Abstract
Forensic mental health services play a key role in the diversion, treatment, rehabilitation, and supervision of offenders with mental health problems in Europe. Private sector providers are increasingly commissioned to provide secure care for service users. Questions have been raised about the effectiveness of the private sector providers. A sample of 229 patients from low, and medium secure psychiatric services in the United Kingdom was analyzed to determine differences in restrictive practices, adverse service user incidents, and service user characteristics between service users admitted to private sector or publicly funded hospitals. Service users with a diagnosis of a personality disorder, and those who had multiple psychiatric diagnoses were disproportionately placed in the private sector. Greater prevalence of seclusion, physical restraint, verbal aggression toward staff, physical violence toward staff and other service users, property damage, and self-harm were observed in private sector service users. Further attention is warranted around the decision-making processes that allocate people to private versus publicly funded care, potential sources of bias in admission characteristics should be taken into account when interpreting poorer clinical outcomes in the private sector.
Introduction
Secure services play an important role in the criminal justice systems of numerous jurisdictions across Europe and must balance the dual aim of treating the mental health needs of service users while also safeguarding the public (Gosek et al., 2020). Like many European countries, secure services in the United Kingdom (UK) are largely commissioned centrally but are often “contracted out” to private service providers 1 (O’Brien, 2015); large corporations (owned by venture capitalists) have increasingly provided secure mental health services for National Health Service (NHS) service users (Pollock, 2010). Little is known about which groups of service users are admitted by different types of service providers, or whether service users are treated equitably across the UK secure care system.
In the UK, secure services are commissioned for different levels of security. These include uncontrolled egress (no restrictions on service user access or egress), controlled egress (control over access and egress, but less than that within a secure unit) low secure units (LSUs), medium secure units (MSUs), and high secure units (HSUs). Definitions of the different levels of security have been proposed (Kennedy, 2002), but debate continues regarding how to match service users to specific levels of security. Gatekeeping decisions regarding where to admit cases continue to be prone to heuristic bias (Lawrence et al., 2018). Two problems arise from this uncertainty; first, people can be placed in unwarranted high levels of security relative to their risk, and second, people may be detained for inappropriately long periods (Davoren et al., 2015). This raises a number of ethical concerns relating to the rights of service users and for their prospect of recovery, particularly when considering the widely reported harmful effects of restrictive practices that are used widely in secure services (Lawrence et al., 2021; Sashidharan et al., 2019). Concern over the misplacement of service users extends beyond welfare and mental health recovery, to the significant drain that funding secure services imposes on the healthcare system.
Excessive security and prolonged care have financial implications, as an example, the total cost of secure services for Welsh nationals during 2019–2020 was £50.5 million (National Collaborative Commissioning Unit [NCCU], 2020) when the total population of Wales was only 3.1 million (Stats Wales, 2022). The mean cost of low and medium secure care in the UK is estimated to be over £325,000 per service user, per annum (Department of Health, 2015). It is, therefore, important that the decision-making processes regarding service user placement are transparent and consistently based on presenting risk and need for treatment in appropriately titrated conditions of security.
Tools such as the Dangerousness Understanding, Recovery and Urgency Manual (DUNDRUM; Kennedy et al., 2016) have been created to address assessment transparency regarding admission to and discharge from secure inpatient services. This toolkit has four structured professional judgment instruments, one of which (DUNDRUM-1) is used to triage service users to appropriate levels of therapeutic security and excellent predictive validity, high internal consistency, and good interrater reliability (Flynn et al., 2011). However, instruments like the DUNDRUM-1 are not used routinely to inform service user placement decision-making for secure service users and instead, clinical judgment determines largely where people are treated.
