supervision and culture - not being identified and dealt with, including those that may not have had a direct impact on a particular death but may be a factor in preventing future ones. This is discussed in greater depth in Chapter 4.

We also looked at how findings were dealt with, including whether recommendations reflected what the investigation had found and the use of non-compliance and supplementary findings categories. This is discussed to a degree in the previous chapter. It is also dealt with in detail in a later chapter, again with findings and recommendations.

Finally, we noted that in many BOI reports there is significant emphasis on the history of the inmate and his/her prior interactions with the prison system. While such information can be useful - and in many cases directly relevant to the death itself - the main focus of a BOI should be how and why the person died, including what if anything could have been done to prevent it and, most importantly, what steps should be taken to minimize the chances of the same thing happening again.

This is an important consideration from an impartiality perspective. If a BOI is going to scrutinize the prior history of the inmate, should it not also be doing the same for any CSC staff involved in the incident? For example, in cases where a Correctional Officer did not make sufficient efforts to determine if an inmate was alive in a cell during rounds, is there any evidence that is not the first time that this Correctional Officer has failed to do this? The purpose of exploring such investigative lines of inquiry from a BOI standpoint would not be disciplinary, nor is it intended to point fingers, rather it may be that there is, for example, a systemic issue with training and supervision that needs to be addressed. We noted that in some instances the same level of scrutiny that was applied to deceased inmates was not being applied to others who were involved, at least on the face of it. This imbalance may adversely impact perceptions of impartiality.

We noted inconsistencies in how BOIs presented information about what investigative steps they had taken to gather evidence. Some reports include a segment entitled Incident Investigation Process and Methodology (IIPM). Others did not. In BOIs that did not include such a segment, it was sometimes possible to reconstruct what steps the BOI had taken but that could be difficult and time consuming to do so.

The IIB Training Manual refers to a 'List of Documents reviewed by the Board'. It sets out a presumably hypothetical list of documentation such as Offender Management Systems items relating to the deceased, statements of individual Correctional Officers, relevant policies and so on. We did not find any such list in the BOIs we reviewed. Each BOI had a list entitled 'Policy Citations' as an appendix. These are policies that relate to the incident, but not a list of actual documents reviewed. We were advised that a list is prepared and is kept in the electronic file for each investigation. It does not appear in the BOI report. It should.

Most IIPMs were approximately a page in length. However, they generally included boilerplate paragraphs at the beginning (about the BOI process in general) and at the end (about protections under s13 of the Inquiries Act and a general statement about the purpose of BOI investigations). Setting out what the BOI had actually done was usually only a paragraph long. In one case, it was just 3 ½ lines.

In the BOIs that did include an IIPM, there was a broad degree of variation in the level of detail describing the steps that the BOI had taken to gather evidence. For example, some described efforts made by the BOI to speak to inmates and whether they were successful or not. Some were silent on that issue, even in cases where it was clear that efforts should have been made to at least contact inmates. In one case, the BOI noted specifically that inmate interviews done by staff prior to the BOIs arrival were sufficient for the BOI 'to complete the mandate'. In another case, the BOI noted it had made efforts to interview the deceased's Parole Officer but were ultimately unsuccessful - though, frustratingly, they did not explain why they were unable to. In some cases, there was no explanation as to why family members had not apparently been interviewed, such as the inmate suicide cases discussed above.

In other cases, the level of detail allowed us to conclude that the investigation was done thoroughly. For example, one BOI described how members had participated in an early morning cell count to fully understand what the process involved and what areas of a particular cell a flashlight illuminated. Excellent. However, that information wasn't easy to find as it was contained in the body of the report - there was no IIPM, unfortunately.

In respect of digitally stored evidence, one IIPM gave the time parameters of CCTV footage reviewed. Others just noted that they had reviewed CCTV. Some included detailed chronologies of events, often based on CCTV evidence. Most, but not all, identified the number of individuals interviewed and their positions. All referred to documents they had obtained, though there were varying degrees of detail about what those documents were.

We encourage BOIs to include as much detail as is reasonably possible in a dedicated section of the report describing in detail what they actually did to gather evidence. There are many reasons why. It is good investigative practice. It lets the reader know precisely what efforts have been made by the BOI to gather potentially relevant evidence, even if those efforts have ultimately been unsuccessful, such as in the case of uncooperative inmates. It renders the investigators accountable for what they have and have not done. Setting out these steps in detail gives the reader confidence that no reasonable investigative stone has been left unturned and therefore the reader can have confidence that any findings and recommendations made by the BOI are based on a suitably comprehensive fact-finding process. Most importantly, it hopefully gives the reader confidence in the overall independence and impartiality of that investigation.

The level of detail in the IIPM will vary from case to case, of course. Unless there is a reason not to, the IIPM should include:

• Any pre-Board activity completed by IIB, including any relevant EIA process
• When the BOI attended the Institution
• Who was interviewed, including the position of any CSC staff? This category should include inmates and anyone else who was spoken to by the BOI. Unless there is a compelling reason, it would not be necessary to include names.
• Efforts to interview any party who was not ultimately interviewed and the reason why
• What was done to identify, gather and review documentary evidence, such as policy documents, inmate files, security information and so on
• A list of those documents, possibly included as an appendix
• Any digitally stored evidence obtained, such as CCTV and telephone records
• Details of site visits and walk-throughs
• Consultation with other parties such as, for example, external experts and/or CSC Research Branch
• Liaison with external agencies such as police and/or Coroner/Medical Examiner
• Any other investigative steps taken
• Any impediments to gathering evidence, such as failure to cooperate or delays in obtaining information, and how they were dealt with.
• If applicable, any direction or input from CSC or other party re the investigation, other than that set out in the Convening Order and ToRs.

We also noted that the more recent the BOI, the more likely that an IIPM segment is included and that the level of detail set out in that segment has increased. We encourage IIB to continue that trend.

There is a lack of consistency when setting out what steps a BOI has taken to gather evidence. Some BOI reports include a segment setting out those steps, others do not.

In cases where a BOI report does include such a segment, there is sometimes an insufficient level of detail about investigative steps taken.

Failure to include this information can potentially undermine confidence in the independence and impartiality of the process.

All Board of Investigation reports should include a detailed segment on what the Board of Investigation did to gather evidence.

The overall CSC process for investigating non-natural deaths is not independent, but we found no evidence that BOIs do not function independently.

In respect of impartiality, we believe that IIB conducts non-natural death investigations reasonably thoroughly and objectively, and therefore, by extension, reasonably impartially, at least based on the content of the BOIs we reviewed. However, we identified several areas of concern. While they were not apparent in every BOI, they did appear often enough to warrant remedy.

In summary, we found no compelling evidence that there is a systemic bias in the way the current system conducts these investigations. Nor did we find sufficient evidence to recommend that the current system is so flawed from an independence and impartiality perspective that it should be completely revamped. Nevertheless, there are steps that CSC should take to increase confidence in how it investigates non-natural deaths. Those steps are set out in our recommendations.

Chapter 3

This chapter is divided into the Current Correctional Information Around the Disclosure of Board of Investigation Reports and our ideas regarding Steps Forward.

With respect to this Term of Reference, the Independent Review Committee considers that sharing the Board of Investigation reports to all Correctional Service Canada staff to be of the utmost importance for two main reasons. First, and most importantly, to prevent deaths in custody across the service. Second, to inform staff of what actually happened as many may have heard about the incident, but do not have access to the actual facts and information discovered during the investigation. This would educate staff on possible infrastructure, policy, training, or staff changes made within the institution or region, because of the incident. In addition to saving lives, this may improve processes, outline clearer policies, and reduce staff or inmate stress.

The Correctional Service Canada is mandated by the Corrections and Conditional Release Act to share Board of Investigation reports with the Office of the Correctional Investigator.^{Footnote 44} As we think of others who may have interest in the BOI report or a Lessons Learned document, we should consider internal government bodies with a particular role in correctional policy and oversight who have the ability to influence change (e.g. Public Safety Canada), coroners/medical examiners, stakeholders, academics, and those interested in corrections. Further significant groups include other inmates; the Next of Kin and family of the deceased inmate; and the victims of the inmate's offense(s), because of their connection and possible ongoing interest in the offender. We appreciate the inclusion of this ToR into our work as it outlines the willingness of Incident Investigations Branch to share the information gleaned from their investigations.

Reports

In September of 2018, the then Federal Minister for Public Safety, released, for the first time, a public mandate letter addressed to the new head of the Correctional Service Canada, Anne Kelly.^{Footnote 45} The letter encouraged her,

Making a mandate letter publicly accessible invites accountability on the progress of the directive. The Committee applauds and encourages the self-reflection and transparency of CSC thus far, with respect to the lessons learned and disclosure of incident information, and believes this gives the Commissioner direction to make as much information as possible, including investigative reports, available to the public.

In terms of CSC's Incident Investigations Branch, the IIB Participant's Training Manual^{Footnote 46} for investigators outlines three main procedures currently in place to share information found during the Board of Investigation reports with CSC staff. First, once the Board members travel to the site and collect their investigation evidence, a debrief is completed at the local level as the final step of the on-site phase where the incident occurred. The Participant's Training Manual states that the objective here is "to share the results of the investigation especially the Recommendations, Compliance Issues and Supplementary Findings that require corrective action by the site."

The Training Manual goes on to outline the second major step in sharing information from the investigation internally as follows, "debriefing the Regional Deputy Commissioner, Assistant Deputy Commissioner, Correctional Operations, Director General, IIB and relevant policy holders, should occur two weeks before the Board's report due date."

The third and final step in the process is when the final report is presented at the National Investigations Meeting (NIM), which generally occurs every three months, chaired by the Senior Deputy Commissioner, with membership including Assistant Commissioners, Regional Deputy Commissioners, relevant Sector heads, and the Director General of IIB.

Without a full staff disclosure of the BOI report, it is currently only this subset of the service who are permitted to view the findings and lessons learned.

During our review of the disclosure process for deaths in custody, the Committee also examined several Commissioner's Directives^{Footnote 47}, including CD 022 Media Relations, CD 530 Death of an Inmate: Notifications and Funerals, and CD 784 Victim Engagement mentioned in ToR #1 as well as the following with a particular focus on policy related to the sharing of information in cases of a death in custody:

• Commissioner's Directive 048 Info Sharing and Support Services Associated with Deaths states that:
  1. "The relevant Regional Deputy Commissioner, or delegate, as instructing officer on specific/regional cases, will immediately share the initial notification of an upcoming inquest/inquiry with the Senior Deputy Commissioner and liaise, as required, with the Regional Administrator, Communications and Executive Services, and the Deputy Commissioner for Women in the case of a woman offender.
  2. The Assistant Commissioner, Communications and Engagement, or delegate, will prepare media responses regarding CSC's involvement in a death investigation or an inquest/inquiry and CSC's response(s) to the recommendations when applicable, ensure liaison with the Minister's Office as necessary, and engage stakeholders, where applicable."

