Research Lead:
Jonathan Luke Austin
jla@ifs.ku.dk

Mental health in detention is a global crisis — fueling cycles of recidivism, poverty, unrest, and trauma. Can design and technology break this cycle? Even in the harshest conditions — humanitarian crises, armed conflict, and settings where mental health is often not a priority?

1. introduction

While political and military cataclysms are dreadful and challenge the mind through the monstrosity of their violence (that of a concentration camp or of an atomic bomb), the shattering of psychic identity, whose intensity is no less violent, remains hard to perceive.

– Julia Kristeva, Black Sun: Depression and Melancholia

At least one in seven people who are incarcerated have a severe mental illness.¹Emilian, C., Al-Juffali, N., & Fazel, S. (2025). Prevalence of severe mental illness among people in prison across 43 countries: a systematic review and meta-analysis. The Lancet Public Health, 10(2), e97-e110. And at least 50% have a history of mental illness more broadly. These rates are double those of the general population. More, evidence suggests that this figure is higher in low and middle-income countries (LMICs), where infrastructure for the diagnosis and treatment of mental illness is very limited even outside prisons.²Ibid. Indeed – generally – between 76% and 85% of people with severe mental disorders receive no treatment for their disorder in LMICs. In prisons, this number is likely much higher.³WHO. 2019. “Mental Disorders Fact Sheet.” World Health Organization. https://www.who.int/en/news-room/fact-sheets/detail/mental-disorders. As such, prisoners with mental illness are a hidden population in multiple ways: stigmatized because of both their detention and their illness.

The prison is a microcosm of society; mental health problems therein are thus not contained within its walls. On the contrary, a psycho-social carceral continuum exists.⁴Wacquant, Loïc. 2001. “Deadly Symbiosis: When Ghetto and Prison Meet and Mesh.” Punishment & Society 3 (1): 95–133. Poverty causes mental illness; institutional weakness leads those suffering from mental illness to be directed to prison; upon leaving prison, re-offending is common; criminal activity in urban neighborhoods increases; risk of engagement in armed violence increases; other forms of violence (intimate partner, child neglect), etc. increase, and so forth. The spiral of this carceral continuum is thus stark and a key global public health concern yet unaddressed.

But mental health screening, triage, and referral systems are absent in prisons or detention settings in both low-income, crisis-affected, or humanitarian settings and many high-income contexts. Levels of psychiatric bed provision in many LMICs have stagnated or declined since 1990. While the prevalence of mental health and other issues has not declined globally for over 25 years.⁵Nadkarni, Abhijit, Charlotte Hanlon, and Vikram Patel. 2024. “Mental Health Care Models in Low- and Middle-Income Countries.” In Tasman’s Psychiatry, edited by Allan Tasman, Michelle B. Riba, Renato D. Alarcón, César A. Alfonso, Shigenobu Kanba, Dusica Lecic-Tosevski, David M. Ndetei, Chee H. Ng, and Thomas G. Schulze, 3347–93. Cham: Springer International Publishing, 3348. One response to this reality has been a rapid – yet contested – turn towards technological solutions.

Key Facts

• At least 1 in 7 incarcerated individuals suffer from a severe mental illness;
• At least 50% have a history of mental illness more broadly;
• These figures are higher in low and middle-income countries, and in situations of humanitarian crisis;
• In low and middle-income countries between 76% and 85% of people receive no treatment for mental illness;
• Levels of psychiatric bed provision in low and middle-income countries have stagnated or declined since 1990;
• The prevalence of mental illness has not declined globally for over 25 years.

2. Contested Technological Turns

Mental health care within high-, middle-, and low-income countries is undergoing a rapid technology-driven transition. These shifts are suffused with the general risks facing the deployment of technology to solve social problems: hype over emerging technologies, data protection risks, technocolonialism, the standardization of human life into algorithmic patterns, rampant commercialization, and more.⁶Madianou, Mirca. 2024. Technocolonialism: When Technology for Good Is Harmful. John Wiley & Sons. These issues are especially controversial vis-à-vis mental health, where the ‘immateriality’ of the phenomenon – the inability to objectively fix what mental health or mental illness actually is, given the complexity of its social, political, environmental, and biological drivers are – make the risks of technological solutionism very acute.⁷Harvey, Philip D., et al. “Technology and mental health: state of the art for assessment and treatment.” American Journal of Psychiatry 179.12 (2022): 897-914.; Berners‐Lee, Ben. “Reconciling healthism and techno‐solutionism: An observational study of a digital mental health trial.” Sociology of Health & Illness 46.1 (2024): 39-58.; Rivera-Segarra, E., Mascayano, F., Alnasser, L., van der Ven, E., Martínez-Alés, G., Durand-Arias, S., … & Susser, E. (2022). Global mental health research and practice: a decolonial approach. The Lancet Psychiatry, 9(7), 595-600; Lovell AM, Read UM, Lang C. Genealogies and Anthropologies of Global Mental Health. Cult Med Psychiatry. 2019 Dec;43(4):519-547. doi: 10.1007/s11013-019-09660-7. PMID: 31729686; Mills C. Strategic universality in the making of global guidelines for mental health. Transcult Psychiatry. 2023 Jun;60(3):591-601. doi: 10.1177/13634615211068605. Epub 2022 Jan 19. PMID: 35043746; PMCID: PMC10486146.

