“Breaking down the barriers”: evaluating practitioners' training and practice with neurodivergent people in police custody Open Access

In the year ending 31 March 2023, more than 764,924 people were detained in police custody in England and Wales (Home Office, 2023). Police custody is a demanding socio-legal setting, in which detainees experience uncertainty and distress (Skinns and Wooff, 2021). Researchers such as Dehaghani (2021) propose that police custody therefore elicits vulnerability. Code C of the Police and Criminal Evidence Act (PACE; Home Office, 2024) defines vulnerability using narrower criteria – focusing upon characteristics such as a detained person's age, mental health and neurodevelopmental conditions (or neurodivergence). Resource scarcity – such as inadequate or minimal training – increases vulnerability for detainees and also custody practitioners (Dehaghani and Newman, 2017). Police often report a lack of neurodiversity-related training, and there is little evidence on how custody practitioners view their work with neurodivergent detainees. The present study (1) evaluates the impact of neurodiversity training on custody practitioners' knowledge and confidence working with neurodivergent detainees; and (2) describes factors that influence practitioners' work with neurodivergent detainees in police custody.

The estimated high prevalence of neurodivergence in the criminal justice system (CJS) has been an area of increasing focus in recent years (e.g. HM Inspectorate of Prisons and HM Inspectorate of Probation, 2021). Neurodivergence (or neurodiversity ^[1]) refers to natural variation in “ways of thinking and experiencing the world” (McLennan et al., 2025, p. 1) and encompasses neurodevelopmental conditions such as autism and attention deficit hyperactivity disorder (ADHD). It is estimated that between 1% and 2% of adults in England and Wales are autistic (O'Nions et al., 2023) and global estimates for adult ADHD range from 1.6% to 5% (Popit et al., 2024). Prevalence rates of neurodivergence are reportedly high in police custody. A recent study reported that of 216 detainees in London-based police custody units who consented to screening, 5% met criteria for further assessment of autism and 50% for ADHD (Brown et al., 2025). Another recent study screened 177 detainees in a police custody unit in Northumbria and reported higher than general population rates of neurodivergence, with 7% suspected to be autistic, and 14% identified as having ADHD (Cranshaw et al., 2025).

Neurodivergent (ND) people – including those who are autistic – experience a range of barriers across the CJS such as two-way communication and sensory-affective difficulties (see Woodhouse et al., 2024). These issues are also reported in police custody and custody-related processes (e.g. processing, detention, investigative interviews) (Bagnall et al., 2025; Holloway-George et al., 2025) with potential vulnerability factors proposed at individual, interpersonal and environmental levels (see Bagnall and Maras, 2025). Subsequently, many ND people in police custody – and the CJS more widely – should receive individualised support (Woodhouse et al., 2024).

Custody practitioners' ability to identify and respond to ND detainees' needs is essential. The majority of custody practitioners are custody detention officers, responsible for tasks such as detainee safety checks, biometric testing and liaising with investigators and solicitors (College of Policing, 2022). Custody officers (sergeant rank or above) hold overall responsibility for the operational running of police custody, including detainee wellbeing and risk screening (College of Policing, 2024a, 2024b). Risk screening processes can be insufficient to identify vulnerability in custody (Stoneman et al., 2019), but greater autism knowledge may aid police to more accurately identify autism and engage in more effective practice (Love et al., 2021). Yet, findings from the United States (Christiansen et al., 2023), England and Wales (Crane et al., 2016) and Australia (Love et al., 2022) show high variability in practitioners' neurodiversity knowledge. Only two studies have examined the neurodiversity knowledge of custody staff and their experience of working with ND detainees in custody. In Holloway et al. (2020), one custody sergeant and two detention officers described the importance of personal experience of autism and autism-related training for their knowledge and practice. Systemic (time pressures), environmental (the custody setting) and interpersonal (communication) factors were also perceived to impact their work with autistic detainees (Holloway et al., 2020). In a later study, Holloway et al. (2022) found that custody practitioners with lower levels of autism knowledge reported the highest perceived increase in knowledge following training (to which we return later). Taken together, neurodiversity knowledge is a key resource for practitioners working with ND detainees in police custody.

