Deaths in prisons

In 2021, over 11.5 million people were held in prisons globally, of whom almost a third were not convicted.¹ The number of people who die in prisons from all causes is largely unknown, though it is estimated that mortality rates for people in prison are at least 50 per cent higher than for people in the wider community. Prisoner deaths are probably increasing because of the increasing size of prison populations, the length of sentences, life sentences² and the proportion of older people in prison.³ Many such deaths are preventable.⁴ Indeed, a 2023 Canadian study that examined prisoner deaths between 2014 and 2021 found almost every life lost could have been prevented.⁵

1. Penal Reform International, Global Prison Trends 2021, available at https://cdn.penalreform.org/wp-content/uploads/2021/05/Global-prison-trends-2021.pdf; see also Helen Fair and Roy Walmsley, “World Prison Population List” (thirteenth edition), Institute for Crime & Justice Policy Research and World Prison Brief (2018), available at https://www.prisonstudies.org/sites/default/files/resources/downloads/world_prison_population_list_13th_edition.pdf. ↩
2. Dirk Van Zyl Smit and Catherine Appleton, Life Imprisonment: A Global Human Rights Analysis (Cambridge, Harvard University Press, 2019). ↩
3. Audrey Roulston and others, “Deaths in prison custody: A scoping review of the experiences of staff and bereaved relatives”, The British Journal of Social Work, vol. 51, No. 1 (2021), pp. 223–245. ↩
4. See https://spcommreports.ohchr.org/Tmsearch/TMDocuments. ↩
5. “An obligation to prevent”, report from the Ontario Chief Coroner’s expert panel on deaths in custody, January 2023. ↩

Deaths in custody are a silent global tragedy. Article 10 of the International Covenant on Civil and Political Rights identifies reformation and social rehabilitation as key aims of the penitentiary system. At the present time, however, instead of rehabilitation, prisoners are dying – and dying needlessly. Upholding the right to life is the minimum that States must achieve; rebuilding lives the ultimate goal. To what extent prisons can contribute to that goal remains unclear.

Prisoner deaths have been subject of reports and recommendations by United Nations bodies and special procedures⁶ and a focus for the Special Rapporteur on extrajudicial, summary or arbitrary executions since the establishment of the mandate.⁷ The present report continues this focus, addressing the prevention and investigation of deaths of people under the control of a detaining authority, including deaths in pretrial detention, psychiatric and medical facilities, whether or not they are part of correctional detention, and during travel to and from such institutions. Also included are deaths that occur outside detention facilities while prisoners are under the control of the detaining authority, for example home visits, compassionate leave and day release.

6. See A/HRC/42/20. ↩
7. See, most recently, A/HRC/47/33 and A/76/264. ↩

When discussing deaths in prisons, the focus tends to be on preventable and unlawful deaths, not the death penalty. However, not all judicial executions are lawful. In its general comment No. 36 (2019), the Human Rights Committee stated that the death sentence may be imposed only for the most serious crimes, understood as intentional killing.⁸ Death sentences must not be mandatory; judges must be able to exercise discretion by taking individual circumstances into consideration in each case. Additional safeguards under article 6 of the International Covenant on Civil and Political Rights must be observed. In many States where prisoners are executed, the death penalty, as currently performed, breaches those requirements and is, therefore, being imposed unlawfully. Further, as the Special Rapporteur has argued, the nature of its implementation, including both the treatment of individuals on death row and the methods of execution, renders the death penalty tantamount to torture.⁹ States should be moving towards abolition of the death penalty.

8. CCPR/C/GC/36, para. 35. ↩
9. See A/77/270. ↩

The report draws on the literature on deaths in prisons and interviews with experts from countries with limited resources, given the paucity of published information from these settings. Interviews were conducted by the Global Research Initiative on Forensic Medicine and Human Rights and Eleos Justice at Monash University in Australia, in association with the Graduate School of Business and Law, Royal Melbourne Institute of Technology (RMIT) University. The Special Rapporteur would like to thank all who contributed to the report, including the interviewees and those who responded to the call for input. The Special Rapporteur wishes to especially acknowledge the contributions of Professor Bebe Loff, Professor Bronwyn Naylor, Dr. Ashleigh Stewart, Dr. Mai Sato, Dr. Reena Sarkar, Dr. Nicholas Dempsey, Professor Emeritus Stephen Cordner, and Professor Richard Bassed.

The report covers the main activities undertaken by the Special Rapporteur from August 2022 to March 2023. Activities undertaken from April to July 2022 are included in the thematic report of the Special Rapporteur to the General Assembly.¹⁰

10. See A/77/270. ↩

During the period under review, the Special Rapporteur issued, alone or jointly with other special procedure mandate holders, 82 communications to States and non-State actors, as well as 38 press statements.

As part of the efforts of the Special Rapporteur to promote and assist in the implementation of standards relating to the effective prevention and investigation of unlawful deaths, including the Minnesota Protocol on the Investigation of Potentially Unlawful Death (2016),¹¹ as well as the death penalty, he carried out numerous activities, including meetings, lectures and consultations with governmental, intergovernmental, non-governmental and academic stakeholders, as well as with families of victims of unlawful deaths during the reporting period. In addition, the Special Rapporteur provided technical advice on matters related to the mandate, including the following visits, described below.

11. The Minnesota Protocol on the Investigation of Potentially Unlawful Death (2016): The Revised United Nations Manual on the Effective Prevention and Investigation of Extra-Legal, Arbitrary and Summary Executions, available at https://www.ohchr.org/sites/default/files/Documents/Publications/MinnesotaProtocol.pdf. ↩

From 5 to 8 September 2022, the Special Rapporteur conducted an academic visit to Tegucigalpa, Honduras, at the invitation of the Office of the United Nations High Commissioner for Human Rights (OHCHR) regional office for Central America, to participate in the first regional forum for the investigation and accountability for crimes against LGBTQI+ people.

From 12 to 17 September 2022, the Special Rapporteur conducted a technical visit to Yerevan, organized with the support of the Office of the United Nations Resident Coordinator in Armenia, to carry out a capacity needs assessment for the implementation of training activities on the Minnesota Protocol.

On 12 December 2022, the Special Rapporteur co-organized a public event on the role of civil society in the establishment of the mandate, held at La Moneda Palace in Santiago, together with representatives of the Chilean Presidency, the Ministry of Justice and Human Rights, the Office of the United Nations Resident Coordinator, the OHCHR regional office for South America and Fundación Horizontes Ciudadanos.

From 14 to 20 January 2022, the Special Rapporteur travelled to Mexico, invited by Servicios y Asesoría para la Paz and the OHCHR country office in Mexico, to deliver, at a number of public events, lectures on the work carried out under the mandate and offer advice on mandate-related forensic best practices to governmental authorities, non-governmental organizations (NGOs), national human rights institutions and families of victims.

From 4 to 10 February 2023, the Special Rapporteur conducted an academic visit to the Philippines, at the invitation of the joint programme between the Philippines and the United Nations on technical cooperation for the protection and promotion of human rights, to carry out an assessment for the implementation of training activities on the use of the Minnesota Protocol.

States assume direct responsibility for the lives of individuals deprived of liberty and they may not use lack of financial resources, logistical problems or the privatization of prisons to reduce this responsibility.¹² The presumption of State responsibility for prisoner deaths until such responsibility is refuted was first articulated by the inaugural Special Rapporteur appointed to examine the questions related to summary or arbitrary executions¹³ and affirmed by the Human Rights Committee in the case of Dermit Barbato v. Uruguay.¹⁴ It has been consistently reaffirmed by subsequent Special Rapporteurs,¹⁵ the Secretary-General¹⁶ and the Human Rights Committee.¹⁷ The European Court of Human Rights similarly found that the burden of proof rests on prison authorities.¹⁸

12. CCPR/C/GC/36, para. 25. ↩
13. E/CN.4/1986/21, para. 209. ↩
14. Human Rights Committee, Dermit Barbato v. Uruguay, Communication No. 84/1981, para. 9.2. ↩
15. See, for example, A/61/311, para. 50, and A/HRC/32/39, para. 99. ↩
16. A/68/261, para. 52. ↩
17. CCPR/C/GC/36, para. 29. ↩
18. European Court of Human Rights, Tanli v. Turkey, Application No. 26129/95, Judgment, 28 August 2001. ↩

The presumption of State responsibility extends to the apparent voluntary recruitment of prisoners to take part in activities that are likely to result in death. Prisons are coercive environments; prisoner capacity to act autonomously is severely curtailed. Where conditions in prison are deficient, any offer of a potential benefit might appear too good to refuse. Prisoners may legitimately be asked to agree to be part of risky activities, such as medical research, when potential benefits to them outweigh the risks, and all steps are taken to minimize such risks. However, prisoners should not be offered overwhelmingly attractive inducements to undertake activities where their lives are at serious risk and little or no effort is made to minimize such risk. Arguably, extrajudicial execution occurs when poorly informed prisoners, whose prison conditions are sub-standard, are induced to participate in life-threatening activities that result in death.