Commissioner's Directive 041 Incident Investigations, Section 53 outlines that the IIB is mandated to send out Significant Findings from the investigation, but not Incident Investigation reports. Section 53 Sharing of Significant Findings states, "The Incident Investigations Branch creates Significant Findings documents. They provide a general overview of the significant findings and recommendations, corrective measures and action plans, and best practices stemming from various investigations. They are distributed to all CSC staff, relevant unions and the chair of the National Executive Committee of the Citizen Advisory Committees, and are posted on the Hub." ("Hub" refers to CSC's internal network available to staff and unions.)

Correctional Service Canada has clear direction around notification, media responses, and sharing findings that will alert staff and the public that a Board of Investigation report will be forthcoming. While we did find compliance generally with the instructions issued in these Commissioner's Directives, the Committee found evidence that Section 53 of CD 041, which is a key vehicle for sharing information about the incident, is not being fulfilled in a consistent and complete manner. In interviewing CSC staff at various levels, many did not know about the recommendations or actions taken from Board of Investigation reports, or where they might go to find this information. IIB confirmed that the Chair of the National Executive Committee of the Citizen Advisory Committees would receive Significant Findings through the internet or social media but we could find no designated space on CSC's website or evidence of shared Significant Findings through these channels. Further, our review of CSC's intranet HUB, (available to CSC staff and unions) revealed that the last Lessons Learned document was about the Saskatchewan Penitentiary Riot in 2016, four years ago. Clearly, valuable findings are not being routinely shared to the CSC staff nationwide, a situation the Committee feels needs to be rectified for the reasons outlined at the start of this ToR.

Aside from our comprehensive review of written information, the Committee also held several interviews around information sharing and disclosure, and from these sources we identified valuable insights, which are discussed briefly below.

First, in considering the resources and time invested in producing a Board of Investigation report, the Committee takes the position that it is important to make best use of them. Indeed, there is a recognized wealth of information in these investigations that can be utilized in many areas of CSC. Each report contains insights, findings, and recommendations that are extremely valuable to both existing and new CSC staff, and they want to hear about it. Furthermore, not disclosing reports containing recommendations, best practices, policy gaps, compliance issues, and corrective actions hamper prevention efforts and opens the door to a repetition of issues. Nationally, CSC could benefit from both the positive actions outlined in a BOI that may ultimately save lives (e.g. an increase in cell searches after an initial overdose to prevent further overdoses), as well as to identify opportunities for improvement.

Second, during our interviews, the Committee discovered that different people have different needs when it comes to disclosure. Some of the management team, frontline staff, Next of Kin, families, and community organizations we spoke to wanted fact sheets/bulletins and quick information about an incident, while others felt full reports were more valuable. We quickly discovered that Correctional Service Canada has many communication platforms and broadcasting mechanisms both internally and externally to offer better disclosure, and in various forms, to appeal to different people's needs. These include Correctional Service Canada's website; social media channels (Facebook and Twitter); email broadcasts to stakeholders and staff; CSC's internal HUB; POpedia (CSC's training platform for Parole Officers); newsletters; information bulletins; and staff meetings.

Third, as mentioned in our earlier Terms of References above, transparency and accountability go hand in hand with both independence and impartiality. Public reporting is a vital component of both. In speaking with both CSC and the Office of the Correctional Investigator, we learned that there are some limitations to the disclosure of investigation information, including:

• Canada's Privacy Act whose purpose is to "protect the privacy of individuals with respect to personal information about themselves held by a government institution and that provide individuals with a right of access to that information".^{Footnote 48}
• Commissioner's Directive 701 Information Sharing outlines that information may be withheld if it jeopardizes the safety of individuals (e.g. threat to a private citizen), jeopardizes the security of a penitentiary (e.g. information describing a security system), or if it jeopardizes the conduct of any lawful investigation (e.g. information collection methods disclosed).^{Footnote 49}
• Protection of information and sensitivity to victims of crime, members of their family, or other connected individuals, and the impact that disclosing and identified facts may have on them.
• Protection of information and sensitivity to the Next of Kin, families, and friends of those in custody and the impact that disclosing and identified facts may have on them.
• Liabilities arising out the Service's operation or conduct of its staff. (IIB's Training Manual indicates, "Copies of IIB Investigation Files are frequently requested for litigation purposes^{Footnote 50})

The final report of the Office of the Correctional Investigator "An Investigation into the Preventable Death of Matthew Ryan Hines" illustrated that families of the deceased inmate have a particular need to be informed of the circumstances surrounding the death of an inmate, and CSC has a particular responsibility to properly inform them.^{Footnote 51} A concern, we heard several times in our interviews, is that when families are not notified immediately about a death, or not able to get answers about what happened, it leads to a perception of lack of empathy and compassion, and "mistrust of CSC". One CSC staff said, "We shouldn't lose the fact that this is someone's child. It takes so long to give them answers. The more you give them at the front-end the better. We need a high level of compassion too." As much information as possible needs to be communicated in a sensitive and timely manner immediately following a serious incident involving someone in CSC's care. The flow of meaningful information to families that can be given from the investigation facts, if they want it, needs to continue during the investigation.

Once the final Board of Investigation Report is complete, the information moves to a different category. Families used to have to make a request through the Access to Information and Privacy (ATIP) service, which can be a very lengthy process, in order to receive the final report, which was heavily redacted.^{Footnote 52} The Director General of the IIB now has the authority to share some of the facts from the report verbally with the Next of Kin, and the families can access the full report through the Regional Administrator, Communications and Executive Services (RACES) team, who will work with ATIP to provide it. There are two issues here. First, the onus should be on CSC to reach out to families who wish to be engaged and not on the families to continually call and check with IIB about when the report is available. Second, we understand, once the report is received, it is unfortunately still heavily redacted, and more needs to be done on this front.

During our investigation of international practices, the Committee was impressed to learn about the disclosure practices being used by the Prison and Probation Ombudsman in the United Kingdom concerning deaths in Her Majesty's Prison and Probation Service (HMPPS).

In the UK, the family is immediately notified when a death in custody occurs and a PPO Family Liaison Officer (FLO) is the main point of contact throughout the investigation. PPO works to ensure families are an integral part of the process by supporting them through the investigation process, offering them the chance to suggest issues for investigation, giving them the opportunity to ask questions and raise any concerns, and providing them with information on available counselling services. The FLO offers to meet with the family and the investigator and prepares the family for any aspect of the report which is likely to be surprising or distressing. (We find this last point of preparing the family particularly compassionate.)

The PPO sends a draft of the report to the bereaved family (as does the Prisoner Ombudsman of Northern Ireland (PONI)). It is identical to the copy provided to the coroner and the correctional staff of Her Majesty's Prison and Probation Service. The draft report is accompanied by supporting annexes which include a review of healthcare notes, transcripts of interviews, and other relevant documents. Items of security may be redacted in the draft. The coroner, correctional staff, and the bereaved family are all given the opportunity to comment on the factual accuracy of the draft report before the final version is issued.

After consideration of comments, the PPO produces a final report, which is again sent to the family (with what we understand to be minimal redactions), the coroner, and the service. (PONI sends a minimally redacted final report to families.)

The family is invited to a debriefing, if they wish. (Both PPO and PONI offer this.) Unless there are compelling circumstances, such as significant security considerations, the family is provided with exactly the same information as the prison service during the debriefing. (Again, the Prisoner Ombudsman of Northern Ireland debriefing is similar.)

The Prisons and Probation Ombudsman publishes an anonymized version of all fatality investigation reports (and any action plan that deals with the recommendations made by the PPO arising from the investigation), after the Coroner's inquest, online.^{Footnote 53} The names of the deceased appear in the report but other names, such as those of prison staff, are redacted.

PPO's website outlines that since 2012 they have been placing a greater emphasis on learning lessons from collective analysis of their investigations, in order to contribute to improvements in the services and to potentially help to avoid preventable deaths. Their impressive range of Learning Lessons Reports and thematic reports appear on their website.^{Footnote 54} These bulletins outline a certain topic (e.g. Self-inflicted Deaths Among Female Prisoners), then go on to give current statistics, rules and procedures around the topic, a few case studies as practical examples, and the lessons to be learned.

We appreciate that the legislative framework in Canada is different from the UK. Nevertheless, we encourage CSC to review the PPO process in depth, with a view to determining what CSC can adopt and adapt with respect to the public release of death investigation reports and other information provided to stakeholders, and especially to Next of Kin and the family of the deceased. The focus of the PPO review should be looking for ways to maximize the amount of information about a death that is provided, and to proactively and continually challenge any impediments to doing that. Applying these processes would mitigate the skepticism that inevitably arises when an entity investigates itself.

As mentioned previously, the IRC spent a great deal of time studying CSC's material around media relations and information sharing, and reviewed several types of communication platforms and broadcasting mechanisms within the service. From our research and interviews with several CSC staff regarding strategies for getting more information to CSC staff and the public, we identify several possible sharing strategies, as follows:

• Upcoming investigations broadcast on CSC's website and on social media channels
• Full BOI Reports, with Corrective Actions taken, shared publicly on CSC's website, preferably with full disclosure or, if necessary, with minimal redaction. (One redaction method may be to have items that must be anonymized, for privacy or security issues outlined above, appear in an annex in the BOI report. This annex can be removed when the report is uploaded to CSC's website.)
• Detailed Lesson Learned Fact Sheets/Bulletins disclosed on CSC's website for all staff and the public to readily access
• Mention made of newly uploaded Lesson Learned Fact Sheets/Bulletins and BOI reports in "This Week at CSC" (the service's staff email vehicle), in "This Just In" (the service's stakeholder email vehicle), and on social media channels
• New or relevant information gathered from Board of Investigation reports broadcast via Information Bulletins for staff, and to inmates as well via Peer Counselors and Inmate Committees, around suicide prevention tips, new drugs entering the institution, etc.
• Information about the process of sharing information gathered from BOIs written up in CSC's "Let's Talk" newsletter at least once each year for continual awareness.

Following the information garnered with respect to this area of our review, the Committee makes the following recommendations:

That investigation reports with corrective actions from non-natural deaths be made available on Correctional Service Canada's website, with minimal redactions, for staff and public awareness, similar to the Prison and Probation Ombudsman in the United Kingdom.