However, effective practice relies not only upon the acquisition of knowledge, but also practitioners' confidence to work with ND people (Love et al., 2022). Police officers with greater knowledge of autism report higher levels of confidence to work with autistic people (Love et al., 2021; also see Copenhaver et al., 2020). Police officers' confidence working with autistic people is related to the style (and quality) of training they receive, alongside officers' level of experience with autism, as well as personal characteristics such as officers' own demeanour (Love et al., 2022). More broadly, confidence (specifically, self-efficacy) is associated with greater task performance (Stajkovic and Luthans, 1998), and is protective against occupational stress and burnout (Shoji et al., 2016), indicating that confidence working with ND people is an important resource for police to acquire. To date, confidence working with ND people is yet to be examined in the police custody context within England and Wales.

The need for the police to receive autism training has been consistently emphasised in previous work (e.g. Holloway et al., 2020). However, several studies suggest that most police in England and Wales do not receive autism-specific training - despite high reported prevalence of ND custody detainees, and the provision of specialist training for all CJS staff being a core recommendation from the Criminal Justice Joint Inspection's 2021 Neurodiversity in the Criminal Justice System: a review of evidence report. Chown (2010) reported that in a sample of 120 police officers and police staff that none had received autism-specific training from their police force (30% had received police training that referenced autism or autism-specific training outside of the police). In Crane et al. (2016), 37% of 242 police officers had received autism-related training. Holloway et al. (2022) reported that only 32% of 142 custody practitioners (across five police forces in England and Wales) had received any autism training. While international studies indicate that autism training for police may be effective (Railey et al., 2020), only one study to date has examined this in England and Wales. Holloway et al. (2022) developed an autism training package specifically for custody staff, co-produced by autistic people, researchers and police. The training was perceived as helpful and relevant by custody staff, whose self-reported autism knowledge scores increased post-training and who described their intentions to use more autism-informed practice.

High rates of neurodivergence are reported among people detained in police custody, and the custody environment can present substantial challenges to these individuals. Police custody is also demanding for practitioners, and high risk for stress and burnout (Houdmont, 2013; Werner-de-Sondberg et al., 2018). Resource scarcity is likely to increase vulnerability for both detainees and practitioners (Dehaghani and Newman, 2017; Peacock and Hutchinson, 2025). While autism training can increase custody practitioners' perceived knowledge and intentions for autism-informed practice (Holloway et al., 2022), less is known about how neurodiversity training relates to practitioners' confidence working with ND detainees and the broader issues that impact these working relationships. The present study uses a mixed-methods approach, combining inferential statistical and qualitative analysis of survey responses from custody practitioners who attended neurodiversity training, to answer two exploratory research questions (RQs).

Four neurodiversity training sessions were delivered in 2022 to practitioners working in three police custody centres in the South West of England. These one-day sessions were provided by Creased Puddle (www.creasedpuddle.co.uk), a UK-based neurodiversity consultancy who specialise in neurodiversity training and delivered by a former custody officer. The training focused primarily on autism, though also addressed other aspects of neurodivergence, such as ADHD. An overview of the training content is provided in Appendix 1 ^[2]. Ethical approval was provided by the University of Bath Psychology Research Ethics Committee.

Ninety-nine participants completed paper-based surveys at the beginning of the training sessions ^[3]. An additional 19 participants completed an online follow-up survey six months after their training. Demographic information is summarised in Table 1. The survey did not collect participants' age or gender. The largest professional group represented in the sample were detention officers (55.4%) and the majority (64.6%) of participants had worked in a custody environment for five years or less.

Survey data were collected using measures administered at various stages of the training. A summary of the measures and the time points at which they were used is presented in Table 2. The knowledge and confidence measures reflect participants' self-perceived understanding and capability. Measures of actual interactions with ND detainees are based on participants' perceptions or estimates of disclosures during custody encounters (not independently confirmed by the research team).