A corollary of the presumption of State responsibility is that if States do not have practical control of their prisons, they should refrain from sending convicted persons to prison. Alternatives to prison should be developed. As noted in 2008 by Philip Allston, the then Special Rapporteur on extrajudicial, summary or arbitrary executions, the “State has no right to imprison a person in order to subject him or her to the caprices and arbitrariness of thugs, whether in the name of necessity, realism or efficiency”.¹⁹

19. See A/HRC/8/3, para. 87. ↩

Incarcerated people tend to be the most economically and socially marginalized and the least educated, with significantly poorer physical and mental health and higher rates of disability than the general population.²⁰ This degree of background deprivation suggests that prisoners are likely to have significantly greater vulnerability to physical and mental illness caused by poor physical conditions, social exclusion, lack of social and health services, lack of meaningful activity and violence.

20. See Institute for Crime and Justice Policy Research, Towards a health-informed approach to penal reform? Evidence from ten countries, 2019. ↩

Leading proximate causes of preventable prisoner deaths are violence, suicide and communicable diseases. Specifically, deaths may result from lack of food or nutritious food, 12 CCPR/C/GC/36, para. 25. 13 E/CN.4/1986/21, para. 209. 14 Human Rights Committee, Dermit Barbato v. Uruguay, Communication No. 84/1981, para. 9.2. 15 See, for example, A/61/311, para. 50, and A/HRC/32/39, para. 99. 16 A/68/261, para. 52. 17 CCPR/C/GC/36, para. 29. 18 European Court of Human Rights, Tanli v. Turkey, Application No. 26129/95, Judgment, 28 August 2001. 19 See A/HRC/8/3, para. 87. 20 See Institute for Crime and Justice Policy Research, Towards a health-informed approach to penal reform? Evidence from ten countries, 2019. lack of potable water, absent or inadequate hygiene and sanitation, vermin infestations, dilapidated buildings, fires, exposure to temperature extremes and lack of access to health care.²¹ 21 High levels of distress, hopelessness and a sense of disempowerment in these circumstances, as well as social exclusion, may lead to mental illness and suicide. Overcrowding in prisons, which is a serious concern generally, has been highlighted during the coronavirus disease (COVID-19) pandemic.²² Overcrowding exacerbates the negative impact of poor conditions, in many cases being so extreme that prisoners are unable to stretch, stand, walk or sleep, and the spread of infectious disease is far more likely. Inter-prisoner and gang violence and riots leading to deaths are generally an outcome of prison conditions.²³

21. A/HRC/42/40, paras. 29–34; A/68/261; E/C.12/COD/CO/4, para. 32; CAT/OP/BEN/1, paras. 221–222; A/HRC/14/24/Add.3, para. 85; A/HRC/20/22/Add.1, para. 52; E/CN.4/2006/53/Add.3, para. 70; and A/HRC/31/CRP.1. ↩
22. United Nations Office on Drugs and Crime (UNODC), World Health Organization (WHO), Joint United Nations Programme on HIV/AIDS (UNAIDS) and OHCHR joint statement on COVID-19 in prisoners and other closed settings, 13 May 2020. ↩
23. For further discussion, see A/HRC/42/20, paras. 14–21 and A/HRC/4/20/Add.2, paras. 37–41. ↩

The more people are imprisoned, the more people will die in prisons. While some States have taken steps to reduce the numbers of people incarcerated, others have increased rates of imprisonment and sentence length. Imprisonment plays social and symbolic functions: increasing the length of sentences and criminalizing an expanding range of human activity is often a calculated, populist response to perceived community concerns about safety and crime rates.²⁴ Imprisonment does nothing to address the causes of crime or substantively improve community safety. Indeed, there is evidence that high incarceration rates increase crime.²⁵ States also use imprisonment to deal with political opposition. These practices must be challenged if prison deaths are to be reduced. Investment in community and institutional infrastructure is needed to address social and economic drivers of crime, thus reducing the need to commit people to prison and creating alternatives to incarceration.

24. Thomas Mathiesen, Prison on Trial (Waterside Press, London, 2006) and John Pratt, Penal Populism (Routledge, New York, 2007). ↩
25. Doris Layton MacKenzie and Douglas B. Weiss, “Other countries have successfully reduced incarceration rates without increasing crime: we can do it!”, Victims & Offenders, vol. 4, No 4 (2009), pp. 420–426. ↩

A third of the prison population is in pretrial detention; some of these prisoners will be declared innocent. Individuals are often held in pretrial detention for years. In some cases, time spent in pretrial detention is longer than the maximum sentence for the crime. Individuals may be denied bail because they are homeless, are without identity documents, cannot afford to post bail, do not know individuals with sufficient means who might agree to stand surety or lack legal representation. These are not sufficient reasons for imprisonment. States must reduce the number of people in pretrial detention. Mandatory pretrial detention should be abolished. A legislative presumption of release on bail should be introduced so that bail may be granted more routinely. In addition, resourcing and streamlining of prosecution and court processes and criminal procedures may lessen the time accused persons spend in pretrial detention.

A significant proportion of people are imprisoned for low level offences. Offences of poverty, for example vagrancy, drunkenness, inability to pay a debt or fine, minor theft or causing public disorder, should be repealed or be subject to penalties other than imprisonment. Victimless crimes, such as sex work, should be decriminalized, as should religious offences, such as apostacy and blasphemy, and gender-based crimes, such as adultery, “decency” related crimes and pregnancy related crimes, including abortion. Women should not be imprisoned for exercising their reproductive rights. If evidence has been obtained through torture, such evidence should not be permitted to support convictions.

States should also consider decriminalizing minor offences, such as low-level drug use and possession. If such offences are retained, police, prosecutors and the judiciary should have the power to divert cases away from pretrial within the criminal justice system. 21 A/HRC/42/40, paras. 29–34; A/68/261; E/C.12/COD/CO/4, para. 32; CAT/OP/BEN/1, paras. 221– 222; A/HRC/14/24/Add.3, para. 85; A/HRC/20/22/Add.1, para. 52; E/CN.4/2006/53/Add.3, para. 70; and A/HRC/31/CRP.1. 22 United Nations Office on Drugs and Crime (UNODC), World Health Organization (WHO), Joint United Nations Programme on HIV/AIDS (UNAIDS) and OHCHR joint statement on COVID-19 in prisoners and other closed settings, 13 May 2020. 23 For further discussion, see A/HRC/42/20, paras. 14–21 and A/HRC/4/20/Add.2, paras. 37–41. 24 Thomas Mathiesen, Prison on Trial (Waterside Press, London, 2006) and John Pratt, Penal Populism (Routledge, New York, 2007). 25 Doris Layton MacKenzie and Douglas B. Weiss, “Other countries have successfully reduced incarceration rates without increasing crime: we can do it!”, Victims & Offenders, vol. 4, No 4 (2009), pp. 420–426. However, the exercise of diversionary power should be transparent and safeguards should be introduced to ensure that diversion practices are not discriminatory or subject to corruption.

Innovative justice approaches should be introduced as a diversion from or alternative to conventional criminal justice. This includes transformative and restorative²⁶ justice approaches (for example, victim/offender mediation, family group conferencing and sentencing circles), verbal sanctions, arbitrated settlements, victim restitution and community service orders.²⁷ Redirecting prison funds to justice reinvestment²⁸ and community-based correction programmes should be considered.

26. See also basic principles on the use of restorative justice programmes in criminal matters (Economic and Social Council resolution 2002/12). ↩
27. UNODC, Handbook on Women and Imprisonment, Vienna, 2014. ↩
28. David Brown and others, Justice Reinvestment: Winding Back Imprisonment (Palgrave Macmillan, 2016). ↩

Sentencing options should include alternatives to prison. The 1990 United Nations Standard Minimum Rules for Non-custodial Measures (the Tokyo Rules) promote the use of non-custodial measures and sanctions and minimum safeguards for persons subject to alternatives to imprisonment, as do the United Nations Rules for the Treatment of Women Prisoners and Non-custodial Measures for Women Offenders (the Bangkok Rules). Non-custodial options include rehabilitation and treatment programmes,²⁹ community service, fines, electronic monitoring and home detention. Where home detention is imposed as a sentence, the State must monitor its impact on others in the home and provide support where required. Using non-custodial options directly reduces the risks of prison deaths.

