Introduction

Many people in contact with the criminal justice system serve all or part of their sentence in the community [1,2]. Mental health and substance use (alcohol and drugs) problems are prevalent and frequently co-morbid in this population [3–7]. Prevalence estimates vary [3] but typically report over 40% meeting diagnostic criteria for a substance use disorder [5], more than 10% for a major depressive episode or psychotic disorder [5], and 75% for any personality disorder [6]. Mortality rates are higher compared to the general population [8], with a substantial number of avoidable deaths attributed to substance use or self-harm [9]. It is vital to enable people in contact with the criminal justice system to access adequate, timely support in the community to improve outcomes.

Help-seeking is a central component of general and mental healthcare access [10–12]. Models of healthcare access acknowledge individual, social, relational, and structural factors. However, there is little explanation on how these factors influence help-seeking specifically, nor how help-seeking may differ for highly stigmatised conditions, including mental health and substance use problems [13]. Help-seeking behaviour is shaped by individual attitudes, appraisal of social norms, and sense of control over behaviour [14]. These variables (attitudes, norms, ability) are associated with mental health help-seeking intentions [15] and have been used to inform interventions to increase mental health help-seeking [16,17]. Results of general population surveys on help-seeking for mental health and substance use problems demonstrate associations between help-seeking behaviours and gender; perceptions of stigma; and knowledge, attitudes and beliefs towards mental illness/substance use and treatment [18–22]. Whilst large general population surveys give important insights into associations between individual factors and help-seeking, they are limited in their ability to explore social/relational influences (e.g., shared beliefs between network members), surrounding cultures (community suspicion of state providers), or the specific experiences of people in contact with the criminal justice system.

In health research [e.g., 23,24] and other fields, [e.g., 25,26] researchers applying social network theory and analysis have shown that social networks in social contexts influence behaviours via various mechanisms. Selectivity describes how people form networks around themselves made up of people with shared characteristics, experiences, values, or behaviour that reinforce their own beliefs, attitudes and behavioural choices [27]. Alternatively, people may adopt the attitudinal and behavioural norms of the network already surrounding them, described as contagion [28]. Appreciating how networks influence behaviour is important for interventions that go beyond an individual, to consider the wider context which may facilitate or inhibit intervention effects. However, research to understand the relational influences on help-seeking for stigmatised conditions among marginalised populations is limited and the distinct experience of people involved in criminal justice system is minimally explored [29].

Men in prison describe a reluctance to seek help for mental distress because of their perceptions of stigma, fear of showing vulnerability that may place them at risk, a desire to project strength and masculine ideals, mistrust of services, and negatives beliefs about the outcomes of seeking help [30]. On release from prison, competing financial and accommodation needs can contribute to individuals deprioritising help-seeking for mental health and substance use [31]. Scrutiny can cause stress and disrupt help-seeking activities [32,33]. Further, people in contact with the criminal justice system in the community disproportionately live in social contexts affected by socioeconomic deprivation [34,35] where mental health and substance use problems are more likely to be present in their social networks, and services may be limited or oversubscribed [36,37].

There is very limited literature exploring help-seeking among people in contact with the criminal justice system beyond prison. Qualitative investigation is required to understand the individual, relational, and contextual influences on help-seeking among people in contact with the criminal justice system living in the community. This can inform judgements about the applicability of existing help-seeking/healthcare access models, and inform design of optimally effective individual and community level interventions. We aimed to understand the influences on help-seeking for mental health and substance use among people in contact with the criminal justice system in community settings, and to summarise these in an accessible framework that could be used to design or deliver interventions to increase help-seeking and access to care.

Methods

We report in-depth qualitative findings from a mixed-methods social network analysis study that explored how social networks influence help-seeking for mental health and substance use among people in contact with the criminal justice system in the community [38,39]. We collected qualitative data via interviews and applied framework analysis. Reporting is informed by COREQ [40].