Inferential statistics are reported for quantitative data. First, three independent samples t-tests were performed to assess whether participants with and without personal experience of neurodiversity differed in their neurodiversity knowledge, satisfaction with how they had worked with ND detainees and confidence working with ND detainees, respectively. Next, Spearman correlations examined relationships between neurodiversity knowledge, confidence and satisfaction working with ND detainees pre-training ^[4]. Two paired samples t-tests (pre-vs post-training scores) then assessed whether the training improved participants' self-reported neurodiversity knowledge and confidence working with ND people ^[5], respectively, while Wilcoxon signed-rank tests were performed to compare participants' level of agreement with the ten statements related to knowledge and practice with ND people pre- and post-training ^[6].

Only three six-month follow-up responses could be matched to the original pre- and post-training evaluation surveys. Consequently, six-month follow up data are reported in descriptive terms only and should be interpreted with caution. Not all respondents answered every survey item, so the responses across the reported data are not always equal to the total numbers in the sample.

Free-text survey responses from all 99 practitioners were analysed by the first and second authors using inductive thematic analysis (Braun and Clarke, 2022). Our analysis utilised a critical realist approach (see Fletcher, 2017) and aimed primarily to capture semantic (surface) level meaning. However, as the survey responses were often short, we took a pragmatic approach and interpretated more latent (underlying) meaning in the data. We also used a more interpretative approach for the third theme in the framing of “internal” and “external” resources.

Following the six-stages of inductive thematic analysis (Braun and Clarke, 2022), the survey responses were first read and re-read for familiarity. Both coders (the first and second authors) then applied an evolving set of codes to the data with ongoing discussion and review. The codes were then organised into clusters to generate initial candidate themes and sub-themes. The coding and theme generation stages were dynamic and involved re-coding, re-clustering and returning to revised themes and sub-themes. This process continued until both coders were satisfied that the themes reflected meaningful patterns in the data. The final set of themes and sub-themes were reviewed and discussed by all three authors.

We present the study findings in two sections. First, we report the quantitative results addressing the efficacy of the neurodiversity training (RQ 1). Second, we present the qualitative analysis which examines the issues that practitioners described as influencing their work with ND detainees (RQ 2).

Just over half (51.5%) of participants reported having had no personal experience of neurodiversity prior to the training, yet the majority (78.8%) reported encountering ND suspects on a daily or weekly basis in their professional roles. When asked about their overall level of satisfaction with these interactions, participants (n = 95) reported an average rating of 60.11% (SD = 18.42), indicating a moderate level of satisfaction (see Table 3).

Participants with personal experience of neurodiversity reported significantly higher pre-training knowledge of neurodiversity (M = 44.63, SD = 22.15) than participants without personal experience (M = 32.22, SD = 21.20), t(84) = 2.66, p = 0.009. The bootstrapped mean difference was 12.41 (SE = 4.54), BCa 95% CI [3.58, 21.39], Cohen's d = 0.57, 95% CI [0.14, 1.00].

Confidence working with ND detainees did not significantly differ between participants with personal experience of neurodiversity (M = 64.88, SD = 18.99) and those without (M = 59.56, SD = 21.32), t(84) = 1.22, p = 0.227. The bootstrapped mean difference was 5.32 (SE = 4.35), BCa 95% CI [−3.27, 14.17], Cohen's d = 0.26, 95% CI [−0.16, 0.69].

There was no significant difference in satisfaction working with ND detainees between participants who had personal experience of neurodiversity (M = 64.39, SD = 16.59) and those without (M = 57.78, SD = 19.29), t(84) = 1.70, p = 0.094. The bootstrapped mean difference was 6.61 (SE = 3.91), BCa 95% CI [−1.03, 14.22], Cohen's d = 0.37, 95% CI [−0.06, 0.79].

Finally, participants' pre-training neurodiversity knowledge positively correlated with their confidence working with ND people. Satisfaction in working with ND suspects also positively correlated with neurodiversity knowledge and confidence working with ND suspects (see Table 4).

Participants' self-rated neurodiversity knowledge (n = 93) was significantly higher after the training (M = 77.10, SD = 11.57) than before the training (M = 38.60, SD = 22.68), with a mean difference of 38.49, 95% CI [34.29, 42.70], t(92) = 18.17, p < 0.001, d = 1.88. Similarly, participants' confidence working with ND suspects (n = 90) was significantly higher post-training (M = 79.44, SD = 12.39) compared to pre-training (M = 61.89, SD = 20.55), with a mean difference of 17.55, 95% CI [13.78, 21.38], t(89) = 9.25, p < 0.001, d = 0.975.