29. UNODC and WHO, Treatment and care for people with drug use disorders in contact with the criminal justice system: Alternatives to conviction or punishment, 2019. ↩

Some countries introduced specific issue courts, of which drug courts of various kinds are an example. Their success in keeping people out of prison is highly dependent on their aims and how they are implemented. For example, where drug courts have imposed abstinence-based programmes, they have been associated with low success rates and imprisonment following drug use.³⁰ The existence of specific issue courts should not bring more people into contact with the criminal justice system nor interfere with strategies to decriminalize low level and poverty related offences.

30. See, for example, Victoria Stanhope and others, “Understanding service disengagement from the perspective of case managers”, Psychiatric Services, vol. 60, No. 4 (2009), pp. 459–464; see also John Robert Gallagher and others, “A perspective from the field: five interventions to combat the opioid epidemic and ending the dichotomy of harm-reduction versus abstinence-based programs”, Alcoholism Treatment Quarterly, vol. 37, No. 3 (2019), pp. 404–417. ↩

Mandatory prison sentences should be abolished. The judiciary should be encouraged to shorten the length of prison sentences. There are also measures to facilitate early release, including pre-release programmes and greater use of probation. A legislative presumption of early release on parole for good behaviour could be enacted and early release for older prisoners should be possible on compassionate grounds. Courts should consider reinforcing this provision by incorporating early release into sentences.

Another significant step States should take is to adopt measures to provide support to people leaving prison in order to keep them from committing crimes and prevent reoffending. Such support might involve linkage to community health services, for example mental health care and opioid agonist therapy. For prisoners with mental illness, linkage to community mental health-care services has shown to reduce reimprisonment.³¹ Placement in housing and the provision of income support should also be considered.

31. Ashleigh C. Stewart and others, “Criminal justice involvement after release from prison following exposure to community mental health services among people who use illicit drugs and have mental illness: a systematic review”, Journal of Urban Health, vol. 99, No. 4 (2022), pp. 635–654. ↩

Non-custodial alternatives reduce the use of imprisonment, with a potential reduction in prison deaths. The Kingdom of the Netherlands has drastically reduced its prison population by replacing short prison sentences with community service.³² As a result, there has been a steady decrease in the number of prisoners,³³ from 20,463 in 2006, to 11,623 in 2021, and 23 prisons have been closed since 2014.³⁴ In the State of Texas, the 2007 Whitmire/Madden correctional treatment and diversion plan reduced the length of sentences, mandated probation and rehabilitation as alternatives to incarceration and gave judges the discretion to sentence non-violent offenders to alternative dispositions.³⁵ Since enacting the reforms a decade ago, Texas has reduced the size of its prison population and has closed 10 prisons.³⁶ Further, police report that crime rates have dropped by 29 per cent, the lowest level since 1968.³⁷

32. In the 1980s, the Community Service Order was introduced in the Netherlands and became popular as a substitute for short-term imprisonment. In 2001, it was replaced by the so-called task penalty, which was designed to be used in place of prison sentences of up to three months; see Rob Allen, Reducing the use of imprisonment: What can we learn from Europe? (Criminal Justice Alliance, 2012) p. 13. ↩
33. The data include pretrial detainees/remand prisoners. ↩
34. World Prison Brief, available at https://www.prisonstudies.org/country/netherlands. ↩
35. Nolan Center for Justice, American Conservative Union Foundation, “Criminal justice reform in Texas”, available at https://conservativejusticereform.org/state/texas/. ↩
36. Ibid. ↩
37. Ibid. ↩

In 1998 the Colombian Constitutional Court declared conditions in Colombian prisons to be unconstitutional.³⁸ In 2023, the Justice Minister proposed changes to Colombia’s penal code,³⁹ including a ban on pretrial detention and prison sentences for non-violent crimes and the incorporation of reintegration programmes for prisoners during parole. In certain instances, restorative justice processes will be applied in cases that would ordinarily be dealt with as part of the conventional criminal justice system. This follows the passage of a law in 2022 that limited the incarceration of sentenced women to eight years or less in prison.⁴⁰ The law applies to marginalized women who are single heads of households responsible for minor children, older persons or people with permanent disability.

38. Colombian Constitutional Court, Judgment No. T-153/98. ↩
39. Adriaan Alsema, “How Colombia’s judicial reform seeks to solve prison crisis”, Colombia Reports, 7 February 2023, available at https://colombiareports.com/how-colombias-judicial-reform-seeks-to-solve-prison-crisis/. ↩
40. Coletta Youngers, “Colombia to implement law on alternatives to incarceration for women heads of household”, Advocacy for Human Rights in the Americas, 11 August 2022, available at https://www.wola.org/analysis/colombia-law-alternatives-incarceration-women-heads-household/. ↩

The Malawi Bail Project helps people convicted of minor offences to apply for bail.⁴¹ The project addresses the lack of legal aid lawyers in Malawi, where the majority of those arrested lack legal representation or knowledge of their right to bail. Through various means such as theatre, plays, booklets and audio-recordings, the project educates people in local communities about the right to bail.

41. Malawi Bail Project: Access to Justice, available at https://www.malawibailproject.com/project-overview. ↩

Since 2014, the Californian elderly parole programme has provided that prisoners aged 60 or older who have served a minimum of 25 years may be referred to parole board hearings to determine suitability for parole. In 2021, the qualification for the programme was lowered to age 50 and years served to 20 years or more of continuous incarceration.⁴²

42. State of California, Board of Parole Hearings, “Fact Sheet”, available at https://www.cdcr.ca.gov/bph/wp-content/uploads/sites/161/2022/03/Elderly-Parole-Fact-Sheet3_18-1.pdf. From February 2014 through the end of September 2020, 1,377 paroles have been granted in the State of California, see Prison Law Office, Information About the Elderly Parole Program (2020), available at https://prisonlaw.com/wp-content/uploads/2021/01/Elder-Parole-Nov-2020v2.pdf. ↩

Decriminalization of drug-related offences significantly reduces the use of imprisonment. In Portugal, drug use and possession were decriminalized in 2001: the proportion of individuals sentenced to prison for drug offences declined from over 40 per cent in 2001 to 16 per cent in⁴³ 2019.43 In 2013, Costa Rica adopted a law providing that 32 In the 1980s, the Community Service Order was introduced in the Netherlands and became popular as a substitute for short-term imprisonment. In 2001, it was replaced by the so-called task penalty, which was designed to be used in place of prison sentences of up to three months; see Rob Allen, Reducing the use of imprisonment: What can we learn from Europe? (Criminal Justice Alliance, 2012) p. 13. 33 The data include pretrial detainees/remand prisoners. 34 World Prison Brief, available at https://www.prisonstudies.org/country/netherlands. 35 Nolan Center for Justice, American Conservative Union Foundation, “Criminal justice reform in Texas”, available at https://conservativejusticereform.org/state/texas/. 36 Ibid. 37 Ibid. 38 Colombian Constitutional Court, Judgment No. T-153/98. 39 Adriaan Alsema, “How Colombia’s judicial reform seeks to solve prison crisis”, Colombia Reports, 7 February 2023, available at https://colombiareports.com/how-colombias-judicial-reform-seeks-tosolve-prison-crisis/. 40 Coletta Youngers, “Colombia to implement law on alternatives to incarceration for women heads of household”, Advocacy for Human Rights in the Americas, 11 August 2022, available at https://www.wola.org/analysis/colombia-law-alternatives-incarceration-women-heads-household/. 41 Malawi Bail Project: Access to Justice, available at https://www.malawibailproject.com/projectoverview. 42 State of California, Board of Parole Hearings, “Fact Sheet”, available at https://www.cdcr.ca.gov/bph/wp-content/uploads/sites/161/2022/03/Elderly-Parole-Fact-Sheet3_181.pdf. From February 2014 through the end of September 2020, 1,377 paroles have been granted in the State of California, see Prison Law Office, Information About the Elderly Parole Program (2020), available at https://prisonlaw.com/wp-content/uploads/2021/01/Elder-Parole-Nov-2020v2.pdf. 43 Transform Drug Policy Foundation, “Drug decriminalization in Portugal: setting the record straight” (2021), available at https://transformdrugs.org/blog/drug-decriminalisation-in-portugal-setting-the qualified women serving a sentence of eight years or more for smuggling drugs into prison would be subject to a reduced sentence, from 8–20 years to 3–8 years. To qualify, women need to meet one of the following criteria: live in poverty; be head of a household responsible for minors or the elderly; living with a disability; or be an elderly person.⁴⁴ In Australia, since 1999, all six States and two territories have introduced diversion programmes for minor drugrelated offences.⁴⁵