Current practice in the UK is that NHS-based consultant psychiatrists decide the level of security that an individual requires following their referral to secure services. This is referred to as the “gatekeeping” process and the psychiatrists making the decisions are most frequently referred to as the “gatekeepers” (Kennedy, 2022). While already a complex process, gatekeeping has been further complicated by increased demand for secure services. In the past 20 to 30 years, forensic bed provision has been increasing across Europe (Gosek et al., 2021). As the number of people occupying secure psychiatric beds has increased, the proportion of government-funded care provided by private services has also increased (Keown et al., 2018), and currently, the majority of Welsh secure service users are treated in the private sector (Mills & Davies, 2022). As well as deciding on the level of security required, gatekeeping clinicians are also required to make placement decisions regarding which service users to admit to publicly funded secure services and which to refer instead to private providers. Just like risk-based decision-making in this context, these decisions are also based to some extent on clinical judgment and may be prone to bias.
Clinicians base their decision to admit a service user to either a public or private sector service on two primary factors: bed availability and the need for specialist treatment. In the UK, priority is given to public sector services; however, if no beds are available, clinicians may opt for an equally suitable private sector alternative. Specialist treatment plays a crucial role, as many service users require tailored care; private sector services in the UK offer a broader range of specialized treatments (Zimbron et al., 2022) making them a viable option. Selection between public and private sectors hinges on the balance of cost and quality. The Quality Assurance Improvement Service (QAIS) of the NCCU evaluates these aspects using their three Qs framework (NCCU, 2022). Services with lower costs and higher quality receive higher priority for referrals from the public sector.
Increased use of private sector secure services has led to debate regarding the quality and necessity of such care. Criticism of private sector care has been raised by numerous bodies because service users are more often accommodated long distances away from their home area and families (NCCU, 2021) and because private services are costly (NCCU, 2022; Pollock & Godden, 2008). Some argue that public and private services should be comparable in quality because they are subject to the same inspection and monitoring processes (Murphy & Sugarman, 2010). For example, in the UK standards of care within secure services for both private and public sector services are set by the QAIS in Wales, and NHS England in England. Training levels of staff are also monitored by the public sector; for Wales, this comes under the remit of the QAIS and Healthcare Inspectorate Wales, and the Care Quality Commission in England. Despite the sample standards of care being applied within the public and private sector, differences in the quality of these services have nevertheless been reported in the media, with NHS services performing better on several indicators including service user access to staff (Mills & Davies, 2022) and the number of complaints made by service users (Health Inspectorate Wales, 2021). These considerations are problematic for all who are concerned about the quality of secure psychiatric care and are an additional complicating factor for gatekeepers who play a key role in deciding the level of security people are treated in and whether they are allocated to private or NHS services. Currently, little is known about the factors which distinguish between service users admitted to publicly funded forensic mental health services, and those admitted to privately owned services. Furthermore, it is unclear as to whether service users are treated equitably across both types of service.
A limited body of research has contrasted the characteristics, treatment and behavior of people admitted to public versus private secure services in the UK. De Taranto et al. (1998), Lelliott et al. (2001), and Kasmi et al. (2020) all report that service users are more likely to be involuntarily admitted to private secure services under civil sections of the UK Mental Health Act (1983/2007) than their NHS counterparts. Unlike in other countries, in the UK, as well as detaining individuals in secure mental health services under criminal legislation, the UK Mental Health Act allows for service users to be detained in secure services under civil legislation, if they are assessed by approved practitioners as presenting as a risk of harm to themselves or others that cannot be safely managed in the community or in less restrictive hospital settings (Hatfield & Antcliff, 2001). A bias for admitting female service users into the private sector has been reported (De Taranto et al., 1998) along with a similar bias for service users diagnosed with personality disorder or intellectual disability (Kasmi et al., 2020). Differences have also been reported in behaviors that challenge pre-admission; with private sector service users showing higher levels of aggression (Lelliott et al., 2001) along with increased use of restrictive practices by staff in the private sector (e.g. seclusion; Kasmi et al., 2020).
It should be noted that the findings in relation to restrictive practices are not consistent; Chaudhry and Pereira (2009) in a cross-sectional analysis of low secure services found higher levels of seclusion and high-dose medication in the NHS. Overall, the existing literature suggests that private services admit more women than the NHS, people diagnosed with personality disorder or intellectual disability, and people who have already proved difficult to accommodate in general psychiatric care. This may explain why there is mixed evidence surrounding more frequent use of restrictive practices within private sector services.