Participants' pre-training level of agreement with all statements related to their knowledge of neurodiversity and practice with ND suspects increased significantly following the training (all ps < 0.001) with medium to large effect sizes (see Table 5).

A subset of participants provided feedback on their practice with ND suspects during the six months post-training. As stated, some participants could not be matched to the original pre-post responses, and so the data here are presented descriptively and should be interpreted with caution. Nineteen participants reported how frequently they had encountered ND suspects in this period; the majority (57.89%) reported contact of once a week, 15.79% reported daily contact, 15.79% reported monthly contact, 5.26% had encountered a ND suspect only once in total and 5.26% were unsure. Twelve participants rated their overall satisfaction in interactions with ND suspects since completing the training, and their average satisfaction (M = 77.83, SD = 11.51) was higher than the average reported by the wider group prior to training (M = 60.11, SD = 18.42). These participants also reported greater confidence in working with ND suspects six months post-training (M = 76.67, SD = 14.26), compared to pre-training (M = 61.89, SD = 20.55). Nearly all participants (92.86% of 14 respondents) reported that they had considered aspects of the training during their subsequent interactions with ND suspects.

Participants rated the training as helpful both immediately after the session (M = 83.83%, SD = 12.79) and at the six-month follow-up (M = 80.10%, SD = 14.24). The majority (63.5%) reported that the course did not need to be longer and most (91.7%) would recommend the training (see Table 6).

Thematic analysis of participants' free-text survey responses produced three overarching themes: (1) ND detainees have diverse needs; (2) Recognising and responding to ND detainees' needs is challenging; (3) Internal and external resources are required to work with ND detainees. Sub-themes are named using illustrative participant quotes (see Table 7).

Theme 1a: “Everyone is different”. A wide variability in the situations and needs presented by people detained in police custody was emphasised by participants. For example, a Detention Officer [16] ^[7] stated, “Every person is different and has different issues and problems”. Moreover, participants specified that there is no single or typical profile for ND detainees: “The spectrum is wider [sic] and no two people will present the same. It would be easy to assume that one size fits all but this isn't the case.” [Inspector 4]. As the previous quote suggests, this variability can present challenges for working effectively with ND detainees – as a Sergeant [31] reflected, “everyone is different. There will always be cases I will be at a loss to deal with regardless of my knowledge/training.”.

Theme 1b: “Feelings of vulnerability”. The diversity of ND detainees' needs was also reflected in their “feelings of vulnerability” [Detention Officer 2]. Police custody was sometimes specified as inducing vulnerability and distress in ND detainees. An Appropriate Adult [4] described the challenge of “The custody environment” and “The stress levels of everyone especially [the detained person]”. ND detainees can experience “anxiety” due to an “environment outside of normal routines” [Sergeant 21]. A Detention Officer [48] specified the impact of features of the custody environment, “over-stimulated with surrounding noise and amount of people, bright lighting”. The nature of ND detainee distress was also described: “emotions vary wildly without warning” [Inspector 2] and “erratic behaviour” [Detention Officer 25] “can escalate quickly” [Detention Officer 26]. This behaviour could turn “aggressive quickly if […] demands were not met” leading to acts of “self harm” [Detention Officer 34].

Theme 1c: “Under the influence”. Substance use and medication were sometimes highlighted as further diversity of needs of ND detainees. As a Detention Officer [32] put it, “Drink or drugs usually make my job hard, neurodiverse or not”. A Sergeant [18] suggested that “Alcohol. Drugs” can be challenging for staff, particularly when the detained person withholds this information from practitioners. A Detention Officer [7] similarly described the challenge of ND detainees being “under the influence” whilst in custody, and also when detainees do not have access to prescribed drugs and are “not correctly medicated”. The lines between intoxication and neurodivergent presentation can also be blurred. Some participants referred to their difficulty distinguishing between the “key differences between neurodiversity and drug induced behaviours” [Detention Officer 30]. As an Inspector [4] asked, “Drink/drugs – how would that impact?”.