43. Transform Drug Policy Foundation, “Drug decriminalization in Portugal: setting the record straight” (2021), available at https://transformdrugs.org/blog/drug-decriminalisation-in-portugal-setting-the-record-straight. Note that the actual number of individuals in prison for drug offences “has remained relatively steady, but a rise in overall prison numbers means the proportion of people serving sentences for drug offences has continued to fall”. ↩
44. Niamh Eastwood and others, A Quiet Revolution: Drug Decriminalisation Across the Globe (Release: Drugs, the Law and Human Rights, 2016) pp. 19–20. ↩
45. The diversion schemes in the different jurisdictions vary, with a range of interventions including or warning schemes or attendance at drug treatment or counselling sessions in lieu of prosecution. ↩

Prisoners should be treated with dignity. States should ensure that prisons are safe, habitable, not overcrowded, clean and fireproof and that utilities are operational; that nutrition is sufficient; that potable water is available in sufficient quantity; that prisoners’ health is monitored and that necessary steps are taken to prevent physical and mental illness and injury, including self-harm; that those who have existing health needs are treated, with their consent; that the requirements of those who have a disability are provided for; that those who need counselling and support receive it; that prisoners are engaged in meaningful activity; and that prisoners are able to maintain contact with family and friends. Age and gender-based vulnerabilities and requirements (including LGBTQI+ prisoners), should be considered in the allocation of prisoners, as well as in access to programmes and health care. There is an increased risk of suicide early in a prison sentence⁴⁶ and shortly after release,⁴⁷ with stressors of community reintegration associated with poor psychiatric well-being. Reception and pre-release assessment programmes should determine the mental health status of prisoners, their suicide risk and general vulnerability. Suitable support should be provided to prevent suicides. Possible hanging points should be removed. Inhumane suicide prevention methods such as solitary confinement should be abandoned. The use of physical and/or pharmacological restraints should be a last resort. States should shield prisoners from inter-prisoner violence and protect those prisoners known to be at particular risk.⁴⁸

46. Shaoling Zhong and others, “Risk factors for suicide in prisons: a systematic review and meta-analysis”, Lancet Public Health, vol. No. (2020), pp. 164–174. ↩
47. See, for example, Erin Morgan and others, “Incarceration and subsequent risk of suicide: A statewide cohort study”, Suicide and Life-Threatening Behaviour, vol. 52, No. 3 (2022), pp. 467–477. ↩
48. CCPR/C/GC/36, para. 25 (references omitted). ↩

There should be sufficient, well-trained prison staff, with appropriate skills to meet prisoner needs, including the ability to build positive relationships and respond without violence to challenging situations. The rights of prison staff should also be observed, including the provision of reasonable wages and conditions of work. If the rights of prison staff are not respected, they may be less inclined to respect prisoners’ rights. States should ensure that skilled management is in place and that prisons are well governed. These requirements are unexceptional – it can only be concluded that when these basic requirements are not met, it is intentional.

Governments are not exempt because they lack resources. If basic expectations cannot be met in prison, imprisonment itself will breach human rights, including the right to life, and should not be imposed. As stated by a previous Special Rapporteur, “Where a Government insists that a regular prison system run by trained, disciplined and humane authorities is beyond its financial means, the alternatives are to revamp the criminal justice system to institute other forms of punishment, to place less reliance upon imprisonment and to instigate a more efficient court system which processes cases more rapidly”.49 record-straight. Note that the actual number of individuals in prison for drug offences “has remained relatively steady, but a rise in overall prison numbers means the proportion of people serving sentences for drug offences has continued to fall”. 44 Niamh Eastwood and others, A Quiet Revolution: Drug Decriminalisation Across the Globe (Release: Drugs, the Law and Human Rights, 2016) pp. 19–20. 45 The diversion schemes in the different jurisdictions vary, with a range of interventions including or warning schemes or attendance at drug treatment or counselling sessions in lieu of prosecution. 46 Shaoling Zhong and others, “Risk factors for suicide in prisons: a systematic review and metaanalysis”, Lancet Public Health, vol. No. (2020), pp. 164–174. 47 See, for example, Erin Morgan and others, “Incarceration and subsequent risk of suicide: A statewide cohort study”, Suicide and Life-Threatening Behaviour, vol. 52, No. 3 (2022), pp. 467–477. 48 CCPR/C/GC/36, para. 25 (references omitted). 49 A/HRC/8/3, para.

Accessible, competent and appropriate health and medical care for prisoners is crucial in preventing deaths in prisons. In Ghana, intensive HIV prevention programmes in all prisons have resulted in declining HIV prevalence. The prisons aids control programme was established in May 2001 to create awareness among prisoners, prison officers and dependants about their vulnerability to HIV infection.⁵⁰ In 2008, HIV prevalence among prisoners was 19.2 per cent⁵¹ but this had been reduced to 0.4 per cent by 2017, with an antiretroviral therapy coverage rate of 100 per cent.⁵² In South Africa, detainees are required to be screened for tuberculosis on admission, biannually and on release, resulting in early detection and the initiation of prompt treatment.⁵³

50. Ghana Prisons Service, HIV/TB Workplace Policy and Implementation Strategy (2011). ↩
51. UNODC, report on the UNODC prisons assessment mission to Ghana (15–19 March 2010). ↩
52. Ghana AIDS Commission, Strategic plan for a comprehensive response to human rights-related barriers to HIV and TB services in Ghana 2020–2024 (2019). ↩
53. Kathleen Baird and others, “Tuberculosis control at a South African correctional centre: Diagnosis, treatment and strain characterisation”, PLoS One (11 November 2022), available at https://doi.org/10.1371/journal.pone.0277459. ↩

During the course of the COVID-19 pandemic in several countries, prisoners serving short sentences or who committed non-violent crimes have been granted early release. Some countries have experienced significant reductions in prison populations, with the region of Catalonia in Spain experiencing a 17 per cent reduction. Other factors contributing to the reduction included a decline in court activity and a possible decrease in crime during periods of lockdown. In Spain, the incidence of COVID-19-related deaths among prisoners has been low. The impact of COVID-19 among prisoners has been less than the wider community. In 2020, total infections were four times less, hospitalizations seven times less and deaths 10 times less: three out of 49,998 prisoners died.⁵⁴ The Health Ministry, in partnership with the Secretary of Prisons, devised specific strategies to be implemented in prisons. Strategies were underpinned by a policy of non-discrimination and the provision of treatment equal to that available to the general public. Information campaigns were implemented, with special attention to vulnerable groups. Prisoners and prison staff were regarded as high priority groups for vaccination.⁵⁵

54. Santiago Redondo and others, “Corrections and crime in Spain and Portugal during the COVID-19 pandemic: Impact, prevention and lessons for the future”, Victims & Offenders, vol. 15, No. 7-8 (2020). ↩
55. Vicente Martín, “COVID-19 and Prisons in Spain: is there any good news?”, Revista Española de Sanidad Penitenciaria, vol. 24, No. 3 (2022), pp. 77–78. ↩

There are several programmes to prevent drug-related deaths in prisons. Needle and syringe programmes are an essential harm reduction measure for the prevention of blood-borne virus transmission. While needle and syringe programmes exist in 92 countries, only 9 offer such programmes in prison.⁵⁶ Canada provides the world’s only prison-based drug consumption room.⁵⁷ In response to the high rates of overdose among those detained in Canadian prisons, as well as immediately following prison release, a drug consumption room was created in the Drumheller Institution in the Province of Alberta. The idea behind the establishment of a drug consumption room was to prevent overdoses among prisoners using illicit substances, prevent needle sharing, limit transmission of blood-borne viruses and facilitate referrals to health-care services. The success of the first prison-based drug consumption room in saving lives resulted in the implementation of other such rooms in federal prisons in Canada. The 2022 Global State of Harm Reduction report estimated that 59 countries offer prison-based opioid agonist therapy to prevent deaths from overdose. The administration of this therapy stabilizes people during imprisonment and safeguards them from overdosing immediately following release.