An objective of the current study was to build on the previous work that has compared service users in private and public sector secure services. To date, this research has been mainly confined to comparing service users in medium security and the findings have been inconsistent. Through using a national database of Welsh service users detained in low and medium-secure care throughout the UK we aimed to identify the factors that predicted admission to public versus private services. Not only did this allow for further comparisons of the clinical characteristics and outcomes for those who use the different types of service, it also allowed for the identification of factors that may be influencing the decision-making process of gatekeeping clinicians. The existing literature suggests that the private sector may provide a lower standard of care, but it also suggests that this may be confounded by bias in the allocation of service user groups that have often been characterized as “difficult to treat” (Bodner et al., 2011; Kohen, 2014).
We also aimed to explore the relationships between admission characteristics, service user challenging behavior and the use of restrictive practices in the public versus private sectors in order to address this hitherto confounded issue in the literature. To date, only limited research has considered bias in these gatekeeping decision-making processes (Lawrence et al., 2018). Understanding the allocation of secure service users to private and publicly funded care and describing any disparities in subsequent care is of interest to service commissioners, policy makers, to service users and the public alike, across all jurisdictions where forensic mental health service provision operates in such a manner. There are obvious potential benefits in trying to make such processes as fair, equitable and transparent as possible, but the first step in this process is identifying the source and nature of any disparities in current practice.
Method
Participants
The current study used data collected as part of standard practice evaluation by the Quality Assurance Improvement Service (QAIS) department of the National Collaborative Commissioning Unit (NCCU), a collaborative commissioning service of NHS Wales. The sample included Welsh service users treated in medium- or low-secure psychiatric services across Wales and England. Table 1 presents participant demographic information. Of the 229 service users, the majority were male, were treated in a low secure service, had a primary diagnosis of a psychotic disorder, had a history of Adverse Childhood Experiences (ACEs), were admitted into their current placement under a criminal section of the Mental Health Act (1983/2007) and had been in their current placement between one and two years. Service users were most commonly treated within a private low-secure service.
Participant demographics.
Ethics approval
Ethics approval was obtained from Cardiff Metropolitan University School of Sport and Health Sciences Ethics Panel (application number PGT-4347). The study was also approved as a service audit by Cardiff and Vale University Health Board, Cwm Taf Morgannwg University Health Board, Hywel Dda University Health Board, Powys Teaching Health Board and Swansea Bay University Health Board Research and Development departments. The audit took place in line with General Data Protection Regulation (GDPR) requirements for the use of confidential service user data without consent. The data were anonymized prior to all analysis.
Procedure
All data were collected by the QAIS audit team as part of their routine service review as commissioners of care provision for Welsh secure inservice users. Auditors collected the data by reviewing service user medical records and entries on the Commissioning Care Assurance and Performance System (CCAPS). This system is used by care providers to report information as required by their contractual agreement with the NCCU for quality assurance purposes. For example, staff are required to enter behaviors that challenge, or any restrictive interventions used with service users into CCAPS. Where multiple behaviors that challenge occur, the most severe of these are entered into the system. Where multiple restrictive interventions take place, all of these are recorded. In the event that a service user is placed in seclusion, the entirety of the time spent in seclusion is recorded as one incident. The system includes objective definitions of challenging behavior and restrictive interventions (see Table 2 for definitions), which are based on the NCCU collaborative framework and informed by Welsh Government (Welsh Government, 2022a). Medical records were either physical documents which were kept in a secure filing room on each treatment site, or, electronic and kept on a secure health board or company intranet server. Records were reviewed between November 2019 and November 2020. All data were completely anonymized prior to statistical analysis using SPSS.
Definitions of restrictive interventions and challenging behaviors.