Theme 2a: “Understand(ing) the signs”. As indicated in the previous sub-themes, custody practitioners described difficulty identifying and responding appropriately to the individual needs of ND detainees. For some a lack of knowledge of neurodiversity made this more difficult: “Because we have minimal knowledge it is hard to assess each person” [Detention Officer 17]. Participants hoped that the neurodiversity training would help them to understand the “Signs and effects of neurodevelopmental conditions” [Inspector 1] and be able to “identify a suspect with a neurodevelopmental condition and to help make their stay as stress free as possible” [Sergeant 13]. This was echoed in the difficult experiences of day-to-day practice with ND detainees: “To understand them. How they behave. How to deal with them.” [Detention Officer 13]. Some described how “Not knowing […] triggers” [Detention Officer 14] is challenging as this could lead to “a meltdown or problems with detention” [Inspector 4].

Theme 2b: “Breaking down the barriers”. Practitioners described the difficulties of engaging with ND detainees more broadly. The challenge of effective communication and “Ensuring understanding” [Sergeant 23] were often highlighted, as were the need to explain custody-related “processes” [Detention Officer 12] and “complex terms” [Sergeant 14] to ND detainees. This communication may be more difficult when the ND detainees' needs are not well understood. A Sergeant [7] further elaborated on the challenges of “Initial communication and breaking down the barriers to establish their needs”. Another Sergeant [31] described how, “Prior to fully understanding individual needs there can be a communication breakdown”. Indeed, the challenge of “Keeping them calm and engaged” [Detention Officer 28] was described, as was “Erratic behaviour, failure to listen when speaking to them” [Detention Officer 25]. Other responses reflected concerns about ND detained people's ability to “comprehend what's happening” [Detention Officer 31] and “really understand what is going on” [Sergeant 10] in custody.

Theme 3 encapsulates the factors that custody practitioners view as assisting in their work with ND detainees. We conceptualised these factors as resources which can be internal (knowledge, approaches and skills) and external (training, environmental and systemic).

Theme 3a: “Some knowledge and therefore some understanding”. Experience and knowledge of neurodiversity (internal resources) assisted the participants to work with ND detained persons. A Sergeant [24] referred to interactions that had worked well with ND detainees: “I have some knowledge and therefore some understanding”. Others' own personal neurodivergence helped them to understand and interact with ND detainees. For example, a Detention Officer's [18] own ADHD helped them to relate to ND detainees and described how they would “take my time and explain”. Another Detention Officer [54] described that they “use [their] coping methods with detainees” from their own personal experience of neurodivergence. More broadly, participants' personal and professional experiences of neurodivergence contributed to their confidence working with ND detainees, “I feel confident because of dealing with it in the family, but feel more knowledge needed” [Appropriate Adult 1].

Those who lack lived or professional experience may particularly benefit from external resources such as training. Participants described how they felt the neurodiversity training will help them identify and respond to the individual needs of ND detainees. For example, a Detention Officer [14] described how after the training they had a “Much more in-depth knowledge of neurodiverse conditions and what I can do to support individuals with a neurodiverse condition through the custody process.” Another participant described how the training “opened my eyes on how fast that kind of person get overload, confused, and now I know how to avoid that.” [Detention Officer 16].

Theme 3b: “More tools”. Participants described specific approaches, skills and strategies (internal resources) that help them to work with ND detainees. Effective communication was often raised as key for good practice, such as using “clear explanations” [Sergeant 14] in “simple terms” [Detention Officer 31]. Participants described responding flexibly to detainees through “Treating everyone as an individual” [Detention Officer 22], “Asking them directly” [Sergeant 30] how to assist them and adapting their practice “to make the process as easy and stress free” [Detention Officer 14] as possible. Others emphasised an empathic approach, as a Sergeant [18] reflected: “Being patient and listening can often reveal what's required”. A Detention Officer [13] described how interactions had gone well when they had tried “To understand their need, how they feel. Their behaviour, body language.” [Detention Officer 13]. The neurodiversity training (external resource) was widely positively viewed as providing new skills and strategies, such as how to make “adjustments” [Sergeant 27] and this helped participants to feel “More prepared” [Detention Officer 11].