56. Harm Reduction International, The Global State of Harm Reduction, London 2022. ↩
57. Government of Canada, The Overdose Prevention Service, available at https://www.csc-scc.gc.ca/health/002006-2002-en.shtml. ↩

Various initiatives exist to prevent suicide in prisons. In the United Kingdom of Great Britain and Northern Ireland, the prevention of suicide in prisons: enhancing access to therapy programme, a collaboration between the Greater Manchester Mental Health NHS 50 Ghana Prisons Service, HIV/TB Workplace Policy and Implementation Strategy (2011). 51 UNODC, report on the UNODC prisons assessment mission to Ghana (15–19 March 2010). 52 Ghana AIDS Commission, Strategic plan for a comprehensive response to human rights-related barriers to HIV and TB services in Ghana 2020–2024 (2019). 53 Kathleen Baird and others, “Tuberculosis control at a South African correctional centre: Diagnosis, treatment and strain characterisation”, PLoS One (11 November 2022), available at https://doi.org/10.1371/journal.pone.0277459. 54 Santiago Redondo and others, “Corrections and crime in Spain and Portugal during the COVID-19 pandemic: Impact, prevention and lessons for the future”, Victims & Offenders, vol. 15, No. 7-8 (2020). 55 Vicente Martín, “COVID-19 and Prisons in Spain: is there any good news?”, Revista Española de Sanidad Penitenciaria, vol. 24, No. 3 (2022), pp. 77–78. 56 Harm Reduction International, The Global State of Harm Reduction, London 2022. 57 Government of Canada, The Overdose Prevention Service, available at https://www.cscscc.gc.ca/health/002006-2002-en.shtml. [National Health Service] Foundation Trust and various universities, offers therapy to those in prison and their families. A talk therapy session, delivered once or twice a week, enhances patient understanding of the reasons for suicide. The national charity, the Prison Advice and Care Trust, works with prison authorities to reduce distress among prisoners in the early days of custody.⁵⁸ The Prison Advice and Care Trust’s first-night workers sit with prisoners when they first arrive and assess whether they are at risk of self-harm or suicide.

58. See website of Prisoners, Families, Communities, A Fresh Start Together, available at https://www.prisonadvice.org.uk/who-we-are. ↩

Finally, while these examples of best practices specifically aim to prevent deaths in prison, a holistic approach to prison management is key to rebuilding lives, as enshrined in article 10 of the International Covenant on Civil and Political Rights. From 2003, under its “new model for prison management”, the Dominican Republic has introduced a significant systemic prison reform, which was complemented by the 2018 plan for the humanization of the penitentiary system. The Criminal and Penal Law of the Dominican Republic (law 224) requires that facilities be habitable and separated by sex. The law prescribes standards for hygiene, food, visits, access to lawyers and other matters. The new prison model is focused on prisoner rehabilitation and seeks to achieve consistency with the United Nations Standard Minimum Rules for the Treatment of Prisoners (the Nelson Mandela Rules). The law prescribes three phases of incarceration: intake/observation, treatment, and the “prueba”, or testing, phase. During the intake/observation period, the prisoner is separated from other prisoners. A personal history is taken and the prisoner is assessed to determine where they should be assigned and the services that should be offered to them. Rehabilitative programmes, including education, work opportunities, social welfare, spiritual assistance, health promoting and other activities, are offered during the next stage. During the final phase, the prisoner may be entitled to privileges, including day and weekend release, and prisoners may qualify for parole. A national penitentiary school provides a one-year intensive training course for “surveillance and treatment personnel” who staff correction and rehabilitation centres. Unfortunately, to date, only around half of the prisons in the Dominican Republic come under this programme although more will gradually do so. While the system has some shortcomings, it represents a significant advance. Commentators report that the new model for prison management demonstrates the importance of going beyond international minimum standards and placing prisoners’ autonomy, participation and dignity at the centre of reforms.⁵⁹ A 2019 study by the Global Foundation for Democracy and Development noted that the model is a remarkable achievement of building and implementing an entirely new vision and practice on how countries could handle punishment, rehabilitation and policy implementation, while acknowledging that, in some areas, there are shortages of services and access to key resources as well as weak oversight mechanisms.

59. Jennifer Peirce, “From rulay to rules: perceptions of prison life and reforms in the Dominican Republic’s traditional and new prisons”, presentation at the University of Toronto, 4 March 2022. ↩

Since 1863, the International Committee of the Red Cross (ICRC) has been concerned with the protection of detainees, initially prisoners of international armed conflicts. In 1870, during the Franco-Prussian war, it began distributing relief parcels to prisoners and enabled correspondence between prisoners and their families. By World War I, internment camps were visited, reports written, information concerning prisoners centralized and attempts made to improve conditions. The 1929 Geneva Convention relative to the Treatment of Prisoners of War formalized a monitoring role for ICRC. Subsequent to the failure of ICRC to disclose the condition of civilians held in the Theresienstadt concentration camp, the Geneva Conventions of 1949 enabled interviews in private with prisoners of war and civilian detainees.⁶⁰ The monitoring principles and practice of ICRC have informed contemporary prison monitoring requirements, including those set out in the Optional Protocol to the Convention against Torture and other Cruel, Inhuman or Degrading Treatment or Punishment.⁶¹

60. Alain Aeschlimann, Protection of detainees: ICRC action behind bars, International Review of the Red Cross, vol. 87, No. 857 (March 2005), pp 83–122. ↩
61. See General Assembly resolution 57/199. ↩

States that have not ratified the Optional Protocol should do so. Consistent with the Optional Protocol, States should ensure that external prison monitoring takes place on a regular basis by an adequately resourced expert body, functionally independent of national governments and prison authorities. The expert body should have the power to enter prisons with and without notice, speak to people and inspect documents without limitation and in private and to report publicly. People communicating with the expert body must be protected.

All prisoner deaths should be mandatorily reported to a properly resourced external expert body empowered to conduct an independent investigation. When possible, the expert body should form conclusions and make recommendations regarding prevention and civil and criminal culpability. However, simply making recommendations is insufficient. States should commit to timely consideration of all recommendations. Recommendations should be implemented when they are likely to make prisons more rights compliant. Further, the responses of States and their reasoning should be made publicly accessible. The expert body should have the power to report publicly and to follow up on recommendations.

Internal prison monitoring is essential for prisoner safety. In the United Nations Office on Drugs and Crime (UNODC) Handbook on Dynamic Security and Prison Intelligence (2015), the Office asserts that staff working with prisoners should have a high level of interpersonal skills and develop and sustain constructive relationships with prisoners. The development of systematized prison information and intelligence through the maintenance of constructive relationships is described as a means of, inter alia, identifying vulnerable prisoners and those who prey upon them and prison staff who are corrupt or violent. The means and degree of monitoring should be proportionate to real or reasonably perceived, rather than fanciful threats. While audiovisual surveillance may be valuable, it requires a balance between protecting prisoners’ privacy and safety. UNODC argues that video surveillance and recording should be accompanied by safeguards, including in relation to storage of and access to footage. It is equally important to have comprehensive, reliable prisoner management systems to ensure safe and appropriate allocation of prisoners and responsive sentence management.

It is presumed that States are responsible for the deaths of prisoners. It follows that all such deaths must be investigated. The death of person under the control of prison authorities is a prison death.⁶² Further, deaths are prison-related when prisoners die outside prison, during transport or following admission to a hospital from prison and should be reported to investigating authorities. Delayed prison-related fatalities may occur after formal prison release. Indeed, because the early post-release period is high risk, it should be presumed that all deaths occurring within 30 days of release are prison-related unless that presumption can be rebutted. However, prison-related deaths may also occur after this period of time.

62. See United Nations Standard Minimum Rules for the Treatment of Prisoners (the Nelson Mandela Rules), rules 69, 70–71. ↩

Globally, relatively few prison deaths are investigated, let alone properly investigated. In some countries a death will only be investigated if it involves a high-profile prisoner or if families agitate publicly for an investigation. Deaths may not be officially recorded or, if recorded, may lack detail. Countries may lack forensic medical services. Investigations may be carried out by people lacking expertise or by different agencies, without coordination. Agencies and investigators may not be independent of prison authorities. There may be no public accountability or transparency. Family members may be ignored or mistreated. The body of the deceased may be treated disrespectfully. Rarely are patterns or practices causing or contributing to the death identified, let alone the knowledge derived from investigations used to prevent future deaths.