Correctional Service Canada should routinely develop and share Lessons Learned Fact Sheets/Bulletins from non-natural deaths on their website, similar to the Prison and Probation Ombudsman in the United Kingdom.

Incident Investigations Branch take steps, similar to the Prison and Probation Ombudsman in the United Kingdom, to further include families and Next of Kin into the investigation process. Offerings need to include the opportunity to suggest ideas for investigation; the sharing of draft reports for accuracy; a final report debriefing (with as much information as possible); and notification of the final report being made public.

Chapter 4

First, it is important to note in this Term of Reference, as outlined previously, that in reviewing the Board of Investigation reports given to us, that the Committee did not re-investigate the incidents in any way. We did not look at how they gathered evidence or who they interviewed to decide the underlying causes and systemic issues. We took the findings at face value and reviewed the underlying causes and their direct link to the recommendations.

Second, it is important to give some background on the BOI reporting structure and the parameters around Recommendations. There are several terms used in a BOI report that are outlined in the Incident Investigation Branch Participant Training Manual for investigators.^{Footnote 55} A few that are necessary to mention here, include:

• Supporting Facts: The evidence or the facts from the incident that demonstrate Compliance Issue(s), Policy Gap(s) and/or Underlying Issue(s).
• Compliance Issue: Evidence outlined in a Supporting Fact wherein a person did not act in conformity with a law, policy, or procedure that was directly related to the incident under investigation.
• Underlying Issue: A Supporting Fact that is not linked to policy.
• Findings: A summary of the Supporting Facts that contain an overarching global, major, or key conclusion to the Investigation Area.
• Significant Findings: Issues and/or recommendations from national investigations which could have a national impact on CSC.
• Supplementary Findings: Compliance Issue(s) that had no direct impact on the incident.
• Recommendations: An issue that is clearly identified in a Supporting Fact as a Policy Gap or an Underlying Issue that is directly related to the incident and directly linked to the Investigation Areas outlined in the Convening Order. There are further parameters around the context of a recommendation as outlined in the next few paragraphs.

Third, we need to introduce the difference between non-compliance and discipline. During the course of an investigation, in describing clearly what happened, Compliance Issues may be found that could be the result of a lack of knowledge, negligence, and/or lack of appropriate management or supervisory systems to deal with compliance. However, the Participant Training Manual for investigators outlines that the "IIB does not conduct disciplinary investigations, but a lessons learned model", and thus, "A recommendation cannot be made to support or to correct a Compliance Issue".^{Footnote 56} This is important to keep in mind, during this chapter. The Compliance Issue may create the need for a corrective or disciplinary action but these are processes separate from an incident investigation as we outline below in further detail.

Fourth, the focus of a Board of Investigation is on the incident and its prevention and not on anything that is extraneous to that. So, for example, a Compliance Issue that involves someone not filling out a form after a death has occurred, would not have any effect on the incident itself, and would be considered a Supplementary Finding in a BOI.

In approaching this Term of Reference, the Committee felt that when reviewing recommendations that address Underlying Causes, the focus needs to be on what happened; why; and what can we do to prevent it from reoccurring. (i.e. Were practices and policies followed? Was the staff at full capacity that day? Were there impediments to performance in or out of their control? Were there underlying issues that management, supervisors, or staff should have noted?).

When reviewing recommendations that address Systemic Improvements, the focus needs to be on ensuring CSC has everything in place to prevent and handle the incidents. (i.e. Are the policies and procedures appropriate or do we have identified gaps? Is the policy fit for its purpose? Do the staff have the resources, knowledge, and strategies in place to successfully handle situations that may arise? Is there an issue within the culture of the institution, or the morale of the staff, such as a tolerance to indifference or not enough focus and determination to ensure a sense of urgency in knowing and following policy when staff respond to the incident that CSC can learn from?).

The Committee is grateful that CSC's Investigations Branch has been willing to outline their training and investigative processes to the IRC and is actively engaged in improving their processes. IIB, in the last two years, has made changes to advance their reporting, and we find this a valuable step. IIB informed us that they have streamlined the long inmate histories, and worked to add more about the context and the environment of an institution. They have created reports that are more succinct and are now using the 'Top-Down Approach'. According to the IIB Participant Training Manual, this approach is "not a Chronological/sequential ordered approach with a beginning, middle and the conclusion at the end" but rather "gets to the point immediately. It allows the writer to rewrite the top of the story continually, keeping it up-to-date. It is evidence-based writing supported by information/facts".^{Footnote 57}

The reasoning behind the changes to the Investigation Report format, we were told by the IIB, were partially in response to the intent and recommendations of the 4th Independent Review Committee.^{Footnote 58} IIB took further action beyond this, they told us, to a more action-based approach with reports that are robust and analytical in nature. These changes assist CSC's internal stakeholders in taking some ownership of the findings and recommendations.

CSC's Research Branch have an ongoing project of collecting information on deaths in custody for their Annual Report, and we were told, in order to assist in their data collection, they have asked IIB to ensure that BOIs collect certain information to help around contextualization and trends (e.g. what medications an inmate was on, if mental health treatment was received, etc.) and lay it out in the BOIs so that it can be collected. While it is not the purpose of the Board of Investigative team to assist with the collection of research data, if it can be done with minimal additional effort, we see the overall value of this for CSC.

We were told by the Office of the Correctional Investigator, operating in its oversight capacity, that the new streamlined process, in their view, "diminishes the reader's ability to assess the underlying investigative process", and "it has become much more difficult to discern what sources led to which conclusion", which forces the external reviewer to source the information themselves, which can be challenging. To this first point, the recommendation by this IRC in ToR #2, to include a detailed segment on what the BOI did to gather evidence, should assist in understanding the underlying investigative process. To the second point here, if the new process leads to a better understanding of the underlying issues and how to address them, this may outweigh the disadvantage of the additional effort for the OCI to discern which sources led to which conclusion.

Out of the 25 reports reviewed by the IRC, those written later in our review period were generally clearer, with more relevant information about the inmate, and had focused areas (e.g. Key Issues of Non-Compliance and Underlying Issues) that added value. The Edmonton Institution Protected Status Inmate Report, in the new style of writing, while not a death in custody, is particularly noteworthy.

The recommendations in the BOI incidents the Independent Review Committee studied varied from robust and on-point, to less vigorous, to none at all; all of which could be appropriate, depending on the circumstances. We found variability in the reports which would be expected in any area of operations, and we are not in a position to explain the variability. No two incidents or BOI teams are exactly alike. Without re-investigating each incident, it is hard to know if the recommendation section should have been proposed differently. Nevertheless, we found that most of the recommendations in the Board of Investigation reports reviewed flowed well from the underlying and systemic issues noted. That being said, we offer a few additional observations.

a) The Opportunity to Raise Systemic Issues Should not be Missed

Each BOI has the opportunity to be used as a vehicle to continually raise systemic issues in and of itself. CSC does not need to wait for an Annual Report on Deaths in Custody or other special thematic reports from the Incident Investigations Branch, CSC Research Branch, or a response to the Office of the Correctional Investigator. One BOI report can lead to a recommendation that addresses a systemic issue with a far-reaching impact on CSC operations. The following case illustrates this point.

Several of the Board of Investigation reports that we reviewed raised the issue of lapses in communication that detrimentally affected the care of inmates. Security information not being relayed, staff leaving communication devices behind, and late reports are examples of these failures. The following BOI, a case of a suicide at a medium security institution, elevated these Compliance Issues to a robust Recommendation.

The case was a suicide by an inmate with a suicide/self injury history, physical health deterioration, nervousness, social withdrawal, and weight loss. He had admitted to manufacturing a noose and having difficulties coping in the last days of his life. The day before he committed suicide, his Parole Officer submitted a Mental Health Services Psychological referral by email, marked high importance, with a subject line of "Psychological Referral for ….". It was sent to the entire Mental Health Department (nine staff). The email outlined that the inmate had admitted to attempting suicide twice while on release and that his physical health concerns were an ongoing issue that were causing him mental health challenges on a daily basis and that he had commented that, "if he were to die today, it would not matter to him".

Unfortunately, the request form that accompanied the email, incorrectly listed a different inmate and the box marked urgent was unchecked. The Psychological Referral for this man was never actioned on that day or the next day, the day he committed suicide. The Board of Investigation found that errors notwithstanding, the referral, "could have, and should have, been actioned immediately as per the Integration Mental Health Guidelines".

There were not sufficient systems in place to ensure this didn't happen. The A/Chief of Mental Health Services said he never received the referral and, in fact, never saw it until the Investigation Board showed it to him. Three other members of the Mental Health Services email distribution list were aware of the referral email prior to the suicide but, as this fact was filtered through another staff person, the Board could not determine the identity of those that received the email. Several others recalled seeing the referral, but could not remember if this was before, or after, the suicide. None of the nine staff informed took action. Several matters within this situation were justifiably noted as Compliance Issues in the BOI and the lack of communication and integration of services as a policy gap.

At the end of this investigation, the Board made a recommendation to clearly address a systemic communication issue in caring for those in CSCs custody. It outlines that the:

"Assistant Commissioner, Correctional Operations and Programs in consultation with Assistant Commissioner Health Services consider, that in cases wherein the risk for an individual to commit suicide and/or self injurious behaviour is believed to be elevated to a high level of risk and/or imminent risk, that an inter/intra-multidisciplinary meeting with institutional/contract staff assigned to manage the specific case (and available/on site at the time) be mandated to occur immediately. The recommendation is consistent with the report of the Third Independent Review of Deaths in Custody document of October 2015 and consistent with findings of previous boards of investigation."

The result of the recommendation was the Case Management Bulletin titled "Managing Offenders Who Are At Risk Of Suicide and/or Engage In Self Injurious Behaviour", dated December 16, 2019, reminding case management staff of the timely and effective information sharing among front-line staff when dealing with offenders who present a risk of suicide and/or self injury. This Case Study shows an example where a BOI took the Compliance Issues and underlying issues to a well-considered BOI recommendation with implications across the service.