Design
The study used a retrospective cohort design. There were several independent- and one dependent variable. The dependent variable—service provider—was binominal; service users were being treated in either “private” or “public” services. Independent variables were categorized as 1) frequency of restrictive interventions, 2) severity of service user incidents, and 3) service user characteristics. Scales for variables related to frequency of restrictive interventions and severity of service user incidents were developed based on Government auditing criteria (NCCU, 2022). Independent variables are outlined in greater detail below.
Frequency of restrictive interventions
Six restrictive interventions were defined in the current study (see Table 2). Restrictive interventions were coded as ordinal variables and included: 1) physical intervention, 2) restraint (not floor), 3) supine floor restraint, 4) prone floor restraint, 5) seclusion, and 6) rapid tranquilization. The frequency of experiencing each of these restrictive interventions was measured on a 0 to 4 ordinal scale where 0 = no history during the audit period, 1 = once to twice in the last 90 days, 2 = monthly, 3 = weekly, and 4 = daily. An “Overall Restraint Score” was also calculated by the sum of each restrictive intervention score with a minimum possible score of 0 and a maximum possible score of 24.
Severity of service user incidents
The following analysis was designed to determine if challenging service user behavior was more or less common in the public or private sector services. Twelve types of service user incidents were recorded as ordinal variables in the current study. The ordinal scale reflected a composite of incident frequency and level of harm (see Table 3). Level of harm was a measure of harm to the service or others, was based on the levels of harm framework (Welsh Government, 2022b) and was entered into CCAPS by the reporting member of staff. Incidents included: 1) verbal aggression toward service users, 2) verbal aggression toward staff, 3) physical violence toward service users, 4) physical violence toward staff, 5) sexually inappropriate behavior toward service users, 6) sexually inappropriate behavior toward staff, 7) purposeful property damage, 8) disruptive and/or intimidating behavior, 9) absconding, 10) illicit substance misuse, 11) self-harm, 12) being the victim of harm. The severity of service user incidents represented a composite score established by the sum of the incident frequency and level of harm. An “Overall Incident Severity Score” was calculated by the sum of each incident severity score, with a minimum possible score of 0 and a maximum possible score of 72. The creation of a composite incident severity score was completed on the basis that the objective of risk assessment is to consider both the likelihood and levels of harm, with assessors in psychiatry typically being biased by excess consideration of latter (Nilsson et al., 2009). By establishing a variable inclusive of both frequency and levels of harm we are going a small way to address this source of bias in our data.
Scoring of frequency and level of harm of patient incidents.
Service user characteristics
Several service user characteristics were included in the current study as categorical variables. For example, “service user sex” was binominal with levels male or female. To reflect whether service users had been admitted from a secure service of the same level of security as their current placement (i.e. from one medium secure service in to another medium secure service), “moved across the secure pathway” was a binominal variable with levels yes or no. “Mental Health Act section” was a nominal variable with levels none, civil, criminal, whereby participants were categorized based on whether they had been admitted to their current placement under civil sections of the Mental Health Act (1983/2007) (e.g. sections 2 and 3), under criminal sections (e.g. section 37, 47 and 37/41), or were not admitted under the Act. Three binominal variables pertaining to service users’ primary psychiatric diagnosis were established and included “psychotic disorder,” “affective disorder,” and “personality disorder” all with levels yes or no. To capture participants that had multiple diagnoses, this was included as another binominal variable with levels yes or no. Number of ACEs were reported by staff members in each hospital prior to audit and were extracted from medical records. Participants were also categorized as either having a history of ACEs with a binominal variable with levels yes or no.
Materials
Service user medical records were utilized in this study, the medical records included physical and digital copies of care and treatment plans, peer review reports, nursing notes and medical charts. Relevant data was stored in an encrypted Microsoft Excel document and analyzed using the Statistical Package for Social Sciences (SPSS version 29; IBM, 2022).