All investigations should comply with the Principles on the Effective Prevention and Investigation of Extra-legal, Arbitrary and Summary Executions,⁶³ as supplemented by the Minnesota Protocol on the Investigation of Potentially Unlawful Death.⁶⁴ Investigations must be undertaken by an independent, impartial expert body, in a prompt, effective, thorough and transparent manner. The main purpose is to discover the truth of the circumstances leading to the death. This includes identifying the deceased, determining the cause and manner of death, and thus distinguishing between homicide, suicide, accidental death and natural death, always remembering that what appears to be a natural death may well have been caused or contributed to by prison conditions. The investigation may inform prosecution and punishment of those responsible, lead to an effective remedy for the next of kin and prevent the recurrence of similar deaths. While the detail of investigatory mechanisms is a matter for the State, it must accord with the Principles on the Effective Prevention and Investigation of Extra-legal, Arbitrary and Summary Executions and the Minnesota Protocol.

63. See https://www.ohchr.org/sites/default/files/Documents/ProfessionalInterest/executions.pdf. ↩
64. Minnesota Protocol on the Investigation of Potentially Unlawful Death (2016), p. 22. ↩

Unfortunately, in much of the world the Minnesota Protocol is unknown or ignored. At a minimum, States should ensure that training programmes are provided for all prison staff, police, forensic doctors and the staff of independent investigating agencies. Ideally the Protocol should form part of medical and legal curricula, the postgraduate training of forensic doctors and criminal lawyers and continuing professional development for the judiciary.

When a prisoner dies, prison staff often determine what is to be done, including whether the death will be reported to an investigating authority. In such instances, the potential for conflicts of interest is unacceptable. Prison staff may call a prison doctor or local doctor to certify the death and make an initial assessment of its cause. A decision about whether further investigation should take place may depend on a doctor’s initial assessment. Classifying a death as natural ordinarily means that there is no investigation, obscuring deaths that might have been caused by maltreatment, poor prison conditions and other human rights violations. However, whether prison-related or not, in the absence of an autopsy, a doctor’s determination is frequently mistaken. Reliable causes of death can only be determined by autopsies. Nonetheless, autopsies are not the norm, let alone by properly trained forensic doctors. This deficiency must be remedied.

As it is presumed that all prisoner deaths are unlawful, the scene of a prisoner death should be regarded as crime scene. A trained forensic doctor should attend the scene, preferably with the body in situ, prior to undertaking an autopsy. An autopsy must be conducted in all cases. Forensic doctors should have access to all evidence and information they deem relevant to their investigation. If a forensic doctor is unavailable, a medically qualified person may be called upon to conduct the autopsy. However, very few non-forensic doctors have experience of autopsies and fewer still have performed one. Their findings are unlikely to be reliable.

Forensic doctors should undertake the processes required to identify the prisoner, especially in cases when the body is not recognizable. The autopsy should identify all external and internal injuries and diseases present. Toxicological, histological, microbiological and other tests may be needed. X-rays and CT scans may be required. If not locally available, arrangements may be made to have necessary tests and examinations done elsewhere, taking care to secure the body of the deceased and any samples taken from the body. With information about the context and circumstances of the death and based on the autopsy findings and the results of other examinations and tests, forensic doctors should make evidence-based conclusions about the cause of the death, the factors contributing to it and the circumstances surrounding the death and should include this information in the final report. All information, findings and conclusions should be recorded, including by photography, so that they are reviewable, that is, another forensic doctor should be able, at another time and place, to come to their own conclusions about the death. Along with the prompt production of reliable results and conclusions, reviewability of autopsies is a key international standard of a satisfactory investigation. Delays of a year or more in producing autopsy results are not acceptable.

The independent investigating authority must be advised of a death immediately. In advance of the arrival of independent investigative officers, prison authorities should prepare themselves to assist. They should secure the scene of every death to preserve evidence. They should gather records and secure all potential sources of evidence. This includes any person involved in or who may have witnessed the death, including prisoners in the same or adjacent wings or cells. Other potential sources of evidence, such as records pertaining to the prisoner who has died, including medical records, visitor logs, closed circuit television records, the deceased’s belongings and other relevant sources of information should be gathered and secured and all such evidence should be made available to investigators.

The independent investigators should further secure and record the scene, take photographs of and film or draw, if necessary, the body and its surrounding, collect and secure exhibits, including relevant records, and interview potential witnesses. Access by independent investigating authorities must be unhindered. The chain of custody recording of all exhibits, including the body of the deceased, must be maintained.

The next of kin of the deceased should be informed of the death of the prisoner at the earliest opportunity, be consulted with and kept regularly informed. All significant developments should be brought to their attention by the independent investigating authority in a coordinated and respectful fashion. The next of kin should also be properly supported throughout the investigation, taking account of their religion and culture. They should be permitted to have a representative present at the autopsy. If they wish to speak with the forensic doctor, either before and/or after the autopsy, they should have the opportunity to do so. Unfortunately, in practice, this is the exception rather than the rule. The next of kin should also have access to all information relevant to the investigations and to any hearings and should be entitled to present evidence.

Upholding the right to life requires that there be full transparency of investigations and accountability for all deaths in prison. If criminal activity is suspected, police and the relevant prosecutorial authority should investigate and pursue domestic criminal justice processes or provide a publicly accessible explanation as to what is taking place. Investigation findings must be publicly available to provide transparency for all prisoners, their families and the public. Investigation findings should be studied to ascertain the nature of the liability of the State and/or others and to ensure there are appropriate consequences. Investigation findings should be studied to inform possible prevention policies and system reform.

Throughout the investigation, all involved must respect the dignity of the dead. Once the investigation is complete, the body of the deceased, including all personal possessions, should be delivered to the next of kin in a compassionate manner, as soon as possible. Appropriate redress must be provided to families.

In drafting the present report, the Special Rapporteur faced difficulties in identifying clear best practices for investigating deaths in prison. Many countries do not have relevant laws or guidance. Where they exist, guidelines leave considerable discretion to prison authorities to determine whether an investigation is necessary. However, some countries have developed detailed protocols for the investigation of deaths in prison.

In India, a judicial inquiry is required every time there is a death in custody. The Indian code of criminal procedure requires that if a person dies in custody, an investigation must be launched, and an inquiry must be held, either by the metropolitan magistrate or judicial magistrate, depending upon the jurisdiction. Regardless of whether the cause of death is natural or unnatural, the National Human Rights Commission must be notified within two months. A videotaped post-mortem (by or under the supervision of a civil surgeon) is mandated for all deaths in custody. Where custodial torture or negligence by prison authorities are found, sanctions are imposed on the person(s) responsible and an independent criminal procedure may be initiated.⁶⁵ While India has a solid legal framework for the investigation of deaths in prison, implementation is unclear.

65. With inputs from interviewee M.D. (India). ↩

In South Africa, the Judicial Inspectorate for Correctional Services is tasked with upholding prisoner rights. It investigates, reports and makes recommendations on the conditions in correctional centres.⁶⁶ Under the Correctional Services Act of 1998, all deaths of inmates must be reported to the Judicial Inspectorate, irrespective of the cause or circumstances. The 2021–2022 annual report of the Judicial Inspectorate details the 46 investigations it conducted during that time, a large majority of which related to deaths caused by suicide, use of force and homicide.⁶⁷ The reports of the Judicial Inspectorate contain significant details, which enable it to identify patterns and suggest steps that might be taken to prevent future deaths.

66. Judicial Inspectorate for Correctional Services, “About us”, available at http://jics.dcs.gov.za/jics/?page_id=116. ↩
67. See http://jics.dcs.gov.za/jics/wp-content/uploads/2022/11/JICS-2021-22-Annual-Report.pdf. ↩

In New Zealand, the Department of Corrections has a clear procedure for investigating deaths in custody.⁶⁸ Immediately after a death in prison, a prison doctor is notified. The police, the coroner, the national corrections office and family members are also informed. Every death is investigated by the coroner and an inspector of corrections; however, autopsies may not be ordered by the coroner in all cases. The prison may also conduct an internal investigation. The police also investigate if they suspect that a death is unlawful. After the coroner’s investigation has been completed, an ombudsman will examine the report of the inspector of corrections, either confirming that the inspector’s investigation was carried out thoroughly and fairly or making additional recommendations. The police may conduct an independent investigation. After hearing the evidence at an inquest, the coroner will make a finding as to the cause and circumstances of the death and make recommendations. The Department of Corrections website includes statistics on deaths in prisons.⁶⁹

68. New Zealand, Department of Corrections, “Deaths in Custody”, available at https://www.corrections.govt.nz/our_work/in_prison/managing_offenders/deaths_in_custody. ↩
69. New Zealand, Department of Corrections, Statistics, available at https://www.corrections.govt.nz/resources/statistics. ↩

In 2023, the Ministry of Justice of Spain, together with the national Council of Forensic Medicine, published a guide on best practices for the medico-legal investigation of deaths in custody, which is based on international human rights standards and forensic best practices, including the Minnesota Protocol.⁷⁰ As set out in the guide, every death in custody should be investigated according to those standards.