Further, two other valuable practices during this investigation are noted here. The investigators on this BOI reviewed the report of the 3^rd IRC, which outlined that communication is vital, and in fact a core element, of suicide prevention.^{Footnote 59} This investigation team also reviewed several past BOIs that reinforced effective communication for suicide prevention and gave site managers a clear plan to follow up on the recommendations. Reviewing other BOI and IIB reports that are relevant to underlying issues increases the value of the reporting, offers credibility to the investigation team's research, and identifies reoccurring issues. We were told by IIB that these documents, where relevant to the BOI, were supplied. Any materials the investigative team wants to see, in addition to the provided documents, can be requested as well. The IIB Participant Training Manual has an Annex that contains an "Investigation Documents List" but neither previous BOI reports nor Independent Review Committee reports appear there.^{Footnote 60} We would like to see these two items added to the IIB Training Manual so Board members know they exist and can ask for them.

That "Previous Relevant Board of Investigation Reports" and "Relevant Independent Review Committee Reports" be added to the Investigation Incident Branch's Participant Training Manual's "Investigation Documents List" Annex.

b) Accountability is a Focus

The Incident Investigations Branch states that its purpose is, "To ensure that the Correctional Service of Canada takes appropriate action following an incident. To ensure that the review and analysis of investigation reports influence organizational policy and practices where appropriate, and significant findings/recommendations from these reports are shared in order to prevent similar incidents from occurring in the future. To ensure that quality of care reviews are conducted when an inmate dies from natural causes in a CSC facility, excluding in Community Correctional Centres."^{Footnote 61} So, while the IIB conducts a lesson learned model and the focus is on what has transpired in an incident and what can be gleaned in order to prevent similar incidents from occurring in the future, there is an overarching issue of CSC's accountability in this area.

The Office of the Correctional Investigator^{Footnote 62} has criticized Correctional Service Canada for downplaying accountability, notably in their Annual Report from 2017-2018 about the Saskatchewan Penitentiary riot. Their recent 2018-2019 Annual Report, a special focus on the Matthew Hines case, contains a table of non-compliance around the Use of Force, and states that the 4th IRC Committee^{Footnote 63} indicated, following its review, that the Hines' Board of Investigation report and their 21 areas for improvement were "not commensurate with the totality and gravity of the findings." This is an important point to consider.

While a large majority of the BOIs that the committee reviewed appropriately addressed the underlying issues with suitable recommendations, there was one instance where in our view, the BOI clearly fell short in this regard. In addition, there were other cases where there were numerous compliance issues with no recommendations. We are not in a position to indicate what recommendations, if any, should have been made in these cases, as we did not reinvestigate each BOI in detail, but we are concerned that the numerous findings of non-compliance not be lost in the process. These cases illustrate the need to review reoccurring Compliance Issues and gather them across BOIs to conduct analyses to inform the organization of possible systemic issues. This would allow the organization to learn from the BOIs collectively and take appropriate action and, in this way, better meet its accountability in this critical area of non-natural deaths in custody.

This was a case of a suicide at a medium security institution. The inmate was found hanging from a ligature attached to an electrical conduit pipe on the ceiling of the cell. The officer who found him while conducting a security patrol, immediately radioed for assistance, and emergency medical attention was provided to no avail. The BOI made no recommendations and did not identify any areas for improvement: however, it made five supplementary findings.

The first supplementary finding noted that an earlier version (2012) of the National Cell Condition Checklist (CSC/SCC 1448) was used to examine the cell instead of the 2016 version of the form. The result of this was that it was not recorded whether there were suspension points in the cell, which the BOI noted was not consistent with CD 550 Inmate Accommodation (August 20, 2018). Consistent with identifying it as a supplementary finding, the BOI concluded that this error did not have an impact on the incident under investigation since all other cells on the range had similar suspension points. They further noted that a Correctional Manager reported that he would take immediate corrective action to replace the outdated form with the revised version.

Suspension points have been the subject of numerous investigations of suicides in the past. BOI reports have recommended the removal of suspension points in an effort to reduce suicides, and CSC has undertaken to remove suspension points from cells. Relegating this finding to a "Supplementary Finding" (a Compliance Issue that had no direct impact on the incident), did not give it the attention it required. Since there were no recommendations or areas for improvement flowing from this investigation there was no plan of action initiated by CSC. In our view, rather than dismissing this finding, it would have been appropriate to draw attention to it with a strong recommendation to reinforce earlier recommendations that all suspension points be removed from inmate's cells. Although we see this as a shortcoming in this specific BOI report, we were informed by IIB that it tracks suspension points and forwards the information to the technical services division when one is identified in a BOI.

As mentioned previously, CSC's Incident Investigations Branch has made strides in moving their investigative reporting forward, which is commendable. In the process of these changes though, there has also been a shift in the approach to Compliance Issues. Although it may seem that not filling out forms to ensure the sharing of information, or not carrying communications equipment, may be a stand-alone policy Compliance Issue, CSC is under obligation to prevent deaths in custody and this concept, in all cases, compliance has to remain in the forefront as the BOI does its work to maintain accountability.

In order to understand Compliance Issues, we need to consider the separate matter of Disciplinary action. While meting out discipline is not one of the purposes of a Board of Investigation, if the public, or families of the person who died, perceive that the disciplinary action determined is not commensurate with the seriousness of the behaviour outlined in the Board of Investigation reports, it is challenging.

In speaking with the Office of the Correctional Investigator, this point was mentioned with the example of the Edmonton Institution Protected Status Inmate Report, where we were told, there was no disciplinary action for four managers.

It needs to be noted that the Board of Investigation and the Disciplinary Investigation are two separate processes and it was not part of the mandate of this committee to review the Disciplinary Investigation process in any way. Nevertheless, given concerns that have been raised, CSC should consider examining the relationship between BOIs and the disciplinary process, perhaps as a term of reference for a future Independent Review Committee.

There were numerous Compliance Issues in the BOIs reviewed by the Committee. We applaud the thoroughness of the investigative teams in these BOIs in finding and reporting non-compliance. From our interviews with CSC staff, we learned that compliance used to be a large focus of the BOI and staff were fearful of repercussions in speaking directly. IIB told us that they are focusing more on the incident and policy related directly to the incident versus previous efforts of five or so years ago when the process was a "compliance" audit. They feel that now the focus is on action (inaction) that contributed to the event being investigated.

The IRC feels that, in some aspects, the current method of dealing with Compliance Issues is not as effective as it could be. The sheer number of Compliance Issues found in some BOIs was, on the face of it, concerning. In moving away from a focus on compliance toward Underlying Issues, Compliance Issues are not being adequately attended to in the reports. In a number of BOIs, we saw repeated issues of delays entering cell, staff not carrying communication equipment, rounds not being completed properly, delays in the communication of information, etc. and we asked ourselves if it was possible that these failures to comply with policy, in some cases, could have had a significant impact on the outcome? Is there a pattern and are BOIs using their opportunity to make recommendations around repeated and systemic issues? With the number of issues found, is the lack of compliance itself becoming a systemic issue? What are the learnings to be found to assist with these issues?

This was a case of an unknown cause of death in a maximum-security institution. There were 12 Compliance issues including the performance by a Correctional Officer who responded to an emergency cell call alarm, from another unit where he was serving meals, to find an inmate was lying face down on the floor of his cell. The Correctional Officer looked into the cell, walked toward the end of the range, turned and returned to the cell, and kicked the door a few times. He quickly moved to the control post and advised them of a medical emergency. He was not able to raise the alarm from the cell door because he left his radio elsewhere and immediately left the unit, during the crisis, to retrieve it.

Other Compliance issues include Correctional Officers not stopping at each cell to ensure the presence of a live breathing body, a delay of over four minutes in opening the cell door as they waited for staff to arrive, inter-cell visiting between inmates allowed the night before, an outdated policy bulletin being used (that concerned a psychological risk assessment which could have meant a potential transfer to lower security for this inmate), quarterly onsite medical simulations not occurring, the scene not being adequately controlled, etc. The BOI had three underlying issues and no recommendations. There was no information on the reasoning behind the lack of recommendations.

This was a case of an overdose where they found an inmate locked in a minimum-security house bathroom. This BOI had identified nine Compliance Issues including Correctional Officers failing to recognize an overdose, staff arriving on the scene with no medical equipment (even though the call for assistance mentioned "medical distress"), urinalysis not administered following Escorted Temporary Absences, a Risk Assessment not done, house searches not done, search activities not recorded, quarterly medical simulations not done. The BOI had five underlying issues, one of which was a 10-minute time delay in staff arriving on the scene with an Automated External Defibrillator or Narcan, and again, no recommendations. Again, there was no information on the reasoning behind the lack of recommendations.

This was a case of a man found hanging in the cell of a medium-security institution. Once again, the Correctional Officer had to leave the scene to summon for help as he did not have a radio. This time-critical issue of not carrying a communication device "had already been addressed by the Institution at the time that the Board was onsite and therefore, was included as a Supplementary Finding" (a Compliance Issue Not having a direct impact on the incident).

Further information in the BOI outlines that there was a delay of five minutes in calling for an ambulance, "due to the Correctional Manager's involvement in another pressing issue". On the face of it, it is disturbing to note that another issue may be more pressing than a man hanging in his cell and the BOI does not identify what the more pressing matter is nor why a manager was needed to make the 911 call. Even though the investigators outline in the BOI that they find this delay to be "excessive and non-compliant with policy", it was also downgraded to a Supplementary Finding. When the paramedics arrive, they continue to perform CPR on the inmate and transport him to the hospital, so, in not knowing whether, or not, the paramedics would have been able to sustain life in those five minutes, it is unclear how not calling 911 in a timely manner is considered a Compliance Issue not having a direct impact on the incident.

Again, there were no recommendations that addressed these points and no reasoning as to why. The one recommendation that followed from this BOI concerned the issue of inmate assessments by an Elder.

Compliance issues indicate that there was a breach of a policy and/or Commissioner's Directive. While it is understood that CSC is a large organization with thousands of employees and errors during a death incident may happen, when there are a significant number of Compliance Issues in one investigation, confidence and trust in CSC is diminished. When the issues are repeated, in different regions across the service, this confidence and the perception that CSC is contributing to the safety of offenders is reduced further.

In each of the last four Cases noted above, there are numerous, repeated, Compliance Issues with no Recommendations to deal with them and, by not addressing these Compliance Issues further, it may create a sense that CSC is missing an opportunity to learn fully from these incidents and take appropriate actions. The purpose of a BOI is to focus on lessons learned and a prevention of further or repeated incidents, rather than to lay blame on individual staff.

With this in mind, we would like to see a change to the current IIB policy of, "A recommendation cannot be made to support or to correct a Compliance Issue". In areas where there are repeated Compliance Issues and an opportunity for site, regional, or national learning exists, we would like to see recommendations that include CSC's Mission, Values, and Ethics^{Footnote 64}; that call for an internal audit of the site of the investigation in a particular area; or that are the focus of additional training. (This builds on the preceding recommendation to review previous BOI and IRC reports that may be relevant in determining reoccurring issues.) In our research, we found that The Prisoner Ombudsman of Northern Ireland makes recommendations on non-compliance with a policy issue related to a death and has a separate Prison Service Disciplinary process.

