The impact of immigration detention on mental health: a systematic review

Studies were included if they (a) included male or female adults or children detained for immigration purposes; (b) reported the prevalence of mental health problems; and (c) presented results of published peer-reviewed research based on one or more of the following study designs: cross-sectional survey; case–control study; cohort study; case series analysis; experimental study with baseline measures for the outcomes of interest; or secondary analysis of organisational records (Table 1). No restrictions were placed on country setting, country of origin of the immigration detainees, or method of measuring mental health outcomes. Since research in this area remains scarce, studies were not excluded on the grounds of their quality (though this is reported in Table 2). Qualitative studies will be presented in a subsequent paper.

Studies that did not report specifically on people held in immigration detention or removal centres (i.e. facilities detaining foreign national citizens for administrative purposes under immigration powers) were excluded. Studies that did not examine the impact of detention on mental health specifically were also excluded. Therefore any studies examining multi-morbidity involving both mental health disorders and physical disorders (e.g. infectious diseases) were excluded, as were studies focused on the attitudes and experiences of health professionals working within immigration detention. Reviews, grey literature, commentaries, conference abstracts, letters, editorials, books and short surveys were also excluded. We have however referred to relevant grey literature in the introduction and discussion sections.

Studies concerning child and adult detainees have been presented separately for the sake of clarity.

Relevant studies were identified through electronic searches of Embase (1980 to 2018 week 22), Ovid MEDLINE (1946 to May week 3 2018) and PsycINFO (1806 to May week 3 2018). Searches were limited to English language only. No other restrictions were applied to the searches. All search terms related to the following two main areas: mental health and immigration detention. The full electronic search strategy is presented in Additional file 1.

Electronic searches were supplemented by screening of reference lists of all included primary studies. In addition, the following journals published between 2013 and 2018 were hand searched: Forced Migrant Review, International Migration Review, International Migration, Journal of Refugee Studies, Journal of Traumatic Stress, Refuge, and Torture. Six International experts working in the field of forced migration and health (i.e. psychiatry, psychology and criminology) were contacted to nominate further papers eligible for inclusion. One response was received.

Three reviewers (CM, RS, ZC) were involved in the process of independently screening titles and available abstracts of identified articles. If it was unclear whether a reference was relevant for the review, texts were retrieved in full. The same team of reviewers (CM, RS, ZC) independently assessed full texts to verify eligibility based on the above selection criteria. Disagreement was resolved either by discussion or with assistance of a third reviewer (CK, KR or MW). Data from all included studies was double extracted by a team of four reviewers (CM, MW, RS, ZC) into a standardised electronic spreadsheet. The following information was extracted from each study: first author, year of publication, population sampled, country where study was carried out, study design, measures used to assess mental health or psychological domains of quality of life, outcomes and length of immigration detention. The quality of studies which met the inclusion criteria was appraised independently by two from a team of three reviewers (CK, CM, MW) using a modified version of the qualitative Critical Appraisal Skills Programme Checklist [35].

Each item was rated as high (2), intermediate (1) or low (0) quality, resulting in a maximum potential total score of 26 (see Table 2 for breakdown of scores). Discrepancies in scores between reviewers were resolved by discussion.

A total of 17 studies reported data on the mental health consequences of immigration detention in adults. These studies involved a total of 1168 participants (Note: this number only includes studies in which primary data was collected). Three studies on parental health are discussed separately in the ‘Child and Families’ subsection below. Six studies were conducted as comparison studies, with a total of 331 non-detained controls. Ten studies were conducted during detention, six were conducted following detention, and one included both current and past detainees. Reasons given for detention varied widely and were not consistently reported. Many were asylum seekers and some studies also included foreign national prisoners or ex-prisoners awaiting deportation, illegal workers and undocumented entrants. Participant ages ranged from 15 to 80 years of age. Approximately two thirds of all participants were male. One study focused exclusively on females, who were illegal sex workers detained in Israel and awaiting deportation [37]. Studies differed widely in the participants’ duration of detention. Some studies reported average duration in terms of the mean, others reported median durations. For those reporting mean durations, these ranged between 4 days to 3.2 years. Reported median durations ranged between 6 to 8 months. Reported ranges fell between 2- to 4 months. Various pre-migratory traumatic events were reported, including but not limited to torture. Where reported, subjection to torture ranged from 12.2–100%.

Overall, 11 studies employed self-report measures, one used self-report measures along with clinical assessment, one used clinical assessments only and four analysed health records. These studies are reviewed below in terms of method of measuring mental health outcomes (self-report and clinical assessments), use of a control design, detention duration and release, and gender.