Method of analysis
Initially, a binominal logistic regression was planned to examine the variance explained in service type by frequency of restrictive interventions, service user incidents, and clinical characteristics of service providers. However, collinearity between predictor variables was detected through assessment of Variance Inflation Factors (VIF) scores during pre-analysis testing. Most VIF scores for our study variables were greater than 5, indicating moderate correlations (James et al., 2021). Therefore, alternative analysis methods were employed to explore associations between service user experiences, characteristics, and service provider type. Mann-Whitney U tests assessed differences in intervention frequency and incident severity between public and private services. Chi-squared tests explored associations between service user characteristics and sector allocation (public vs. private). Non-parametric analysis was chosen for its robustness and conservative nature to minimize Type I error (Gaito, 1959). The Holm Bonferroni method (Holm, 1979) was used to guard against alpha inflation.
Results
Differences in use of restrictive interventions between public and private services
A number of Mann-Whitney U tests determined whether the frequency of restrictive interventions differed between public and private secure services. These are detailed in Table 4 below. Public and private sector services differed significantly in terms of the frequency of seclusion, supine floor restraint, and physical intervention (p < .05); private services utilized these types of restrictive interventions more than public services. Service users treated in the private sector services also experienced significantly greater restrictive interventions overall.
Differences in frequency of use of restrictive interventions between NHS and private.
U: Mann–Whitney’s U statistics; z: standardized test statistic, *: statistically significant after adjusting for alpha inflation, p <.05.
The analysis thus far suggests that, at first glance, private services appear to be more restrictive than those in the public sector. The next set of analyses explored the possibility that this difference might be attributable to increased severity of behaviors that challenge being displayed by service users in the private sector, rather than private hospitals being more restrictive per se.
Differences in incidents between NHS and private services
Mann-Whitney U tests were conducted to determine whether the severity of untoward incidents differed between those treated in public versus private secure services. This is outlined in Table 5. Service users treated in the private sector services displayed significantly higher levels of self-harm, they also showed increased severity of purposeful property damage, verbal aggression toward staff, and physical violence toward staff and service users, compared to people treated in public services (p < .05). Service users treated in private sector services displayed significantly greater severity of incidents overall than those treated in public services, this finding suggests that their behavior was generally more challenging.
Differences in frequency/severity of patient incidents between NHS and private.
U: Mann-Whitney’s U statistics; z: standardized test statistic; *: statistically significant after adjusting for alpha inflation, p <.05.
Differences in service user characteristics between NHS and private services
Several Chi-squared analyses were conducted to explore the prevalence of service user characteristics between service users treated in public and private sector services. These are presented in Table 6. Significantly more service users with a diagnosis of a personality disorder, and service users with multiple psychiatric diagnoses were admitted into private sector services compared to public services (p < .05).
Prevalence of patient characteristics between NHS and private sector services.
X 2: Pearson Chi-sqaure; V: Cramer’s V; *: statistically significant after adjusting for alpha inflation, p <.05.
Discussion
The current research had two objectives that were addressed using routine audit data from NCCU’s Quality Assurance Department regarding all Welsh secure service users detained in low, or medium security in the UK whose care was funded by NCCU during the 2020 calendar year. First, we explored differences in the combined frequency and severity of service user incidents (e.g. verbal aggression, self-harm, and violence) between private and publicly funded services along with the use of a range of restrictive practices (e.g. restraint, physical intervention, and seclusion) in these two sectors. Both incidents and restrictive practices were more prevalent in the private sector. In the second set of analyses, we explored potential differences in the presenting characteristics of service users treated in the public versus the private sector as a potential confound of the preceding analysis. We found significant bias for allocating service users to private services who were diagnosed with personality disorder, and who had multiple psychiatric diagnoses.