As mentioned in the chapter (above) on ToR #1, CSC Values and Ethics offers ad-hoc sessions to meet team specific needs/requests. CSC spends a great deal of time developing and updating policies in encompassing detail to ensure an effectively run correctional organization. They have decided, in these policies, what they consider needs to happen in order for a staff person to do their job with a high quality of work, professionalism, commitment, and ethical behaviour. We learned of scenarios being used in staff recruitment, refresher training, and, in CSC's Guideline entitled, "Response to Medical Emergencies", (Section 800-4, under Responsibilities and Procedures), that "there are quarterly on-site simulations of medical emergencies that allow staff to practice and remain current in their skills".^{Footnote 65} We understand this practice of using reporting findings in specific cases being incorporated in recruit training is used by the Prisoner Ombudsman of Northern Ireland and appreciate that the use of these learning opportunities, to ensure the preservation of life, is of the utmost importance.

The following recommendations are meant to reduce Compliance Issues and strengthen CSC learnings from the numerous Compliance Issues identified. They allow the current Discipline process to remain intact to deal with individual non-compliance.

That Boards of Investigation be allowed to make Recommendations, to correct reoccurring Compliance Issues, to advance Correctional Service Canada learnings. When a Board of Investigation identifies reoccurring issues, then the Board of Investigation should make a recommendation to address this, or explain why they have not done so.

That Correctional Service Canada incorporate scenarios and lessons learned into staff training at all levels from reoccurring Compliance Issues that are noted by Boards of Investigation.

When Compliance Issues are moved to Supplementary Findings, they are determined to have had no direct impact on the incident and therefore of less importance. They are no longer tracked in terms of follow-up or corrective measures by IIB. The Investigations Training Manual on the criteria to make this important distinction on whether to move a Compliance Issue from the main focus of the report to a Supplementary Finding says, "Supplementary Findings are Compliance Issue(s) identified by the BOI that target the areas that were specified in the Investigation Areas as listed in the Convening Order but had no direct impact on the incident, corrective measures were already taken, there are no National implications and it is not identified as a recurrent theme."^{Footnote 66} The judgement on whether a Compliance Issue "had no direct impact on the incident" is left to the Board of Investigation, and based on some of the reports we reviewed, we see the need for guidance to assist BOI teams in making this decision.

That the Incident Investigations Branch work to create further criteria to define the wording "had no direct impact on the incident", as it relates to Compliance Issues.

If a Compliance Issue is perceived to have an impact on the death by the reader of a BOI report, it can be challenging to understand why it was moved to a Supplementary Finding. This is the situation in the Cases above, and in particular in the case where there was a delay in calling the paramedics. The reader is left wondering if placing this call earlier may, or may not, have made an impact on the inmate's death. In circumstances such as these, it is important to articulate exactly why a Compliance Issue is considered to have no direct impact on the death and is therefore moved to a Supplementary Finding.

If a Board of Investigation moves a Compliance Issue to a Supplementary Finding, it should outline the reason for the decision.

The Office of the Correctional Investigator has long-noted Compliance Issues. The noteworthy 2011-2012 Annual Report contains several recommendations to CSC on compliance around suspension points, emergency response protocols, use of force, etc.^{Footnote 67} In their "Investigation into the Preventable Death of Matthew Ryan Hines - Final Report February 15, 2017" it states, "Given that CSC investigates itself largely on the basis of compliance with policy and procedure rather than accountability, most Boards of Investigation do not issue recommendations of national significance. Consequently, at the site level, the Office sees the same mistakes repeated over and over again." This is an important issue we consider with our recommendations below.

Repeated lack of compliance begins to move toward a systemic issue and serves to reduce the trust placed in CSC as a whole, and in the IIB to conduct independent and impartial investigations. As one interviewee told us, "At some point you need to look at, 'Why is my staff not complying?' Is it the culture, indifference, labour relations, lack of compassion, etc.? This needs to be addressed to achieve CSC's goals." In the Case examined more fully in ToR #5, a nurse, who realizes the inmate before her is deceased, and further CPR is futile, did not feel comfortable in pronouncing his death due to "a culture of fear in CSC". A second nurse at the incident confirms that there was "fear of reprisal", if nurses pronounced death at that institution. Serious comments like this bear further investigation.

The 2nd IRC report, in 2012, also looked at this issue of compliance.^{Footnote 68} IRC #2 discussed a Systems approach to death investigations, outlining that,

"If a human error is discovered as part of an investigation, the investigation should learn more about why that error occurred. Similarly, if policy was not complied with, the investigation should try to understand why the policy was not adhered to. The answers to those why questions will help identify the systemic errors that can be remedied through recommendations that are more effective than policy and education alone."

The 2nd IRC goes on to recommend,

"A more in-depth exploration of organizational gaps in the service is required by exploring why compliance issues occur, and systems, and environmental factors that lead to human errors, when they occur."

and, on balance, a further recommendation that CSC,

"Implement a new mandatory section in all BOI reports, outlining what went well in the management of the offender and the response to the incident."

This 5th IRC concurs with these recommendations. While we did note items that went well in the text of the BOIs reviewed (for example, a Best Practice of leaving a smoking smudge shell in front of a deceased inmate's cell), we agree that a dedicated section would outline positive strategies for other sites to use and advance learnings.

Our final recommendation moves to further reduce Compliance Issues within CSC on a national level.

That the Incident Investigations Branch develop a thematic report focused on Board of Investigation reports into deaths in custody to determine if there exist any patterns and/or potential systemic issues in relation to compliance. Based on the outcome of this review, it may be appropriate to include this topic as a Term of Reference for a future Independent Review Committee.

Underlying causes and recommendations need to be validated in that there needs to be an understanding of what the standards are so that accountability of being held to them is obvious. BOI information was not provided in some cases with no information outlined as to why it was omitted. An example of this, as noted above, is if reoccurring or systemic Compliance Issues, that may potentially be related to a death, cannot be advanced forward to a recommendation, the reasoning behind this should be stated.

In the BOIs reviewed, references to Commissioner's Directives and Standing Orders^{Footnote 69} were noted in the "Policy and Reference Citations" section of the BOIs and we found this practice ensures thoroughness and assists in the credibility and impartiality of the investigation We did find one reoccurring issue however. One of the particular questions that arose several times as we reviewed the 25 reports within our purview was, "How many staff must be present to open a cell door when you suspect someone inside is in distress or see someone hanging"? During the reviewed incidents, some staff waited until two people were in front of the cell, some waited for four staff, and one incident allowed two inmates to rush in first. There may be valid reasons for a waiting, or moving ahead, at such a critical time, but this should be outlined in the BOI. Without knowing the Policy or Standing Order around a decision, how can an investigation team (or someone reading their report) understand the underlying causes and the appropriate ensuing recommendation?

Having BOI Terms of Reference, and thus findings and recommendations, tied to CSC's Mission, Values, and Ethics, as recommended in the ToR #1 chapter of this report, will also assist in knowing the standards that must be held to.

c) Other ToR Recommendations Noted

Recommendations made elsewhere in this report will also will address the issue of accountability, in actuality or in perception, as they may strengthen in determining the underlying causes of an investigation. Examples include:

• In ToR #2, our recommendation to allow ToRs to be developed mid-investigation, if the evidence warrants it.
• In ToR #2, we recommend IIB staff should be dispatched to the place where a death occurred as soon as possible, securing evidence, holding immediate interviews and include inmates, families, and those who may have further information.
• In ToR #2, we recommend that all BOI reports should include a detailed segment on what the BOI did to gather evidence.
• In ToR #2, we recommend that CSC create a formal training programme for Community Board Members on how to plan and conduct investigations, how to conduct interviews, the challenges of systemic investigations, CSC's Mission, Values, and Ethics, and the vital role of a CBM challenging and questioning.
• In ToR #3, we recommend that CSC follow PPO's method of publicly disclosing their BOI reports (with full recommendations and corrective actions).

a) SMART Method

CSC uses both the SMART and Hierarchy of Effectiveness concepts when responding to recommendations.^{Footnote 70} The SMART concept, in CSC's Guidance Tools are defined as:

• Specific (What exactly are you trying to correct/improve?)
• Measurable (Will you know if the action(s) has/have been implemented and if the desired outcome has been achieved?)
• Accountable (Determine a lead for the proposed action(s))
• Realistic (Can it be done?)
• Timely (Break the job down and assign a reasonable time period for completion) In some of the BOIs reviewed, each SMART concept was broken out and addressed specifically, and in others it was addressed as in paragraph manner with the SMART heading.

The Hierarchy of Effectiveness, identified in the Guidance Tools, relies on the concepts of sensitivity to operations; preoccupation with failure; deference to expertise; commitment to resilience; and reluctance to simplify.

When they were addressed in the corrective actions following a recommendation, they was found to be valuable. Inasmuch as possible, knowing that the corrective actions will be take this form, the BOI recommendations should be written with these concepts in mind.

b) The Value of Varied Thinking and External Eyes

Recommendations do not always have to be directly causal to the death of an inmate, but may lead to other ideas that may be advantageous during the death process. The Committee heard, during one of our interviews, about how one of the investigations recommended that CSC consider "codes" to identify the condition of an inmate in hospital (e.g. red/black) to indicate the seriousness of the situation. Simple ideas used elsewhere, both in correctional sites and in our communities, and brought into CSC recommendations can lead to important changes. This speaks to the value of thinking outside the box and bringing Community Board Members, who may have different perspectives taken from community situations, onto an Investigation team and into valuable recommendations.

c) The Value of Including Supplementary Information

As noted in previous ToRs, IIB has instituted an Enhanced Investigation Analysis approach to investigating fatal and non-fatal overdose incidents using focus groups, interviews, and questionnaires centred around the four pillars (prevention, treatment, harm reduction and enforcement strategies) of addressing problematic substance. We were pleased to see this information used in our BOI review but would have appreciated the survey being attached to the report as an annex. This, and any other supplementary information found during a Board of Investigation and attached to the report, assists with the clarity of the underlying causes and recommendations.

In summary, this Committee believes that, for the most part, the Board of Investigation reports reviewed developed appropriate recommendations from their evidence. We described one case where, in our view, the BOI fell short in making a suitable recommendation to address the underlying causes. We found some other cases that raised concerns about how compliance issues are addressed, and what might be done to capture recommendations from recurring compliance issues, elucidate patterns from the findings of BOIs and, thereby better extract the lessons learned from these investigations.