Ten studies reported data on the mental health of children and/or families within the context of detention. This review reports on 629 children and young people (across 9 studies) and 302 parents (across 5 studies). Two studies were conducted as comparison studies, with a total of 96 controls. Seven studies were conducted while participants were held in detention, and three were conducted following detention. Again, detention duration was differentially reported. Where reported, the means fell between 6 months and 2 years and 4 months and the ranges spanned from 11 days to 2 years and 8 months across studies. The ages of children spanned from childhood to adolescence, ranging from 11 months – 19 years. Both male and female children were represented, yet gender distributions were largely unreported.

Overall, 5 studies used either self-report or parent/teacher report measures Two studies combined clinical assessment with self-report measures to evaluate psychiatric vulnerability [23, 52]. This allowed not only for an estimation of prevalence rates of psychiatric disorders, but also for retrospective accounts of mental health difficulties preceding detention. One study reported clinical outcomes of referrals into Child and Adolescent Mental Health Services (CAMHS) from a detention centre [53]. Two further studies reported on data collected as part of an Australian Human Rights Commission Data Set [48, 54].

A wide range of psychological disturbances mirroring and extending those of adults was found across all three studies which also all included clinical assessments (see Table 4). All children evidenced at least one psychiatric disorder, most frequently depression, anxiety, PTSD and somatization, depending on the diagnostic categories used. The overwhelming majority also struggled with sleeping (65–100%) and eating problems (100%), suicidal ideation (50%), and self-harm (25–80%). Drawing on data from telephone-administered psychiatric interviews and self-report measures from a near-complete sample of detained families of unspecified ethnicity, Steel et al. [23] found psychiatric disorders to be ten times more likely to occur subsequent to detention when compared to prior to detention. Although based on retrospective reports, such difficulties appeared to be specifically linked to detention experiences. Combining clinical assessments with parent-report and, where possible, triangulating these with self-report measures of mental health in a small sample of child and adolescent detainees, Lorek et al. [52] further identified age-related differences in mental health difficulties. Specifically, whilst elevated levels of emotional and behavioural problems and somatic complaints were common across all age groups, developmental delay and regression were more apparent in younger detainees (aged 3–6). These included frequent crying and withdrawal (100%), disturbed sleep routines (100%), language delays (50%,), loss of previously acquired cognitive skills (33%) return to nappies (38%), enuresis (13%) and encopresis (13%). Amongst older children (aged 7–11), self-report questionnaires suggested high levels of depression (100% not within normal range, 50% above the cut off for clinical depression), anxiety (66% above cut off for clinical anxiety) and PTSD symptomatology (17% approaching clinical cut off) [52]. Though this study did not have a control group, the reported lack of pre-existing mental health problems again suggests that sudden mental health deterioration may be attributable to detention experiences.

Similar developmental differences in psychiatric morbidity were also found in the study of CAMHS assessments of 10 detained families [53]. Developmental concerns such as language regression (50%), attachment problems (30%) and enuresis (30%) were reported in children under 5 years of age, whereas symptoms of PTSD (100% above clinical cut off), depression with suicidal ideation (100% above clinical cut-off), self-harm (80%) anxiety (70%), and somatic symptoms (50%) were more common in older children (aged 6–17).

Regressive behaviours such as enuresis (45% boys, 48% girls) and encopresis (8% boys, 11% girls) were also self-reported by a group of adolescents referred into an infirmary during their time in detention in one study, alongside the more common symptoms of severe to very severe PTSD (very severe 94% boys, 96% girls; severe 6% boys, 4% girls; Rothe et al., 2002a). Following release from detention, 57% of adolescents within the same detention facility were considered to continue to meet a diagnosis of PTSD [56].

A secondary analysis of an Australian Human Rights Commission data set indicated very high rates of severe psychological distress amongst adolescents, evidenced by severe disorders for 85.7% of adolescents [54]. All met criteria for mixed anxiety and depression. Additionally, parent reports suggested a high probability of psychiatric disorder in 75% of their children, and significant emotional symptoms int 78.6%. Drawing on a different subset of the collected data, it was estimated that a third of detained children had sufficient clinical symptoms to necessitate tertiary outpatient assessments [48].