Data analysis

We adopted framework analysis combining deductive and inductive approaches, assisted by Nvivo software [42]. Our initial framework synthesised common help-seeking elements from conceptual models of mental healthcare pathways [11,12,43] and general healthcare access [10] to form initial categories (themes): desire to seek help, ability to seek help, and help-seeking context. Researchers deliberately attended to relational processes and social context, in addition to individual factors. Three researchers (CC/JG/MT) independently coded three transcripts to the initial framework, inductively creating sub-themes within each main category (desire, ability, context). We (CC/JG/MT) met to discuss utility and consistency, which was high at the main category level. A key area of discussion was how to represent the highly inter-related relational and individual elements of help-seeking behaviour. We considered an additional category/theme for relational factors but ultimately decided to represent relational elements as a sub-theme within each of desire and ability, highlighting its integral presence rather than something that can be separated. After refining and adding inductively identified sub-themes into the framework, and discussing and agreeing the optimal structure, two researchers (JG/MT) independently coded two further transcripts. Consistency remained high, including at sub-theme level, and all transcripts were then independently double coded to the framework. A third researcher (CC) checked coding and created a matrix, with cases as rows and themes/sub-themes as columns. Coded segments were pulled into the relevant cells and summarised, before synthesising the main elements of each theme/sub-theme, and final review by all researchers. Overlapping sub-themes were then combined with input from all analysing researchers (CC/JG/MT) to produce the final framework (See Fig 1). We maintained a record of each iteration of the framework.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Framework of factors influencing help-seeking for mental health and substance use.

(A) Coding framework showing themes and subthemes, representing influences on help-seeking and where these interrelate.

https://doi.org/10.1371/journal.pone.0354996.g001

Findings

We interviewed 50 people: 23 in LA1 and 27 in LA2. An additional two wished to participate but did not have capacity to consent or did not meet inclusion criteria. No participants declined to participate or dropped out after agreement to meet a researcher. One participant opted to be interviewed with their support worker and one participant agreed for their interview to be used for analysis only (no verbatim quotation). Interviews lasted up to two hours, with one conducted over two meetings.

Participants were mostly white men living in areas of high socioeconomic deprivation. Most were serving a community sentence (e.g., unpaid work). Almost all self-identified as having current mental health or substance use needs, and the majority were accessing some support. Several were seeking help, and most had at least one unsuccessful help-seeking attempt. A substantial proportion of participants’ social networks had experienced problems with their mental health, substance use, or with the criminal justice system. Most network members were perceived to be supportive of help-seeking for mental health and substance use. See Table 1. Social network members were more equally split by gender and almost half had been known to the participant for over 10 years. They were slightly older on average and similar in ethnicity. Network members were described most frequently as friends, followed by family members and service providers. Full information of network members characteristics and the variation by gender is included in S1 Table in the S1 File.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Sample characteristics.

https://doi.org/10.1371/journal.pone.0354996.t001

Our framework has three categories (themes): desire to seek help, ability to seek help, and help-seeking context, each with further subthemes. ‘Desire to seek help’ is conceptualised as the yearning for something different (or not) and the motivation to achieve this (or not), whereas ‘ability to seek help’ describes the extent to which participants had the skills and resources available to them to achieve their desires. Desire and ability are further divided between individual and network influences, reflecting the integral importance of relational influences. ‘Help-seeking context’ described the wider community environment in which participants lived, including physical, social and cultural elements. See Fig 1.

Discussion

We present findings from the first in-depth qualitative study into the influences on help-seeking for mental health and substance use among people in contact with the criminal justice system in the community. Rigorous qualitative data analysis of 50 interviews undertaken within a mixed-methods social network analysis study, confirmed that desire to seek help, ability to seek help and help-seeking context are important influences on help-seeking in this population, consistent with theories of human behaviour [14] and early healthcare access models [10]. Our attention to social networks, in addition to individual and contextual factors, revealed the importance of relational influences operating across participants’ social networks and communities. This is a previously under-explored topic and thus, our findings make threesignificant contributions. First, we have advanced understandings of the factors influencing help-seeking for mental health and substance use among people in contact with the criminal justice system, particularly highlighting the need for nuanced appreciation of relational and contextual influences beyond the individual. Second, we make a methodological contribution to research examining behaviour and behaviour change with under-served/marginalised populations by highlighting the enhanced understanding attained by explicitly asking about peoples’ social networks in their distinct communities. Third, the framework provides a resource to inform the design and delivery of individual, social network, community-level, and public health interventions to optimise mental health and substance use-related help-seeking among people in contact with the criminal justice system living in the community. This could ultimately lead to more people securing timely access to appropriate services that reduce the risk of health crises or avoidable deaths.