Despite these concerns, the deployment of technology in this area has become more-or-less inevitable for several key reasons:

1. Medical resources – infrastructural, financial, logistical, and human (see below) etc. – to diagnose and treat mental health conditions in detention are severely lacking globally and evidence suggests that digital solutions are effective in filling this gap;
2. Staff training to diagnose and treat mental health conditions is especially limited in humanitarian and detention settings in both LMICs and HICs, but again evidence suggests technology can help;
3. The complexity of mental health conditions, which require ‘trans-diagnostic’ approaches that consider the prevalence of comorbidities and socio-ecological (as opposed to physical) aetiologies, etc. is a problem that is possibly especially amenable to technological intervention.

Given all this, emerging technologies are seen as having the potential to work as cost-effective, scalable, and context-sensitive tools to improve mental health conditions in prisons and detention across the world. _­Consider the following remark from a Nigerian forensic psychiatrist:

Akin to other international domains, there is an urgent need to expand the uniformity and coverage of psychiatric screening programmes in correctional environments in Africa, AI could offer viable solutions for this, though at the time of writing, AI-driven screening remains untested within African prison settings.⁸Ogunwale, A., A. Smith, O. Fakorede, and A. O. Ogunlesi. n.d. “Artificial Intelligence and Forensic Mental Health in Africa: A Narrative Review.” International Review of Psychiatry, OnlineFirst, 2.

The key words here are “could offer… solutions.” Technology has been deployed in the realm of mental health in and beyond LMICs for decades. But it has been harder to deploy these technologies in ways that are effective: producing tangible change in different scalable contexts across the world. The turn towards ‘Artificial Intelligence,’ for instance, is simply the latest over-hyped solution to the problems identified above. There are a series of serious demand- and supply-side barriers to technology in this area, barriers which must first be fully taken in to account to produce positive rather than simply ad-hoc palliative change in this area.

3. HUD’s Approach: Malembe, Malembe

How can we work-with technology, yet mitigate its evident-risks, to improve mental health care in prison and detention settings in humanitarian contexts? At one level, it is clear that we need to develop a multi-sectoral, socio-ecologically embedded, participatory, and context-sensitive approach to developing technological innovations in this area. But how to get there? What do these buzz words really mean for us?

Given the range of social and technical problems to be addressed, HUD begins with the Lingala mantra in the Democratic Republic of the Congo that one should deal with things malembe malembe – slowly, slowly – calmly, calmly. Against the dream of all-encompassing technical solutions, we must first take a step back, returning to fundamental user-needs before making technological choices. While AI may play a role, it is not useful to start thinking about AI yet. Instead, more basic obstacles must first be overcome.

Indeed, one of the main blockages to mental health care in detention settings rests in the simple fact that detainees are rarely effectively screened for mental health conditions upon their detention, during their detention, and after their detention. There are few real-time diagnostic tools that can assess the impact of mental health in detention settings, and even fewer in settings of armed conflict and humanitarian crisis. But without having this second order capacity to screen defined populations for mental health and psychosocial issues it is far more difficult to mobilize direct primary care interventions of any kind and/or to advance tertiary enablers of medical treatment that require such granular data on the scope, extent, and precise nature of a problem.

A focus on a low-intensity intervention at the level of screening and clinical decision support is thus a slow yet critical and pragmatic starting point: something that may increase the core capacity to screen, diagnose, and triage MHPS issues within detention settings. Indeed, working to develop a clinical decision support system in this area has the potential to scale to improving the quality of care, the control of attrition rates, rehabilitation services, etc., but also has virtues even without these other benefits necessarily materializing.