At face value, the observation of increased prevalence of incidents and restrictive practices in the private sector is consistent with previous findings (Kasmi et al., 2020; Lelliott et al., 2001) that have questioned the quality of secure care provided by the private sector. However, by exploring possible confounding by service user presenting characteristics, we have called this interpretation of these differences into question. We found that gatekeepers were allocating subtly different groups of service users to these two sectors. Specifically, it is important to note that characteristics with increased frequency in the private sector group factors are known to be associated with treatment resistance, unstable recovery, high levels of self-harm, institutional aggression, and re-hospitalization (e.g. personality disorder; Sulzer, 2015). It is also important to note that the gatekeeping function in Wales is undertaken exclusively by NHS staff. A cynical interpretation of bias in this key decision-making process is that the NHS gatekeepers were diverting difficult-to-manage cases away from the services in which they worked and into private sector beds. An alternative cynical view is that private sector providers have been acting for their shareholders to optimize profit and will, therefore, have applied more relaxed criteria for accepting particularly complex cases.
An alternative explanation may be that private sector services are better equipped to manage patients exhibiting characteristics commonly associated with challenging behavior. For instance, provision of specialist personality disorder treatment is predominantly offered by the private sector (Zimbron et al., 2022). However, it is worth noting that, according to Zimbron et al. (2022), there is currently no nationally agreed service specification for “specialist personality disorder” services in the UK, with providers themselves defining this term. Additionally, a prerequisite for staff selection in secure services is the ability to effectively work with individuals demonstrating challenging behaviors, requiring specific training tailored to support service users with personality disorders (National Health Service England, 2018).
Initially, frequency of restrictive interventions was compared between service users in private and public services. Service users treated within private sector services experienced restrictive interventions with increased frequency, these included seclusion, supine floor restraint, physical intervention, and overall restraint, compared to those treated within the public sector. Our primary interpretation of this is that this observation stems from the higher levels of complexity and associated increased rates of challenging behavior among those referred to private services. Caution should be taken with this interpretation, however, as due to the methodological limitations of the current study we were not able to conduct an analysis to directly establish this link. Furthermore, it should be noted that private services have been criticized for providing less quality care than NHS services. Therefore, it is currently not possible to rule out that the increased levels of restrictiveness observed in private services reflects lower quality of care, for example, in relation to a lack of individualized care plans (Koekkoek et al., 2006). Service users treated in private sector services also displayed higher average levels of severity in incidents that included overall incident severity, self-harm, purposeful property damage, physical violence toward service users and staff, and verbal aggression toward staff.
Besides differences found in restrictive interventions and service user incidents, the results suggest that more service users with personality disorders and multiple psychiatric diagnoses were treated in private sector services compared to publicly funded services. No significant differences were found between service users with diagnoses of psychotic or affective disorders. While females were over-represented within the private sector, this difference was not significant. The greater population of women in private sector services may be the result of greater numbers of female beds available within private sector services, compared to the public sector (Bartlett et al., 2014) and may also explain why there are more service users with a personality disorder being cared for in these services, given that many females within secure care are diagnosed with personality disorder (Karsten et al., 2015). The greater number of female patients being treated in the private sector may also have contributed to the greater prevalence of restrictive interventions being employed within the private sector. There is literature to suggest that females are more likely to be secluded, and more likely to have experienced at least one restrictive practice compared to men (National Health Service Digital, 2015). However, it is important to acknowledge these findings are inconsistent and conflicting within the evidence base where, for example, Barr et al. (2023) found that young white men were at the highest risk of being secluded in their study.
Overall, differences in service user admission characteristics and behaviors that challenge provide convergent evidence that there may have been some bias in the allocation of more complex and challenging service users toward the private sector and away from public sector services. Due to small sample size and collinearity between variables used in the current study, we were unable to assess interactions between service user incidents, sector type, and use of restraint, while controlling for other factors such as patient sex. Future research should be conducted to address this.