The importance of individual knowledge, attitudes, beliefs, and perception of stigma are well-documented as influencing help-seeking for mental health and substance use [18–22], including among people in prison [45]. Demonstrating the relevance of these factors in a population involved with the criminal justice system in the community supports applying established cognitive-behavioural models to explain mental health and substance use-related help-seeking within this population [14,46]. However, our findings highlight the importance of areas that are only beginning to be considered in healthcare access models [10], and the potential limitations of considering help-seeking as a rational, individually-driven choice [43,47]. If we are to fully appreciate the factors influencing mental health and substance use-related help-seeking among people in contact with the criminal justice system, focused attention on relational influences operating across social networks and communities is required.

Social networks shape behaviour by providing (or withholding) resources, information and emotional support, and by subtly (or not) enforcing attitudinal and behavioural norms [48–51]. Perry and Pescosolido [43] demonstrated that average network-level trust in service providers was associated with mental health help-seeking. Our findings support this by qualitatively illustrating how the inverse, network mistrust of service providers can deter help-seeking. Epistemic mistrust is rooted in negative experiences in early relationships and is characterised as a self-protective position of dismissing others as unlikely to be safe information recipients/holders, to really care, or to be able to offer anything useful [52]. Adverse and abusive childhood experiences are associated with mistrust of others and thus avoidance of help-seeking among people sentenced to imprisonment [53]. We corroborate this in a large community sample and highlight how, not only services, but the knowledge they hold and their potential to be useful was mistrusted across peoples’ networks, further entrenching norms around rejecting help-seeking. Our findings indicate that epistemic mistrust, whilst often understood individually, can be shared across social networks. Thus, developing the trust needed to facilitate help-seeking for stigmatised conditions in under-served communities will likely require community-level cultural change.

A culture that characterises help-seeking as an indicator of weakness, beyond the stigma towards experiencing problems in the first place, has been shown to influence mental health help-seeking in other populations [54], but there are particular considerations when people are also involved in the criminal justice system. The need to project strength and autonomy to protect oneself from potential adversaries is articulated by men in prison [55]. In our community setting there was less emphasis on physical survival, and greater concern with community rejection, but both populations seek to preserve their ‘strong’ identities. Ramesh et al. [55] described a need to project strength as associated with the prison environment’s demands to perform masculinity. However, the presence of this in our community sample, and amongst women, suggests this may reflect common intergenerational norms in the networks of people in contact with the justice system. That is, norms that mental illness, substance use problems, and help-seeking are viewed as weakness.

Increasing public understanding of mental health and substance use problems and destigmatising help-seeking is an ongoing aim of public health campaigns in Scotland and elsewhere [56,57]. Despite participants’ networks being characterised by high levels of mental health and substance use problems, knowledge about mental health and substance use problems and how to address these was limited. This suggests that traditional strategies of information dissemination and anti-stigma messaging may fail to reach those in greatest need, potentially linked to the shared epistemic mistrust of information providers. Co-produced communication campaigns that capitalise on local culture and use trusted credible messengers and common communication formats, may be more likely to have success in increasing knowledge, problem recognition, and awareness of available support [58]. However, additional investigation of the mechanisms of influence and how best to increase knowledge are first needed.