70. See https://www.mjusticia.gob.es/es/AreaTematica/DocumentacionPublicaciones/InstListDownload/Buenas%20pr%C3%A1cticas_Muerte%20en%20custodia_ok.pdf. ↩

Comprehensive, reliable data on prisoner deaths are necessary to inform policy and practice and to enable accurate monitoring and evaluation of interventions aimed at preventing prisoner deaths.⁷² In 2019, the Human Rights Council recommended that States implement systems to collect and analyse data on prisoner deaths.⁷³ However, currently, the number of prisoners who die globally and their causes of death are unknown. To the extent that records exist, data are often missing and inadequate, tending to obscure what has happened.74 66 Judicial Inspectorate for Correctional Services, “About us”, available at http://jics.dcs.gov.za/jics/?page_id=116. 67 See http://jics.dcs.gov.za/jics/wp-content/uploads/2022/11/JICS-2021-22-Annual-Report.pdf. 68 New Zealand, Department of Corrections, “Deaths in Custody”, available at https://www.corrections.govt.nz/our_work/in_prison/managing_offenders/deaths_in_custody. 69 New Zealand, Department of Corrections, Statistics, available at https://www.corrections.govt.nz/resources/statistics. 70 See https://www.mjusticia.gob.es/es/AreaTematica/DocumentacionPublicaciones/InstListDownload/Buen as%20pr%C3%A1cticas_Muerte%20en%20custodia_ok.pdf. 71 The Special Rapporteur also wishes to acknowledge the contributions of Roisin Mulgrew, University of Galway, and Philippa Tomczak, University of Nottingham; see Philippa Tomczak and Róisín Mulgrew, “Making prisoner deaths visible: Towards a new epistemological approach”, Incarceration, vol. 3, No. 1 (2019). 72 Stella Botchway and Seena Fazel, “Determining rates of death in custody in England and Wales”, The Journal of Forensic Psychiatry and Psychology, vol. 33, No 1 (2022), pp. 1–13. 73 A/HRC/42/20, para. 65 (c). 74 CAT/OP/MEX/1, para. 173; CAT/OP/MDV/1, para. 115; CAT/OP/BEN/1, para. 223; and A/HRC/18/32/Add.2, para. 54.

72. Stella Botchway and Seena Fazel, “Determining rates of death in custody in England and Wales”, The Journal of Forensic Psychiatry and Psychology, vol. 33, No 1 (2022), pp. 1–13. ↩
73. A/HRC/42/20, para. 65 (c). ↩

There is no standard classification of causes of deaths in prisons.⁷⁵,⁷⁶ Prisoner deaths may be dichotomized into natural and unnatural deaths or, in the medical-legal literature, into natural, accidental, suicide, homicide and undetermined. Some definitions exclude deaths of prisoners transferred to external medical facilities on temporary leave and awaiting trial.⁷⁷

75. Tenzin Wangmo and others, “The investigation of deaths in custody: A qualitative analysis of problems and prospects”, Journal of Forensic and Legal Medicine, vol. 25 (July 2014). ↩
76. Penal Reform International, Deaths in prison: Examining causes, responses, and prevention of deaths in prison worldwide, December 2022. ↩
77. Ghazala Sattar and Martin Killias, “Death of offenders in Switzerland”, Journal of Criminology, vol. 2, No. 3 (July 2005), pp. 317–340. ↩

Prisoner deaths are often highest among disadvantaged and marginalized populations, including First Nations communities.⁷⁸ For example, transgender prisoners are more likely to experience problems in prison, although the rate at which transgender prisoners die is unknown.⁷⁹ This underscores the need to collect detailed, disaggregated data on who is dying, the immediate and underlying causes of death and the contexts in which all prisoners die.⁸⁰ This, in turn, requires that baseline data about prisoners be recorded on admission. Records should be regularly updated and should include activities undertaken by the prisoner, treatments received and incidents that may have occurred. Data collected and recorded should include the prisoner’s sex, gender, sexual orientation, age, language, ethnic origin, disability, religion and any other characteristic that may be a source of vulnerability in prison.

78. Brittany Friedman, “Toward a critical race theory of prison order in the wake of COVID-19 and its afterlives”, Sociological Perspectives, vol. 64, No. 5 (2021), pp. 689–705; Sherene Razack, “Timely deaths: Medicalizing the deaths of Aboriginal people in police custody”, Law, Culture and the Humanities, vol. 9, No. 2 (2013), pp. 352–374; Chris Cunneen, “Aboriginal deaths in custody: A continuing systematic abuse”, Social Justice, vol. 33, No. 4 (2006), pp. 37–51. ↩
79. Caroline Gorden and others, “A literature review of transgender people in prison: An ‘invisible’ population in England and Wales”, Prison Service Journal, vol. 233 (2017), pp. 11–22. ↩
80. OHCHR, “A human rights-based approach to data”, Geneva, 2018. ↩

Classifying prisoner suicides is difficult as this generally requires evidence of suicidal intent, which is difficult to determine without adequate investigation. Suicides may be classified as accidental, undetermined or other, resulting in underreporting and avoidance of further scrutiny.⁸¹ Deaths classified as homicide may obscure contributing factors, such as prison overcrowding. Death classifications may also fail to accurately reflect incidents resulting in mass fatalities, such as inter-prisoner violence, with each death apparently representing a separate incident if no contextual information is included. Similarly, deaths during or related to physical or chemical restraint(s) are often classified as accidental, natural, undetermined or other, masking excessive use of force or poor management of a medicated prisoner.

81. Ingvild Maria Tøllefsen and others, “The reliability of suicide statistics”, BMC Psychiatry, vol. 12, No. 1 (2012), pp.1–11. ↩

Deaths resulting from infectious disease, poor nutrition, exposure or to decaying, filthy and mould ridden cells without sewerage or running water, lack of health care and general neglect may be classified as natural deaths, as may deaths of prisoners due to prior police violence. Also classified as natural or at least not prison-related deaths – or not classified at all – may be deaths that occur shortly after release from prison, although the cause may be due to prison conditions. As noted, deaths occurring 30 days after release from prison should be assumed to be prison related; prison related deaths may also occur after the 30-day period and should be recorded as such.

The UNODC International Classification of Crime for Statistical Purposes offers a model framework for the collection of data on crime that distinguishes between different types of killings and includes acts that lead to death or are intended to cause death. The International Classification contains useful categories of homicide related to social prejudice, sociopolitical agendas and civil unrest. Social prejudice is described as violence against specific social groups, including hate crimes. This is where victims are targeted because of their characteristics, or ascribed attributes, beliefs or values. Characteristics and attributes include sex, gender, sexual orientation, age, language, ethnic origin, disability and race. Beliefs or values, at a minimum, include religious beliefs and economic and social views. The International Classification of Crime for Statistical Purposes provides further disaggregating variables for the identification of the nature of offences, motives, situational contexts, mechanisms of killing and victim and perpetrator characteristics. However, this statistical model does not necessarily address natural deaths and deaths misclassified as natural, where data collection is equally important.

The WHO International Classification of Diseases (11th revision (ICD-11)) is a comprehensive standardized system of classification of human diseases and causes of death, which can be applied to both natural and unnatural prisoner deaths. The WHO system also includes qualifier and extension codes that can be combined with death codes to provide context. This includes codes for the quality of health care, factors influencing health status and external factors that directly and indirectly contribute to prisoner deaths. Thus, coding for a tuberculosis-related death can also capture problems associated with imprisonment. This enables patterns to be revealed, analyses to take place and interventions to be implemented and tested.

Considerable training is required to successfully apply both coding systems. Achieving successful application of one system is challenging enough. It is unlikely that both can be effectively utilized in the prison context. An abbreviated version of these classification systems that captures information on the characteristics of the deceased and perpetrator, if any, and the cause, manner and context of natural and non-natural deaths is highly desirable. Importantly, a new universally applicable data-collection system must generate sufficient information to enable the identification of preventable deaths. Finally, the privacy of any personal information held in prison and medical records must be protected by law and data should be disclosed only when permitted by law or where there is a serious and imminent threat to life or health.

In addition to prison related data collection, it should be noted that only 68 per cent of countries, territories and areas of the world have death registration coverage of at least 90 per cent.⁸² In other words, approximately one-third of the world’s population dies outside a system registering the medical cause of death. It may be assumed that the lack of death registration also applies to deaths in prison. Studies have also found that, without an autopsy, the cause of death provided on a significant proportion of death certificates is wrong.⁸³ Deficiencies in death registration both inside and outside prisons should be remedied.