Implications
The results of the current study raise a number of questions surrounding the method of allocating care for service users to either public or private secure services. Currently, neither service users nor their families are involved in the decision-making process for treatment placement, the responsibility for such decisions rests solely with the local Clinical Commissioning Group (CCG). The results of this study indicate that this decision-making process would benefit from further review. For example, the development of the DUNDRUM-1 tool has been applied to secure service gatekeeping regarding security level (Lawrence et al., 2018). Lawrence et al. (2018) highlighted the heuristics that may be at play within forensic practice, including representativeness, availability, and anchoring. Given that private sector services should be acting as an “overflow” measure for publicly funded services, minimal differences would be expected with regards to the challenging nature of service user’s clinical presentation between the two sectors. However, the current study has found significant differences in preexisting characteristics of service users, and their outcomes and experiences while in hospital. These results indicate that clinicians may be making decisions about service user’s placement based on such characteristics, either knowingly or unknowingly with the net effect of concentrating more difficult to manage cases in the private sector.
Limitations
While there is increasing public discourse about the roles of public and private psychiatric services, there has been a dearth of evidence-based research or evaluation on the topic. While previous literature has been limited, the current sample has included service users at two levels of security and of differing lengths of stay. However, despite this clear strength, there are a number of limitations to consider.
As data was provided by the NCCU and collected locally, uniformity may be lacking, as also noted in research by Kasmi and colleagues (2020). This is particularly true in a service evaluation in which variables were measured using scales that were not developed by the researcher. For example, scores on negative outcomes such as property damage and aggression were based on staff recording incidents as either “minor harm” or “major harm.” Operational definitions were provided for staff members for these harm classifications; however, it is possible that staff members may have interpreted these differently. Future research should look to include more objective measures of harm caused by challenging behavior.
Further limitations exist regarding the way in which data in the current study were collected, which may impact the validity of variables included. For example, CCAPS requires that services report the most severe behaviors that challenge, where multiple incidents occur, and so the number of incidents captured in the current study may not accurately reflect the total number of incidents that took place. This system also does not include more detailed information on the use of restrictive interventions, for example, the length of time that a service user stayed in seclusion. As such, seclusions that last for several days would be recorded the same as a seclusion that lasted several hours. To address this issue, future research should look to collect more detailed information on behaviors that challenge and restrictive interventions that are used within services.
Another limitation of our study is the lack of captured justifications for admission into private or public secure settings. This limits our ability to understand why service users were placed where they were and forces us to rely solely on specific service user characteristics. Understanding the rationale behind their admission could have enriched our interpretation of how the clinical stability of service users influenced their placement decisions. Investigating the distinct differences in reasoning for admitting service users to public versus private sector services should be a focal point for future research.
An additional limitation of our study is the inability to account for the specific type of personality disorder experienced by service users. For instance, certain personality disorders (e.g. cluster B personality disorders) are linked to heightened levels of challenging behaviors, such as aggression (Genovese et al., 2017). Consequently, we were unable to determine if a service user diagnosed with a personality disorder prone to challenging behavior influenced our results. Future research ought to incorporate more comprehensive data on the service user population, particularly focusing on diagnoses that might influence an individual’s likelihood to exhibit challenging behaviors.
Conclusion
This study has provided further evidence that there are significant differences between service users in public and private secure hospitals. Service users in private hospitals were more likely to have a personality disorder and had multiple psychiatric diagnoses. In terms of outcomes once in hospital, service users in the private sector were more likely to experience restrictive interventions and engage in more severe incidents, which could be a result of the greater clinical complexity observed in the service users most often placed in private services. However, on this basis, it is unclear why service users treated within the public sector were detained for longer. Both outcomes are concerning and suggest some impingement on service user rights. Future research should further explore the decision-making processes behind deciding service providers for service users. Considering the role of NHS staff as gatekeepers to all services, the observed disparities in service user presenting characteristics suggest that current practice is ethically questionable, with the NHS detaining service users displaying fewer challenging behaviors for longer periods of time and sending service users with more challenge behavioral histories to hospitals that may struggle to sufficiently meet their needs.
Footnotes
Acknowledgements
We would like to acknowledge the support of the NCCU and the service users whose data was used to conduct this research.
Declaration of interest
The authors have no conflicts of interest to report.
1
We acknowledge that private services could also include in the charitable sector; however, for the purposes of this manuscript, we have used private sector to refer to privately owned non-public sector services that provide mental healthcare.