Adverse mental health consequences of immigration detention are consistently recognised across the literature. Such findings prevail even in countries where detention standards are regarded as relatively benign [46, 52]. Much of the clinical literature reports high levels of anxiety, depression and PTSD and poor quality of life. Whilst such mental health difficulties cannot be viewed in isolation from past histories and pre-detention traumas, which are not consistently measured across studies, controlled studies with non-detained controls uniformly suggest greater symptoms and ‘caseness’ (i.e. meeting diagnostic criteria for a specific mental health condition) in detained samples. This indicates that detention plays an independent role in contributing to poor mental health outcomes amongst asylum seekers. Nonetheless, being an asylum seeker or having greater trauma exposure of any kind (whether torture or other exposure) prior to detention seems to be associated with higher rates of anxiety, depression and PTSD in the context of such detention (e.g. [41]). Similarly, greater isolation from families and living alone has been found to increase such symptoms in immigration detainees (e.g. [43]).

All adult studies examining the association between detention duration and mental health severity (N = 5) demonstrate a significant relationship between detention duration and mental health deterioration. No such correlation was found for length of detention and quality of life (N = 2), although a negative trend was found in one of these studies [46]. Additionally, perceived support from detention staff was found to impact positively on quality of life, highlighting the importance of staff behaviour on the wellbeing of detainees but not necessarily having any effect on their mental health problems. Symptom severity has been found to decrease and quality of life to increase following release, but mental health difficulties persist well beyond release. Despite the prominence of mental health disorders amongst detainees, significant barriers to the identification and treatment of those needs remain [47].

The comparatively smaller body of research on detained children and families (N = 10 and a total of 629 children studied) echoes the findings in adults summarised above. It should be noted however that the detention of children is fairly uncommon across several countries including the UK, yet remains relatively widely practiced in some European countries [60] and, as recently widely publicised, in the USA. Across those settings and jurisdictions where it remains permitted, profound and far-reaching mental health difficulties are found amongst detained children and young people. Only one in four studies found the severity of mental health difficulties to increase with time the children spent in detention [53]. The other three studies found no correlation between detention duration and mental health. While this has not been researched, this may be explained by the possibility that even small durations of detention are traumatic and harmful to children. The nature of the mental health difficulties found differs across age groups; whereas younger children display developmental regression and primarily externalising behaviours, symptoms of depression, anxiety and PTSD are prominent in older children. Psychosomatic symptoms or somatization appear to be common across developmental stages. This may reflect the difference in experiencing and communicating psychological difficulties in children and young people [61, 62], but may also be understood as a reflection of the somatic component within the symptom pattern of complex PTSD [63]. Whilst detention itself may compromise child development and psychosocial health, as well as aggravating existing trauma, it may also lead to a breakdown of family units. The detention environment may therefore undermine adequate parenting abilities and leave parents feeling disempowered, as they struggle with their own mental health difficulties. Taken together, this may prevent parents from being able to provide support necessary for healthy child development, leaving children at risk of adverse mental health outcomes [64]. It remains unclear however, whether detaining children together with their parents or forcibly separating children from their families is more harmful to the children’s mental health. Further, it is currently unknown how rates of parental mental health difficulties compare to those of single adults and those whose families are not detained. Finally, there is a dearth of quantitative research on the mental health of unaccompanied minors who are detained. With the current number of unaccompanied minors rising across the globe, research into the effects of immigration detention on this subgroup of children is of high importance, particularly with view to the possibility that some unaccompanied minors may be inappropriately treated as though they were adults, consequently increasing their risk of exploitation and harm [65].

Overall, the findings suggest that detention exacerbates the mental health burden of asylum seekers and refugees and that such detention should be viewed as a traumatic experience in and of itself. This may be particularly true for those detainees who are particularly vulnerable prior to detention. Individuals with pre-existing mental health difficulties may also be less effective self-advocates and therefore more likely to be detained [11], despite recommendations and policies to the contrary [31].

A comprehensive search strategy, as well as independent screening, data extraction and quality appraisals were applied to ensure a systematic approach to synthesising the literature on the mental health consequences of immigration detention. Adherence to the PRISMA guidelines further increased the methodological rigour and clarity of the review. We sought to ensure that all research that met out criteria was identified through additional contact with expert researchers in the field. Finally, the exclusion of grey literature, as well as confining the search to papers written in English, inevitably resulted in the exclusion of some papers which may have limited the overall understanding of detainee mental health.