Chapter 5

The Committee was given the BOI reports for 25 cases of non-natural deaths in custody that occurred during the period from April 1, 2017 and March 31, 2019. In addition, we received the report of the "Board of Investigation into the Riot, Related Death of an Inmate and Serious Bodily Injuries Sustained by Multiple Inmates at Saskatchewan Penitentiary (Medium Security) on December 14, 2016."

Our main conclusion from our review of all these reports is that the action plans initiated by CSC in response to the recommendations of the BOIs were generally appropriate and adequate. However, there were some cases where we concluded that the action plans were not adequate and, in our view, failed to address the underlying issues raised by the investigation. Four cases are described below. While we hope that our commentary on these cases will be helpful in the form of "lessons learned", we are not making any recommendations in this section of the report.

The first case was a suicide in the administrative segregation unit of a maximum-security penitentiary. It was a complicated case, and the BOI did an excellent job of elucidating the history and characteristics of this individual, his trajectory through the penitentiary system and the events that led to his death. Briefly, he was a unilingual francophone who identified as Metis, with a complex psychological profile that included a serious history of self-mutilation, attempted suicides, substance abuse, antisocial and borderline personality disorder, and periods of paranoia (exacerbated by intoxication and sometimes rising to the level of conspiracy). Furthermore, he had little insight into his mental health issues, and was generally guarded toward mental health professionals. Among the BOI reports that we reviewed, this one was exemplary in its attention to context and detail.

Over the years of his incarceration, he had periodic stays at mental health centres where he functioned better. Prior to his inter-regional transfer, he had been in a medium secure institution when, highly intoxicated and incoherent, he provoked other inmates and was aggressive and threatening toward staff. This resulted in a reclassification to maximum security. The BOI reviewed this incident in some detail in their report, and noted clear evidence of confused, disorganized and delusion thinking throughout the incident. They concluded that his mental health issues were not given sufficient weight in the reclassification process, and that the incident was assessed as it would be for an individual in full possession of his faculties, which was demonstrably not the case.

The BOI acknowledged that this individual was difficult and challenging to manage; however, they noted that mitigating factors were not adequately considered in arriving at the reclassification decision. Since he was deemed to be unsuitable for transfer to either of the maximum-security institutions in the region, one because he was unable to integrate and the other because he had identified incompatibles (i.e. individuals with whom he had previous dealings that would raise the risk for violence), he was transferred to a maximum-security institution in a different region.

The inmate objected to the transfer and later committed suicide. The BOI noted that the institution that he was to be transferred to would not be able to provide any services in his mother tongue and language of choice, so that none of his needs - whether they be Aboriginal spirituality, mental health or correctional programming - could be met. They concluded that the only factor considered in the inter-regional transfer decision was the need to end his administrative segregation in that institution, and that the process was driven by the imperative to restrict the use of administrative segregation. On the last point, the BOI found that there were a few alternatives that had not been fully explored, including a request to the region of origin to return him there (where he would have had access to services in his functional language). The inmate had expressed an interest in being transferred to an Intermediate Care Unit which was full at the time, and not commensurate with his assessed level of mental health need (i.e. moderate); an inter-regional transfer was viewed as a quicker option. The BOI also noted that consideration could have been given to a reclassification to medium security, given that it had been six months since the incident at the medium secure institution and his behaviour had not been problematic since that time.

From their interviews with institutional staff, the BOI learned that all staff considered the proposed inter-regional transfer to be inadequate and that it should not have been the preferred option. Yet, no one opposed the transfer. Once the decision was made for an inter-regional transfer, the regional protocol set out steps, timelines and each person's role in the transfer process. The inter-regional transfer coordinator was responsible for choosing the institution and this was done on the basis of cell availability. The BOI found that this procedure was inadequate because it did not provide sufficient consideration of the needs of the inmate who was being involuntarily transferred. In this case, it was evident that none of his mental health, program and case management needs could be met with this transfer. Several staff indicated that they believed that he would be placed in segregation upon his arrival at the new institution. In effect, the net result would have been to shift his segregation status from one institution to another.

The BOI also reported that the inmate had filed a grievance objecting to his transfer which was upheld posthumously by the Senior Deputy Commissioner. The response to the grievance noted that the institution had not made sufficient efforts to find an alternative placement to the one foreseen, considering that the proposed transfer would not allow the inmate to access services in his language of choice.

The BOI found that were no immediate precursors of the incident. He had been regularly questioned by the mental health team about possible thoughts of self-injury and denied these. The BOI had access to letters found in his cell which indicated he viewed his imminent transfer as a conspiracy against him, but he was able to mask this in his presentation to staff and therefore this could not have been known prior to the incident. The BOI concluded that the triggering event was the imminent transfer combined with the persecution scenario, which was unknown staff, such that no measures could have been taken to prevent the suicide.

The BOI identified 21 areas for improvement and made eight recommendations, which extensively cover a broad range of issues, including the transfer decision, mental health services, Aboriginal services, official languages, cell window coverings, suspension points and notification of Next of Kin. This Committee is focused on the central issues - the proposed involuntary inter-regional transfer, the consideration of alternatives and how the final decision was made. These issues were the subject of the first three recommendations and numerous ancillary Areas for Improvement. The discussion below examines Recommendation 1, which though tentatively worded, addressed the fundamental issue, i.e., the process that led to the inter-regional transfer decision that ignored the needs of the inmate in this case.

Recommendation #1 states:

CSC supported Recommendation 1 (and the ancillary Areas for Improvement 1 to 6) and provided the following response:

All of the essential services are taken into consideration, including Aboriginal social history, the inmate's language and health factors, in any decision concerning the transfer of inmates, which also means when the transfer is to alleviate the inmate's segregation status.

Prior to transferring an inmate from administrative segregation to another institution, Correctional Service Canada (CSC) must consider many factors including official language, culture, health and family. All of this information should be documented in the assessment for decision through the transfer process as prescribed in Commissioner's Directive 709-1, Administrative Segregation (August 1, 2017) and in conjunction with Commissioner's Directive 710-2-3 Inmate Transfer Processes (January 15, 2018). All efforts are made to avoid transferring an inmate outside of region especially when an inmate's access to interventions, programs or services, i.e., health, Elder services in his official language, or to his community, will be limited or his access impacted. Nevertheless, despite best efforts, in some circumstances a decision has to be made to transfer outside the region when an inmate cannot be safely managed outside of the administrative segregation nor transferred to any regional institutions. Having said that, with the strengthened oversight to segregation by the National Long-Term Segregation Review committee, there are regular, national conversations about complex cases.

In addition, in 2017/18, Parole Officers received a one-day in-class training on Aboriginal Social History (ASH) Considerations through the Parole Officer Continuous Development training. CSC will ensure to include this case as a training component in the Parole Officer Induction Training and/or in the Parole Officer Continuous Development Training.

The Security division will examine CD 709-1 at the next review to ensure that the requirement to consider access to interventions, programs or services is well documented, an analysis of this information is noted, and the rationale is recorded as part of the decision making process for a transfer out of administrative segregation. The expected promulgation should take place in 2019 and an update will be shared in February 2019 to confirm the next steps.

The first paragraph of the response states that the various needs of the inmate are taken into consideration in transfer decisions, but it was clear from the findings of the BOI this is not what happened in this case. The second paragraph states that "all efforts" are made to avoid an inter-regional transfer where the inmate's access to services will be limited or his access to them impacted. Again, the findings of the investigation showed that staff were well aware that virtually all of the services and programs for this inmate would be hugely impacted, and yet the transfer decision was about to be affected at the time that the inmate took his life.

The second paragraph goes on to state that, "despite best efforts" to avoid a transfer, a decision for an inter-regional transfer has to be made when an inmate cannot be safely managed outside of administrative segregation or transferred to another institution. However, in this case, the BOI concluded that the efforts made to find alternatives to the proposed transfer were not satisfactory (e.g., consideration for a transfer to the region of origin where he would have received services in his functional language was not considered); the process unfolded inexorably while the institutional staff believed that it was the wrong decision.

The third paragraph concerns training regarding Aboriginal Social History, which is a positive initiative but does not address the central issue regarding the involuntary inter-regional transfer decision. The fourth paragraph indicates that the Security Division will examine CD 709-1 Administrative Segregation to ensure that important information is included and the rationale for the decision document. Yet, the findings of this investigation indicated that the critical issue was the disconnect between the information documented by the Case Management Team and the decision made at Regional Headquarters to satisfy an organizational imperative to move inmates out of segregation to available cells elsewhere. The fifth paragraph refers to the use of this case to inform the National Indigenous Plan as well as to be included in scenarios in staff training for the Suicide Prevention Strategy; however, based on CSC's response to the recommendations the lessons learned from this case were not as robust as they might have been, as outlined in our discussion below.

We consider the findings from this investigation to be extremely concerning. Due to an organizational imperative to reduce the use of segregation, a decision was made to transfer an inmate with significant mental health issues to an institution on the other side of the country where he would be unable to access services in his functional language, and be further away from his family. Institutional staff knew that this was not the right decision but felt unable to speak out against it. Put simply, the process resulted in a seriously wrong decision.

A major purpose of conducting BOIs is to allow the organization to learn from the incident and make improvements in their policies and procedures so that an incident of this kind will not recur. In our view, CSC missed an important opportunity to improve their policies and procedures in response to the recommendations of this BOI. Rather than maintaining that the existing structure was adequate to address the needs of inmates in these circumstances, the appropriate response would have committed CSC to undertake a fundamental review of the inter-regional transfer process to ensure that the needs of the inmate are fully considered and given significant weight before any decision is taken, that the Case Management Team has a more prominent role in the decision-making process, and that all alternatives to an involuntary transfer are explored and documented before a decision is made.

The finding that staff reported to the BOI that the proposed transfer was not the right option but did not speak up to oppose it, merits additional comment. CSC is an organization with a Mission and Core Values. As such, staff members should be encouraged to speak out when they see something that offends these values, and managers at all levels must support staff in doing so. Otherwise, the risk is that staff become cynical and the organization loses ground in its efforts to operate in accordance with its Mission and Values. The overriding lesson to be learned in this incident is that a specific operational imperative must never be paramount, and should not override the Mission and Values of the organization.