Health care is free at the point-of-use and funded through general taxation in Scotland, so despite availability challenges, cost was not identified as a contextual influence as it may be in other countries. Nonetheless, access barriers remain. Scotland has seen substantial investment in support and treatment for illicit drug use in recent years [59]. However, mental health services remain oversubscribed, inadequately resourced, difficult to navigate, and often exclude people with substance use problems [60,61]. This remains problematic for people in contact with the criminal justice system where experiencing both mental health and substance use problems is common [62]. Given the individual and public health implications of unmet mental health and substance use needs among people in contact with the criminal justice system, it is vital that if people are encouraged to seek help, their contemporary experience is positive. Mental health and substance use services need to be adequately resourced to meet the level of need, and skilled to overcome the collectively remembered negative experiences in this population. Trauma-informed and relational practice approaches are likely to be an important component of more positive experiences [63,64]. Further, services need to be able to accommodate people who cannot meet inflexible demands and strict expectations. Services that take an inclusion health approach [65] may be beneficial, with consideration for remote, rural, and urban contexts, though evidence for the impact of inclusion health approaches is still emerging.

Intervention could be shaped by the framework we have developed. Service providers could utilise the framework to identify which influences may be evident in client’s life, and tailor intervention accordingly. However, the framework should be applied critically, and not be considered definitive at this stage. Future research could further refine and confirm the frameworks utility; identify which factors, in which combinations, are most modifiable and have the greatest impact for different groups in different contexts; and examine the mechanisms by which different factors have their effects on help-seeking. There is potential to use the framework to design and evaluate multi-level complex interventions to increase mental health and substance use-related help-seeking among people in contact with the criminal justice system, and to understand how interventions may need to vary across individuals, networks, and contexts. The framework can also be tested for utility in understanding the needs of individuals experiencing other stigmatised issues or from other marginalised groups, and designing intervention to encourage help-seeking.

Strengths and limitations

Our study has several strengths including a large sample size, recruitment from two distinct geographical areas that highlighted contextual influences, rigorous team analysis, and involvement of people with lived experience in the research team and LEAP. We published protocols a-priori [39]. Our framework was developed with a focus on mental health and substance use among people in contact with the criminal justice system. Given the common intersections between criminal justice system contact and other stigmatised experiences (e.g., homelessness) [66], our framework has potential to explain help-seeking for other stigmatised issues (e.g., HIV/AIDs, victimisation/abuse) in other marginalised populations. Further research would be beneficial to confirm this.

There are four limitations. Firstly, recruiting people who were not already receiving or seeking mental health or substance use support was challenging, due to our primary recruitment strategy being via service providers. However, many participants reflected on past experiences, providing insights into their experience when not seeking or accessing help. This was successful in counterbalancing any potential ‘pro’ help-seeking bias, and rich accounts are evident in the data. Secondly, we did not recruit in a city. Whilst the influences on help-seeking in cities may differ, our between-context comparison allowed us to identify context-specific factors that can be considered in any setting. Thirdly, distinguishing individual, social, and contextual influences on a complex human behaviour was challenging. Readers and those seeking to apply the framework in research and practice should be mindful of its preliminary nature and potential for overlap in the realities of peoples’ lives. Finally, caution will be needed in considering transferability of our findings to countries with different demographic characteristics, and legal, criminal justice, health, and social services. For example, in more ethnically diverse communities, intersecting stigma may have an additionally deterrent effect, or in countries where health insurance is required, cost/registration challenges would also need consideration.

Conclusion

Facilitating early mental health and substance use help-seeking among people in contact with the criminal justice system is an important public health concern, with potential to reduce later needs for emergency intervention and avoidable mortality. Our research demonstrates that the desire and ability to seek help among people in contact with the criminal justice systems are considerably influenced by relational factors, operating across social networks embedded in distinct local and national contexts. This highlights the advantages of taking a deliberately social and relational perspective in understanding help-seeking, particularly in relation to stigmatised conditions among underserved communities.

Whether working at individual, social network, community or a wider societal level, the framework produced can aid in identifying intervention targets for increasing help-seeking among people in contact with the criminal justice system. The complex relationship between individual, relational, and contextual influences indicates that multi-level interventions may be needed and that these should be further tailored to specific individuals, social networks, and localities.

Supporting information

Acknowledgments

We acknowledge the Lived Experience Advisory Panel, and Joshua Dumbrell for coordinating this, as well as our stakeholder advisors and service providers who supported interviews and recruitment.