Numerous methodological problems exist at primary study level, which may restrict the conclusions that can be drawn in this review but also indicate where further research is needed. The evidence to date remains sparse due to the multiple obstacles hindering such research. These include restricted access opportunities into detention facilities as well as the prohibiting of random sampling (e.g. [38, 44]). This may have led to an increased reliance on qualitative rather than quantitative methods, particularly in the study of detained children. Such studies have not been included in this review. With regards to studies conducted within detention centres, participants are often recruited through community areas within the detention centres (e.g. [11]), possibly introducing a bias towards those detainees who are most healthy. Recruitment of participants in studies conducted following release from detention, often occurs through lawyers, community organisations or other support services (e.g. [41]). Most studies did not describe the detention regime in sufficient detail to allow for this to be considered as a potential influencing factor. This, together with clearer differentiation between the traumatic effects of detention itself from those of the detention environment) is an important focus for future research. It also carries the implication that future studies should provide as much information as possible about the detention environment in order to facilitate judgment as to the generalisability of the findings. Participation rates are often unknown. It is therefore possible that the recruited samples differed systematically from detainees who did not access community facilities or come into contact with support services. Samples may therefore represent only the most vulnerable individuals in need of support, or conversely, the most highly functioning who are able to access appropriate services. Further research in more representative populations is needed to allow more confident generalisation of the findings to the detainee population as a whole. Additionally, the probability of being detained depends to some extent to sex and nationality, resulting in predominantly male populations of certain national origins [38]. Whilst a broad range of nationalities is represented across studies, a paucity of research focusing on women is evident. Differing nationalities contribute to the vast heterogeneity of samples, along with varying past experiences and traumatic histories, migration statuses, and detention systems and immigration policies within host countries. Correspondingly, language may be an important complicating factor, acting both as a prominent exclusion criterion or a possible cause of misinterpretation or inability to express mental health difficulties clearly [47]. Communicating mental health problems may additionally be compounded by differing expressions of distress and trauma, alongside differing attitudes and experiences of mental health within countries of origin [67]. However, participants from a wide range of cultural backgrounds seem to endure similar psychological ill-health when subjected to detention.

Multiple designs as well as a diversity of methodology and instruments used are evident across studies. Whilst these may differentially contribute to the overall understanding of mental health problems amongst detainees, they also limit comparability. The majority of research is conducted as cross-sectional studies, which warrant caution in the drawing of casual inferences. The suggested causal link between detention and worsening of mental health is however consistent with a limited number of longitudinal studies and comparison studies, implying that detention not only exacerbates existing mental health disorders, but also contributes independently to the onset of new ones. Nonetheless, isolating the effects of detention alone remains a complex task and additional longitudinal research is needed in order to establish the mental health trajectories of detainees, from the very early days of being arrested and detained. This may also include investigating the adverse effects of being detained multiple separate times at different lengths, as well as the effects that experiences in detention may have on people’s ability to navigate and integrate into new societies. Finally, the frequent reliance on a range of self-report questionnaires, which have often not been validated specifically for the use with detainee populations, may be problematic though it is in our view unlikely that this significantly affects the overall validity of the findings. Moreover, measures often focus solely on symptoms of PTSD, depression and anxiety, which may fail to investigate other disorders such as schizophrenia and other non-affective disorders as may be the case in refugees as a whole [5]. Whilst self-report questionnaires are undoubtedly useful for the purpose of screening for mental health problems, outcomes of such self-report measures should not be equated with formal psychiatric diagnoses, which require clinician-administered interviews. To date, there is a limited use of diagnostic clinical assessments to complement self-report measures, and where these are conducted, the samples are restricted to adult male detainees only.

Taken as a whole, the body of research reviewed in this paper consistently illustrates severe mental health consequences amongst detainees across a wide range of settings and jurisdictions. In view of this, in terms of provision of care, consideration of mental health disorders and their sequelae should become routine in making the decision whether or not to detain, and in monitoring the welfare of those already in detention. Comprehensive sociocultural assessments should be carried out by care providers with transcultural knowledge, skills and practices [68]. These assessments should lead to adequate treatment for individuals in detention who suffer from mental health difficulties. Additionally, there is a need to identify and address vulnerability within an overall asylum policy and more specifically a detention policy that avoids or minimises further harm being done to vulnerable people. For this, it is important to take into account multiple varying expressions of distress and vulnerability, which should not be reduced to Western conceptualisations. Moreover, a broad concept of past trauma is needed which is not restricted to the definition of torture according to Article 1 of the Convention Against Torture [69], but rather focuses on the severity of harm to which the individual has been subjected [70]. Further, there should be a greater focus on minimizing length of detention. Our findings add weight to the call in the UK and other countries to introduce a statutory limit of duration of detention, such as exists in all other EU countries [71]. The increasing evidence for the negative impact of immigration detention on vulnerable populations is particularly policy-relevant. Similarly, there is insufficient attention paid to child protection and safety, or to adequate education and health protection in countries where children continue to be detained, despite the recognised vulnerability and risk of children and young people and their protection under the UN Convention on the Rights of the Child [72].