82. See United Nations Statistics Division coverage of births and deaths registration, available at https://unstats.un.org/unsd/demographic-social/crvs/. ↩
83. See, for example, Lauri McGivern and others, “Death certification errors and the effect on mortality statistics”, Public Health Reports, vol. 132, No. 6 (2017), pp. 669–675; Sangyup Chung and others, “Factors associated with major errors on death certificates”, Healthcare, vol. 10, No. 4 (2022); and Jahanpour Alipour and Abolfazl Payandeh, “Common errors in reporting cause-of-death statement on death certificates: A systematic review and meta-analysis”, Journal of Forensic and Legal Medicine, vol. 82, No. 1 (August 2021). ↩

In his introduction to the 2019–2020 annual report of the South African Judicial Inspectorate for Correctional Services, Justice Edwin Cameron stated:
“Where lives are precious, they are precious everywhere. With rising crime, our country has treated the lives of incarcerated people as though they are not precious. Our correctional system threatens to create a permanent state of exception for those inside – as both physical spaces in society and communities for certain individuals.
“We must ask ourselves to imagine a world in which correctional facilities are rendered unnecessary; a future in which they no longer exist. This may be a utopian hope, but, as we did during apartheid, we must look ahead, and even dream ahead. We need to work and plan as though this world is possible.”

The Special Rapporteur can only echo those thoughts as they have significance beyond South Africa. While prisons remain, the lives of prisoners should be regarded as precious, and the loss of life seen as a tragedy for which the State is, prima facie, responsible. All possible steps should be taken to prevent prison deaths and, if a death does occur, investigations be conducted in a manner that accords with the gravity of the event.

The following recommendations are chiefly directed towards States as the primary actors and bearers of responsibility. Intergovernmental organizations, international agencies, non-governmental organizations, representatives of academia and other stakeholders concerned with the protection of persons imprisoned anywhere should seek to achieve the same ends. Prevention (a) The use of imprisonment must be minimized: the use of imprisonment should be minimized by repealing victimless, poverty-related and minor crimes and mandatory prison sentences; a presumption of release on bail when arrested should be established, sentencing options other than imprisonment should be introduced and a presumption in favour of shorter sentences and parole created; innovative justice options should be adopted, when possible, in place of overburdened conventional criminal justice systems; and the criminal justice system should be resourced and streamlined to reduce delays that produce the overuse of remand and excessive periods of time being spent in custody; (b) If prison conditions breach human rights, imprisonment must not be a sentencing option; if conditions of imprisonment breach human rights standards and norms, including the right to life, imprisonment must not be imposed as a penalty; similarly, if States do not have practical control of their prisons and cannot ensure that human rights standards and norms are observed, States must refrain from sending convicted persons to prison; (c) Prison administrations must protect the rights and welfare of prisoners and prison staff; prison governance must be competent and consistent with the rule of law and human rights; ensuring the dignity and welfare of the prisoner must be the paramount concern; and the human rights of prison staff must be respected; (d) Effective monitoring systems must be implemented; effective internal and independent external monitoring systems are essential; frequent and regular external prison monitoring and unannounced visits must be carried out by adequately resourced expert bodies, functionally independent of government and prison authorities; internal prison monitoring should be proportionate, striking a balance between protecting prisoner privacy and prisoner safety; and States should ratify the Optional Protocol to the Convention against Torture and other Cruel, Inhuman or Degrading Treatment or Punishment; (e) Prison conditions must neither cause nor contribute to physical or mental illness; prisons must be habitable and the prison environment should promote health and be socially inclusive, providing regular access to family and friends, including participation in programmes where social bonds can be made, such as education and work-related and rehabilitation programmes; (f) Prisoners must have access to health services at least equivalent to care provided in the community; prisoners must be examined by qualified medical practitioners and a confidential record made of their physical and mental health status, treatments recommended and potential vulnerabilities addressed (including suicide risk) at admission, during imprisonment and pre-release; prisoners identified as physically or mentally ill or at risk of becoming so should have ready access to medical and psychological care of a standard equivalent to care provided in the community; regular check-ups should take place to ensure that existing and emerging conditions are treated; appropriate health care must be provided for older prisoners, recognizing the challenges of ageing in prison; if a preventable deterioration in prisoner health status is found, the cause should be identified and addressed; given the disproportionately high rate of prisoner suicide, suicide prevention strategies are essential; and harm reduction services offered in community settings should be available to all those in prison; (g) Prisoners must not be asked to undertake activities that put their lives at risk; prisoners should not be induced to participate in activities that may place their well-being at risk, unless the potential benefits outweigh the risks, and all possible steps must be taken to minimize such risks; and prisoners must be fully informed of and comprehend all potential risks; Investigation⁸⁴ (h) Every death in custody must be investigated; (i) Deaths in prison, deaths that occur outside the prison while the prisoner remains under prison authority and deaths of former prisoners that occur within 30 days of release from prison must be fully investigated; investigations should establish the cause, manner, place and time of death, as well as any pattern or practice that may have caused it; investigation findings should be publicly accessible, although the findings may be anonymized; (j) When investigating prison related deaths, the Minnesota Protocol on the Investigation of Potentially Unlawful Death (2016) must be observed; all those involved in investigating deaths in prison, especially prison staff, police and/or other investigators and forensic doctors, must be familiar with, and comply with, the Minnesota Protocol; (k) Forensic doctors must investigate prison-related deaths and an autopsy must be conducted in all cases; forensic doctors must attend the scene of each death with the body in situ and should have access to information they deem necessary; autopsies must be carried out in all cases of prisoner deaths by a trained forensic doctor who has access to suitable facilities, the ability to conduct necessary tests and examinations and other resources that enable the autopsy to be reliably conducted; medico-legal death investigations must be documented so that all findings are reviewable by another forensic doctor at another time and place; undue delays in the production of such reports are in breach of the Minnesota Protocol; (l) Death certificates must be completed by a forensic doctor, a copy provided to the next of kin, the cause of death and other findings explained to the family and deaths registered; (m) The prisoner’s next of kin must be kept fully informed and wellsupported; the prisoner’s next of kin must be informed of the death at the earliest opportunity, consulted with and informed of all developments by knowledgeable persons in a coordinated and respectful fashion; next of kin must also be properly supported throughout the investigation in a respectful manner that takes account of their religion and culture and they should be offered the opportunity to have a representative present at the autopsy; once the investigation is completed, the body of the deceased should be delivered to the next of kin in a compassionate manner at the earliest opportunity; proper reparations should be provided to the family and the family should be advised of the steps being taken to prevent similar prisoner deaths; (n) Unless there are exceptional circumstances, findings of potential criminal liability should lead to prosecutions; findings from death investigations indicating criminal liability should lead to prosecutions of those who had criminal involvement or to a publicly reported decision issued on why a prosecution is not proceeding; (o) Investigation findings must inform prevention strategies; investigation findings that reveal systemic and underlying issues and other prevention opportunities 84 See also, ICRC, Guidelines for Investigating Deaths in Custody (October 2013), annex I, available at https://www.icrc.org/en/doc/assets/files/publications/icrc-002-4126.pdf. must be examined and acted on by the State and all relevant authorities to prevent further deaths; (p) There must be transparency in investigations and responses to findings; responses of the State and other relevant authorities to investigation findings must be publicly reported and followed up by the independent investigating authority; Collection of information (q) Comprehensive, disaggregated data about each prisoner must be recorded and regularly updated; comprehensive disaggregated data about each prisoner must be recorded upon their admission and kept up to date; data collected and recorded should include the prisoners’ sex, gender, sexual orientation, age, language, ethnic origin, disability, religion and any other characteristics that may prove to be a source of vulnerability in prison; records should be regularly updated and include activities undertaken by the prisoner, treatments received and incidents that may have occurred; (r) Comprehensive, disaggregated data on all prisoner deaths must be collected; a universal, easy-to-apply classification system should be developed and applied to natural and non-natural prisoner deaths, which should include detailed disaggregated data about the deceased (and the perpetrator, if any), the situational context and cause and manner of death; deaths should include those occurring in prison and outside prison if prisoners remain under the control of the detaining authority, as well as deaths occurring up to 30 days after release from correctional institutions or those occurring later as a result of illness or injury acquired in prison.

84. See also, ICRC, Guidelines for Investigating Deaths in Custody (October 2013), annex I, available at https://www.icrc.org/en/doc/assets/files/publications/icrc-002-4126.pdf. ↩