The second case was an incident involving multiple overdoses over a period of about three weeks, one of which resulted in death. Illicit substances containing fentanyl had been introduced into the institution. The report included an excellent context section that described the trends in the use of opioids and related deaths in communities across Canada, as well as the particular impacts in CSC institutions. An Enhanced Investigation Analysis (described earlier in this report) had been done by a separate team of staff from the IIB prior to the members of the BOI arriving on site to conduct the investigation. The results of the discussions that the pre-investigation team had with staff members and inmates are presented in the BOI report.

The BOI report contains 11 recommendations that cover a broad range of areas, including management strategies to address an outbreak of lethal drugs, drug detection, handling of highly toxic substances, security intelligence, administration of Narcan, perimeter lighting and criminal networks. Two particular recommendations raised questions for this committee in the context of ToR #5, and these are discussed below.

One of the findings of the BOI was that a substantial number of operational staff (non-health services staff) interviewed viewed the immediate administration of nasal Narcan as the most important element of a response to an unresponsive inmate to preserve life and did not view the administration of a nasal dose of Narcan as a component of an integrated First Aid response as outlined in the response protocol. In addition, many non-health services staff reported to the BOI that further Narcan training (i.e. online video or revised First Aid training) would be helpful to increase their confidence in their capacity to respond to an unresponsive inmate.

This led to a recommendation that the Assistant Commissioner, Human Resource Management consider enhancements to training on the administration of nasal Narcan through an online training video or revision of First Aid training contracts. This recommendation was not supported. The CSC response to the recommendation describes several existing mandatory training requirements, numerous training/instructional materials, and the established policy requiring non-health services staff to initiate CPR/first aid where physically feasible and, in the case of suspected opioid overdose, administer Narcan. The committee appreciates that the training mechanisms and the appropriate policy were already in place, and one might reasonably conclude that enhancements to current training protocols were not necessary as there were training requirements and numerous materials already available. Furthermore, we noted that the BOI report indicated that following the first two interrupted overdoses the Acting Warden sent an all-staff e-mail as a reminder of the signs and symptoms of overdose, including the protocol for use of Narcan by non-health services staff, with step-by-step instructions that specifically reference applying First Aid measures prior to the administration of Narcan. Nevertheless, considering that many non-health services staff indicated that they did not feel confident in carrying out a life saving measure as important as the administration of Narcan, we would have expected the response to be to offer refresher materials/training at this site on its use.

Another area of this investigation focused on the various avenues of introduction of illicit substances (i.e. inmates and staff through the Principal Entrance, the Sally Port, Visits and Correspondence, Admissions and Discharge, Throw Overs and Drone Activity) and unpredictable changes in strategy by criminal networks. In this regard, the BOI observed that the Principal Entrance, considering the high volume of traffic of visitors and staff, represented a logistical challenge and a potential vulnerability in relation to the introduction of illicit substances to the institution where staff come under the influence of criminal organizations. Specifically, the BOI noted that since staff are involved in searching other staff with whom they have or may have a private/personal relationship, the circumstances exist for a real and/or a perceived conflict of interest. Further, this concern regarding staff searches was buttressed by the pre-Board survey findings that included widespread reports that staff were not subject to a consistent quality of searching at the Principal Entrance. The BOI suggested that CSC consider mitigation strategies to address this issue, and mentioned as an example using a random selection mat (used in airports) to identify individuals at random for a full frisk or body scan, thereby eliminating the discretion of staff in regards to rigorously searching a colleague.

This led the BOI to make a recommendation to revisit the practice of the quarterly staff searches and consider further options to enhance the effectiveness of the searching of staff at the Principal Entrance, with the intention of mitigating the efforts of criminal organizations to introduce illicit substances into institutions and reducing or eliminating the real or perceived conflict of interest of staff searching staff with whom they have a personal relationship. This recommendation was not supported. The CSC response correctly notes that "quarterly" refers to the national reporting of staff searches and not the searches themselves, and that CD 566-8 requires that staff be searched every time they enter or leave an institution. The CSC response also states:

In our view, this response failed to address the underlying issue of the perceived conflict of interest surrounding the searching of staff by staff and the widespread reports that there was a lack of consistent quality of searching of staff at the Principal Entrance. This underlying issue was the basis for the BOI recommendation that CSC consider further options to enhance the effectiveness of the searching of staff at the Principal Gate. In its response, CSC missed an opportunity to undertake a review of possible strategies/enhancements to address this important issue.

The third case involved an inmate who was found unresponsive in his cell and died shortly thereafter, despite application of first aid measures and administration of Narcan. The cause of death remained unknown at the time of the writing of the report. Areas for improvement noted in the BOI report covered issues related to quality of security patrols, calling for assistance, availability of cell keys, emergency equipment, emergency medical directives, and cell searches.

The BOI made two recommendations. The first recommendation was that the institution update its emergency response guide "to ensure that it is consistent with national direction that nurses from all Units responding to medical emergency bring emergency medical equipment with them to reduce the possible delay in provision of appropriate assessment and treatment." The second recommendation was that "the Regional Director, Health Services develop a plan to explore and address the Nurses' reluctance to practice to full scope with respect to pronouncing death" at the institution, which was within their parameters of their position.

In response to the first recommendation the institution replaced the existing guide with a new document that sets out the protocols for nurses in emergency situations, and clearly states that the nurse will bring emergency medical equipment with them when responding to an emergency. The action taken in response to the second recommendation was that the new protocol document was sent to unit Managers and Nursing staff via email and the A/Chief of Health services met with staff to discuss and review the pronouncement of death and the new protocol.

It is important to note that the impetus for the second recommendation was a BOI finding that a nurse, whose assessment was that the inmate was deceased and further CPR was futile, did not feel comfortable pronouncing death even though it was within the scope of nursing practice in the relevant jurisdiction and was provided for in relevant legislation and CSC policy. She described the context of her discomfort as "a culture of fear in CSC." In addition, another nurse who responded to the incident informed the BOI that there was "fear of reprisal" following investigations if nurses pronounced death at that institution.

In our view, this was a serious, and possibly systemic issue, that merited a more robust action from CSC. Rather than simply holding a meeting to discuss and review the issue of pronouncement of death, we would have expected a response initiating a further investigation of the matter in order to more fully understand the basis of these fears, how widespread they may be, and ensure that they are properly addressed.

There was one last case where the committee judged that the corrective measures and action plans initiated by CSC did not adequately address the investigation recommendations and underlying issues. In this case, the BOI found that an important contributing factor to a death in custody (due to an overdose of medication) was that information that the offender was seeing a psychiatrist and taking medication for a psychiatric condition was not communicated from the sending institution to the receiving institution upon transfer. The nurse at the receiving institution reviewed the file and noted suicide risk factors, but was unaware that he had been under psychiatric care and erroneously noted that he was not taking any medication. Accordingly, the offender was not put on the list for continuation of his psychiatric care. The BOI noted that the information was not contained on the electronic medical record file, the Open Source Clinical Application Resource (OSCAR), because the region had previously been granted an exemption from participating in the national electronic system. The BOI made only one recommendation, which was that the region be included in the electronic medical record file, OSCAR, so that detailed health information would be reliably communicated to health care staff whenever an inmate was transferred, either within the region, or from one region to another.

The CSC response indicated that the recommendation was supported but the action taken is not what was recommended. The CSC response explained that case conferencing to share information is done by telephone, but in this case, the information relayed was not documented in writing. Mental health staff in the region were reminded of their obligation to communicate and document information in cases of mental health issues and suicide risk. In addition, recent admissions to the institution where the incident occurred would be examined at each mental health committee review in order to identify actual or potential need for mental health services. In our view, this action fell short of what was recommended, and notably, did not address the gap in sharing information in cases of inter-regional transfers.

In summary, this committee believes that the four cases described above are instances where CSC's actions and responses to BOI reports fell short in adequately addressing the recommendations and underlying issues. As noted at the outset of this chapter, we are not making any recommendations to CSC in this area. Nevertheless, we would expect CSC to review these cases with our commentary in mind, and consider whether any further action is warranted.

Conclusion

The Convening Order for this Independent Review Committee included six elements in its Terms of Reference, which were set out in the introduction to this report. We addressed all of these elements in the course of our review. In the preceding chapters, we have summarized the main findings from our review, and made recommendations based on these findings.

In the introduction to this report, we stated an important caveat that established limits on the thoroughness of the review that we were able to conduct. Specifically, we did not have access to the investigative product for each BOI, and therefore our starting point was the BOI report itself (its findings, analyses, identification of underlying issues, and recommendations). We were comfortable proceeding on that basis for the purposes of our review, but considered it important to bring it the attention of the reader at the beginning, and at other points throughout this report.

In our review of the action taken in response to the recommendations of the 4th IRC (ToR 1) we found several examples of initiatives that the IIB had undertaken to improve the investigative process. We reported on these in the first chapter; commended CSC for undertaking these initiatives; and made some recommendations that are intended to incorporate these initiatives more fully into the investigative process. In addition, we make two recommendations concerning the engagement of Next of Kin that were prompted by our interviews with families and their experience when a family member dies in custody. We also make a recommendation regarding research on measuring the quality of prison life, a best practice from the UK that could assist CSC to gain a better understanding of the conditions in institutions as experienced by inmates and staff, take measures to the improve performance for the general well-being of inmates and staff, and possibly contribute to the reduction of incidents of non-natural death.

The Terms of Reference required us to analyze the independence and impartiality of CSC's investigative process (ToR 2). These are complex concepts that are fundamental to proper investigations. Even allowing for the caveat mentioned above, this was a significant undertaking. We meticulously tested CSC's investigative process against each of the criteria that are the hallmarks of independence and made findings from this assessment. We examined the organizational structure, policies and practices for investigations of death in custody in several other jurisdictions and found examples of best practice that could be adopted and adapted to strengthen the actual and perceived independence of CSC's investigative process, and made recommendations accordingly. We approached the issue of impartiality in the same manner - applying established criteria to determine how well CSC's investigative process measures up against these criteria; identifying ways to enhance the impartiality of its investigative process; drawing on best practice from other jurisdictions; and making recommendations that flowed from this analysis.

In Chapter 3, we address the Term of Reference (3) that asked us to provide suggestions on how CSC can share its investigative reports with staff to ensure a broader awareness of "lessons learned" and make them more accessible to the public. In this chapter we noted that there are many audiences that are interested in getting more information about these investigative reports, including CSC staff, and there is a desire to make more information available within the organization and to the public. We listed many platforms that are available currently within CSC but observe that they are not being utilized to the extent that they could be. In our international review of investigative practices regarding deaths in custody, we found that Prison and Probation Ombudsman (PPO) in the United Kingdom had established policies and procedures that provide for extensive sharing of information regarding their investigations of deaths in custody, beyond what is current practice in CSC. While we acknowledge differences between the British and Canadian contexts, we make recommendations that CSC move toward practices similar to those of the PPO.

