The effect of supervision, health and environmental interventions on the health of prisoners in isolation: systematic review

Introduction

Internationally, there are different definitions of isolation. Common to all is the starting point of isolation as a serious, invasive and harmful measure. The UN has established specific time limits to limit the most harmful forms of solitary confinement. They define isolation as the confinement of an inmate for 22 hours or more per day without meaningful human contact, and extended isolation defined as isolation that lasts for more than 15 consecutive days. Norwegian legislation has a number of provisions that allow for isolation or limiting the social contact of inmates. Based on Section 37 of the Execution of Sentences Act, the Norwegian Prison and Probation Service decides that an inmate can be completely or partially excluded from social contact with other inmates if it is necessary to prevent inmates from influencing the prison environment in a particularly negative way. This may be to prevent inmates from harming themselves or others or making threats against others, to prevent significant material damage, to prevent criminal acts or to maintain peace, order and security.

The municipality is responsible for providing necessary health and care services to prisoners, and are to supervise prisoners in isolation, while the specialist health service shall ensure that necessary specialist health services are provided. The municipality's responsibility includes all patient and user groups, including convicted persons and prisoners in custody with somatic or mental illness, injury or disorder, substance abuse problems, social problems or reduced functional capacity. In spring 2026, the Storting adopted legislative amendments establishing supervision of isolated prisoners as a statutory task for the municipal health and care services.

Solitary confinement is used in Scandinavian prisons, and previous research on solitary confinement in prison has shown that solitary confinement can cause health damage to the prisoners. In the Parliamentary Ombud's special report to the Storting on isolation and lack of human contact in Norwegian prisons, it is recommended to revise the Directorate of Health's guide "Health and care services for prisoners in prison", to ensure the identification of isolation injuries and the care of isolated prisoners. It is therefore important to look at which health and environmental measures can help prevent such injuries or reduce the risk of health injuries in prisoners.

Objective

The purpose of the research review was to examine the effects of supervision, environmental and health interventions and measures that are intended to prevent or reduce the risk of health injuries or ease isolation for prisoners. This could provide a more solid evidence base for training health personnel and employees in prisons. The review was to answer the question: What effect do supervision, health and environmental interventions have on the physical and mental health of prisoners in isolation?

Method

We conducted a systematic review. We included studies that investigated the effect of interventions that in various ways were intended to prevent isolation injuries or ease the burden of prison isolation, applied to both inmates in imposed and self-selected isolation. Relevant measures or interventions were measures that examined the frequency and type of supervision, health measures and environmental measures given to prisoners in isolation. Health measures were measures directly related to health or health status, such as mapping or observation of health status, measuring blood pressure, nutritional and fluid intake monitoring, and similar activities. Environmental measures were measures that were intended to ensure meaningful human contact and/or reduce social and sensory deprivation. The work was based on a project plan that was prepared and published before the work on the report began. We searched the following seven databases: Epistemonikos, MEDLINE, PsycINFO, EMBASE, Sociological Abstracts, Scopus and Web of Science. In addition, we performed searches in other sources. We included studies from 2015 to 2026, from countries in Europe, Australia, Canada and the USA, written in Norwegian, Swedish, Danish or English. We assessed the risk of systematic bias in the studies using checklists adapted to the different study designs. Due to the heterogeneity of the studies, we did not perform meta-analyses, rather we presented the results narratively.

Results

The database searches yielded 4529 hits before removal of duplicates. After removal of duplicates, there were 2501 references. We also considered approximately 450 studies from searches in other sources. We considered 24 full-text publications. We ultimately included 6 studies distributed across 6 publications, all from the USA. The studies were published between 2016 and 2025. We included one randomized controlled trial, two quasi-experimental studies and three before-and-after studies. We only identified studies on environmental interventions. The studies had a total of 950 participants, mainly adult men. There was great variation in the type of intervention and outcomes measured. One included RCT was assessed as having some concerns about the risk of bias. The two quasi-experimental studies were assessed as having some concerns and a high risk of bias, respectively. Of the three before-and-after studies, two studies were assessed as having some concerns about the risk of bias and one as having a high risk of bias. We focused on three of the studies that had a control group in the analysis. Overall, the review found that there are no clear or consistent effects of environmental measures for prisoners in solitary confinement. Measures aimed at prisoners in solitary confinement have limited and varying effects, with small effects on short-term and situational factors. For rule violations, the studies showed relatively small effects and divergent results, and there is no clear effect of the interventions. For psychological functioning, work alliance and criminal thinking, no clear effects have been found. Although some studies point to possible improvements on more stable conditions, such as criminal thought patterns and serious mental disorders, were consistently small and fraught with great uncertainties.

Discussion

The evidence answers the research questions to a limited extent. We included studies investigating effects of environmental interventions of on the health of prisoners in isolation. We identified no studies that answered the research questions about the effects of supervision and health measures for prisoners in solitary confinement with a focus on preventing or reducing the risk of health damage or to ease the isolation of prisoners. Furthermore, we identified no studies from Norway or the Nordic countries, no studies that examined the effects of measures aimed at women or youth. We can say little about how interventions work when it comes to different types of isolation, in terms of cause, length and severity.

The applicability of the knowledge base has clear limitations. The studies have been conducted in real prison contexts, which provides practical relevance. However, the studies have been conducted in American prisons, which differ from the Norwegian correctional service in terms of, among other things, framework conditions, sentencing conditions and use of solitary confinement. This limits direct transferability. In addition, methodological limitations and uncertain effect estimates mean that the findings should not be used as a basis for specific measures or decisions.

Overall, the evidence in this review provides indications rather than definitive answers. Some measures may have a positive effect, but it does not provide a sufficiently solid basis to conclude clearly about the effect on either mental or physical health in isolation.

There is a need for research with a stronger design, especially studies with control groups, longer follow-up periods and greater emphasis on physical and mental health outcomes and contextual factors. Until this is available, practice should be developed based on available knowledge, combined with local adaptation and systematic evaluation.

Conclusion

There is very little research on the effects of supervision, health and environmental measures during isolation that are intended to prevent or reduce the risk of health damage or ease isolation for prisoners. This review shows that the evidence is very limited and inadequate, with few studies, methodological limitations, and small samples from contexts that can only be partially compared to the Norwegian one. We found that measures aimed at prisoners in isolation have limited and varying effects, with some small positive effects on short-term and situational outcomes. The overall evidence base shows that there are no clear or consistent effects of environmental interventions for prisoners in solitary confinement on the investigated outcomes. More research is needed, especially studies with stronger designs, longer follow-up periods, and greater emphasis on physical and mental health outcomes and contextual factors.