More, working at this low-intensity level is important socially and politically. Efforts to deploy complex technologies in this area are criticized not only because of the lack of contextual fit they demonstrate in both HICs and LMICs but because they naturalize a particular understanding of mental health. Indeed, the very idea of a ‘mental health gap’ as it is currently understood has been criticized for emerging from a biomedical understanding of mental health that might miss not only the real ‘scale’ of this gap but also where it should be identified more precisely and, in particular, its socio-political content.⁹Littoz-Monnet, Annabelle. 2022. “Exclusivity and Circularity in the Production of Global Governance Expertise: The Making of ‘Global Mental Health’ Knowledge.” International Political Sociology 16 (2).

Related to this, working at this slow, careful, level has other potentially radical benefits. It is not possible to disassociate both contemporary biomedical paradigms and modern forms of incarceration from their colonial roots. Kinshasa’s central prison, Makala, was constructed by Belgian colonialists in 1957. Not far away, the Belgians also constructed a new sanatorium. Both facilities are still in-use, little changed, and their architectural designs are mirrors: both follow an architectural structure whose design explicitly sought to enforce forms of racial segregation, under the guise of good medical-carceral practice.¹⁰Nys-Ketels, Simon De. 2021. “A Hospital Typology Translated: Transnational Flows of Architectural Expertise in the Clinique Reine Elisabeth of Coquilhatville, in the Belgian Congo.” ABE Journal. Architecture beyond Europe, no. 19 (December).; Nys-Ketels, Simon De. 2023. “Myths and Realities of the Belgian Medical Model Colony: A Genealogy.” ABE Journal. Architecture beyond Europe, no. 21 (July).

The roots of biomedical understandings of medical health are – therefore – situated in a desire to limit the decision-making powers of the disenfranchised, as has been widely discussed in social theory. There is thus a further good reason to return to considering how technological change in this area might be leveraged to return different forms of decision-making power to these actors. Moreover, such a step – if integrated into technology design processes – would allow us to see sites like the Congo, Colombia, or elsewhere as new resources for global change. Typically, these contexts remain sites of experimentation – in which corporations abuse the lack of power of communities to expand the scope of biomedical paradigms. Yet, the failure of these approaches even within HICs, where both prisons and mental health care are in equal crisis, suggests that intervening in this area may allow for a paradigm-shift driven by LMICs themselves.

4. Co-designing pathways forward

HUD has begun exploring mental health in detention through, first, a bottom-up needs-assessment that is driven by the insights of actors in the Democratic Republic of the Congo with decades of experience addressing these issues in one of the most complex humanitarian settings in the world. Our goal is to explore the role of technology in addressing mental health in detention and highlight the global relevance of Congolese expertise.

Following this, HUD’s Jonathan Luke Austin and Rachel Howell are organizing a series of co-design workshops drawing together medical professionals (doctors, nurses), civil society members who conduct monitoring visits to prisons and detention centres, and Congolese authorities with the goal of identifying pragmatic, actionable, and scalable interventions for mental health screening in detention settings. The DRC has experienced armed conflict for decades, with severe and deeply negative consequences for its detention system. In one sense, it is a ‘worst-case’ scenario for 1) improving mental health, generally, 2) doing so in conflict-affected detention settings, specifically, and 3) introducing technological interventions. Nonetheless, our co-design process is revealing how this seemingly worst-case scenario offers key lessons on how humanitarianism might integrate evidence-based and context-specific MHPSS programming for detention.

In particular, our co-design work is highlighting how the DRC is an apt context through which to critique existing initiatives in this area, demonstrating how technological interventions into mental health (such as the WHO’s digital app version of its ‘mhGAP’ program) rarely ‘fit’ with sites of humanitarian crisis at both technical and socio-political levels. Second, we also show how contexts like the DRC also nonetheless hold productive potential to provide innovative (technological) alternatives that move beyond the biomedical understandings of mental health still prevalent in Europe and North America. Indeed, contexts like the DRC offer an opportunity to ‘prototype’ alternative technological (or not) approaches to detainee mental health in conflict-affected settings beyond the traditional approach of first piloting in less “complex” settings and, instead, prototyping in a more complex socio-economic and political space that provides the potential for a form of ‘reverse innovation’ of relevance far beyond such a worst-case scenario.

In the coming months and years, we will draw on partnerships with Congolese experts to begin co-designing technological interventions in this area. In doing so, our goal is not only to improve mental health conditions in Congolese detention settings but to create solutions that can scale beyond that context.