The fourth chapter of this report focused on whether the recommendations made by Boards of Investigation appropriately addressed the underlying causes that led to the incident as well as any need for systemic improvements to policy and procedures (ToR 4). While the committee found that the BOI reports generally made appropriate recommendations, some of the BOI reports raised concerns. One concern was that in some cases that were numerous compliance issues, without a recommendation, raising the need to: a) allow a BOI to make recommendations on the basis of compliance issues when there is a recurring pattern of such issues; b) examine compliance issues collectively across BOI reports and prepare thematic reports on systemic issues related to compliance; and c) incorporate scenarios and lessons learned from recurring compliance issues into staff training at all levels of the organization. We make recommendations to address these areas. We also found instances where compliance issues were judged to have had "no direct impact on the incident", and therefore listed as "Supplementary Findings", where our sense was that they may well have had an impact and should have been retained as a focus in the report. We make recommendations to address this concern.

The last chapter of our report addressed the issue of whether the corrective measures and action plans initiated by CSC adequately addressed the recommendations and underlying issues in the BOI report. Our general conclusion, from our review of the BOI reports, was that the corrective measures and action plans were appropriate and adequate in the large majority of cases. However, there were four cases where, in our view, the response from CSC was not adequate and the underlying issues identified in the BOI report were not appropriately addressed. We provided a synopsis of each of these cases. We expect that CSC will review these cases with the benefit of our commentary, but we have not made any recommendations on this subject.

We have one last observation. This was the 5th in the series of Independent Reviews of Non-natural Deaths in Custody. The breadth and complexity of the Terms of Reference for this review were demonstrably greater than for previous reviews. Indeed, one of the Terms of Reference - analyzes the independence and impartiality of CSC's investigative process - could well have been the sole focus of a major review. As it was, the committee was tasked with addressing this issue, along with five other elements resulting in more work than could be done in the allotted timeframe. In order to avoid a reoccurrence of this situation, we urge CSC to ensure that, in the event that a 6th IRC is convened, the amount of time allotted to complete the review is commensurate with the scope of the Terms of Reference, including sufficient time to obtain and review relevant evidence in sufficient depth.

Correctional Service Canada should ensure that any future Independent Review Committee is provided with sufficient time to complete its mandate.

In closing, we sincerely hope that our report and its recommendation will assist CSC in their efforts to improve their investigative process and contribute to the prevention of non-natural deaths in custody.

References

Gabor, Thomas (2007) Deaths in Custody. A report submitted to the Office of the Correctional Investigator: Ottawa.

Correctional Service of Canada (2017), Death of a Person in the Care and Custody of Correctional Service Canada: A Guide for Family and Friends. Ottawa: Correctional Service Canada.

Correctional Service Canada (2019). Annual Report on Deaths in Custody 2016-17. Report SR-01. Ottawa: Correctional Service Canada.

Correctional Service Canada. Conducting Investigations Training Participants Manual 2020-2021. Ottawa: Correctional Service Canada.

Correctional Service Canada (2019). Notification of Death Training Package. Ottawa: Correctional Service Canada.

Office of the Correctional Investigator (2006). Annual Report 2005-2006. Ottawa: Correctional Investigator.

Office of the Correctional Investigator (2016). In the Dark: An Investigation of Death in Custody Information Sharing and Disclosure Practices in Federal Corrections. Ottawa: Correctional Investigator.

Office of the Correctional Investigator (2018). Annual Report 2017-2018. Ottawa: Correctional Investigator.

Nadeau, L., Brochu, S. & Cormier, R. (2018). Fourth Independent Review Committee on Non-natural Deaths in Custody that occurred between April 1^st, 2014 to March 31^st, 2017. Correctional Service Canada: Ottawa.

Liebling, A. (2004). Prisons and their Moral Performance: A Study of Values, Quality, and Prison Life. UK: Oxford University Press.

Measuring the Quality of Prison Life - MQPL+ (2015) Prisons Research: University of Cambridge. https://www.cam.ac.uk/research/impact/measuring-the-quality-of-prison-life

Liebling, A. (2014) Prison quality, moral performance and outcomes. A paper presented at the 19th Council of Europe Conference of the Directors of Prison and Probation Services, 17-18 June, 2014. Helsinki.

Auty, K. & Liebling, A. (2020) Exploring the Relationship between Prison Social Climate and Reoffending. Justice Quarterly, 37 (2) https://www.tandfonline.com/doi/abs/10.1080/07418825.2018.1538421

Ross, M., Liebling, A. & Tait, S. (2011) The Relationships of Prison Climate to Health Service in Correctional Environments: Inmate Health Care Measurement, Satisfaction and Access in Prisons (2011) The Howard Journal of Criminal Justice 50 (3), 262-274.

Annex A: List of Recommendations

1. The Values and Ethics of Correctional Service Canada should be referred to in the Board of Investigation Terms of Reference, be a focus of the investigative process where relevant, and included in findings and recommendations, as applicable.
2. "To examine all four pillars" be included in the Terms of Reference in cases involving overdoses.
3. A context section be included in every Board of Investigation report.
4. That Correctional Service Canada embark on a program of research on the Quality of Prison Life in Correctional Service Canada institutions along the lines of the research undertaken by Professor Alison Liebling and her associates.
5. That the Trends, Analyses and Performance database (TAP) be made available to the Board of Investigation teams to assist them in understanding the institutional context of the incident.
6. That Correctional Service Canada contact Next of Kin in the event of an imminent death in hospital in order to provide the opportunity for the family to travel to be with the inmate at the end of their life or to participate in end-of-life decisions.
7. That there are only two possible letters sent to families. That the Warden, notify the family, offer condolences and indicate that a letter will be sent by the Family Liaison Coordinator who will be their main point of contact for information about the circumstances of death, making funeral arrangements and further disclosure. If there is a subsequent investigation, the Incident Investigations Branch sends a second letter outlining that the family may participate in the Board of Investigation and can contact them for investigation information.
8. The Director General of the Incident Investigation Branch should report directly to the Commissioner of Correctional Service Canada.
9. A Board of Investigation should have the authority to develop additional Terms of Reference as an investigation is underway, without obtaining prior approval from Correctional Service Canada. The rationale for any such decision must be documented.
10. All communications between any Correctional Service Canada party and Board of Investigation members concerning findings and/or recommendations that the Board of Investigation has made or is considering making should be fully documented. If changes are made to any finding or recommendation, the process by which that occurred and the rationale for any changes should be included as an addendum to the Board of Investigation report.
11. An Independent Observer should be appointed to monitor a Board of Investigation in the instances defined by the Office of the Correctional Investigator in Recommendation 10 of its 2017/18 Annual Report, as well as any other death in custody where the Commissioner or the Minister of Public Safety determines it would be in the public interest to do so.
12. Correctional Service Canada should appoint Community Board Members as chairs in high profile Boards of Investigation, where it is reasonable and practical to do so.
13. Correctional Service Canada should work to appoint Community Board Members who have expertise and insight into issues facing disproportionately represented inmate populations, particularly in cases where race and/or cultural background may be an issue in a given investigation.
14. All Community Board Members should be given the same tools to do their job as those provided to Board of Investigation members who are Correctional Service Canada employees.
15. The Incident Investigation Branch should develop an additional component to its existing investigation training programme for all Board of Investigation members that covers contacting parties from outside Correctional Service Canada, who may have relevant evidence, including inmates, their victims, and their families.
16. The Incident Investigation Branch should create a formal training programme for Community Board Members, including training on investigative interviewing, Correctional Service Canada Mission and Core Values and their vital role in questioning and challenging, if necessary, how the investigation is proceeding. This training should be delivered as soon as a Community Board Member is appointed, with refresher training as required.
17. The Incident Investigation Branch should consider expanding the Enhanced Investigation Analysis process to more investigations, where appropriate.
18. Incident Investigations Branch staff should be dispatched to the place where a death occurred as soon as possible. They should ensure that all evidence that may be relevant to the death is identified, secured and gathered. They should also conduct preliminary interviews of parties who may not be easily available for interview by a Board of Investigation at a future date, such as inmates and family members.
19. Boards of Investigation should identify and, if warranted, interview family members, inmates, victims and anyone else, when it appears that they may have information that is relevant to the death.
20. All interviews conducted by a Board of Investigation during the course of its investigation should be digitally voice recorded.
21. All Board of Investigation reports should include a detailed segment on what the Board of Investigation did to gather evidence.
22. That investigation reports with corrective actions from non-natural deaths be made available on Correctional Service Canada's website, with minimal redactions, for staff and public awareness, similar to the Prison and Probation Ombudsman in the United Kingdom.
23. Correctional Service Canada should routinely develop and share Lessons Learned Fact Sheets/Bulletins from non-natural deaths on their website, similar to the Prison and Probation Ombudsman in the United Kingdom.
24. Incident Investigations Branch take steps, similar to the Prison and Probation Ombudsman in the United Kingdom, to further include families and Next of Kin into the investigation process. Offerings need to include the opportunity to suggest ideas for investigation; the sharing of draft reports for accuracy; a final report debriefing (with as much information as possible); and notification of the final report being made public.
25. That "Previous Relevant Board of Investigation Reports" and "Relevant Independent Review Committee Reports" be added to the Investigation Incident Branch's Participant Training Manual's "Investigation Documents List" Annex.
26. That Boards of Investigation be allowed to make Recommendations, to correct reoccurring Compliance Issues, to advance Correctional Service Canada learnings. When a Board of Investigation identifies reoccurring issues, then the Board of Investigation should make a recommendation to address this, or explain why they have not done so.
27. That Correctional Service Canada incorporate scenarios and lessons learned into staff training at all levels from reoccurring Compliance Issues that are noted by Boards of Investigation.
28. That the Incident Investigations Branch work to create further criteria to define the wording "had no direct impact on the incident", as it relates to Compliance Issues.
29. If a Board of Investigation moves a Compliance Issue to a Supplementary Finding, it should outline the reason for the decision.
30. That the Incident Investigations Branch develop a thematic report focused on Board of Investigation reports into deaths in custody to determine if there exist any patterns and/or potential systemic issues in relation to compliance. Based on the outcome of this review, it may be appropriate to include this topic as a Term of Reference for a future Independent Review Committee.
31. Correctional Service Canada should ensure that any future Independent Review Committee is provided with sufficient time to complete its mandate.