However, some practitioners felt that they lacked (external) resources to support them in their work. The challenges of having “no resources” [Detention Officer 5] and adapting “to the suspect's needs with the limited resources available” [Detention Officer 34] were described. The efficacy of work can be influenced by whether they have the “time and space to build a relationship” [Appropriate Adult 4] and “Not having enough time […] prior to interviews” [Appropriate Adult 1]. A Detention Officer [54] raised similar concerns about the challenges of “Identifying what methods work for them in a very limited time. 24 h is not enough to help long term”. As described earlier, the custody environment itself is not “conducive to their needs … “and “ … the need to try to put measures in place.” An Appropriate Adult [3] simply described their challenges as “Environment. Resources”.

Most custody practitioners in the present study reported encountering ND people on a daily or weekly basis. While our sample is limited to South West England, this contributes to increasing evidence of high rates of neurodivergence within police custody and the CJS more broadly. On average, practitioners' pre-training ratings of neurodiversity knowledge were low, and over half of the sample lacked direct experience of neurodiversity in their personal lives. Furthermore, practitioners rated their neurodiversity knowledge as higher if they had personal experience of neurodiversity. The majority reported low to moderate levels of satisfaction in their pre-training work with ND detainees. The qualitative data provided clear illustrations of the demands of police custody (which we return to later).

Self-reported knowledge of neurodiversity and confidence working with ND people increased significantly (with large effect sizes) following the training. Significant increases (moderate-large effect sizes) were also seen in practitioners' post-training agreement with statements relating to neurodiversity-related knowledge and practice. Furthermore, we found a significant, positive relationship between participants' pre-training neurodiversity knowledge and confidence working with ND suspects (moderate effect size) and a strong, significant positive relationship (large effect size) between confidence and satisfaction working with ND suspects. This is consistent with previous research which found police officers' autism knowledge positively relates to their confidence (or self-efficacy) for working with autistic people (e.g. Holke et al., 2025). Indeed, police officers have described the importance of effective training and knowledge gained from personal experience for their confidence working with autistic people (Love et al., 2022).

These improvements should be viewed in light of our qualitative findings. In Theme One, participants perceived great variability in the needs and presentation of ND detainees, consistent with heterogeneity in autism (Jacob et al., 2019) and ADHD (Luo et al., 2019). The social, sensory and procedural nature of the custody environment was often viewed as eliciting or exacerbating vulnerability (as in Holloway et al., 2020; Dehaghani, 2021; Holloway-George et al., 2025). Further complications raised by intoxication and medication were described, such as distinguishing between drug use and neurodivergent presentation (see Gibbs et al., 2023), as well as challenges of engaging with ND suspects who are intoxicated or medicated. Some evidence indicates that substance misuse is lower in autistic than neurotypical people (Weir et al., 2021; but also see Haasbroek and Morojele, 2022), though it does still occur, for example, for self-medication (Ressel et al., 2020) and rates may be higher in police custody given the patterns in neurotypical populations (see Newbury-Birch et al., 2016).

In Theme Two, custody practitioners' difficulties identifying and responding to ND needs were discussed, including how this could lead to distress such as meltdowns. Meltdowns can occur when autistic people are overwhelmed by cognitive, sensory, emotional stresses (Lewis and Stevens, 2023) including in police custody (Slavny-Cross et al., 2022, 2023; Holloway-George et al., 2025). Such distress also exacerbated practitioners' difficulty communicating effectively with ND detainees. Custody practitioners therefore need to be better equipped to identify and respond to ND needs to prevent escalation and support ND detainees' participation in often high-stress custody contexts.

In Theme Three, we conceptualised the resources that custody practitioners need to work effectively with ND detainees as internal and external. Internal resources, such as neurodiversity knowledge, communication skills and confidence appear important for working with ND detainees. External resources such as training can help to support these, as shown in the present study and previous work (Love et al., 2022; Christiansen et al., 2023). Neurodiversity knowledge was significantly higher (pre-training) when participants had personal experience of neurodiversity (and neurodiversity knowledge was significantly related to confidence and satisfaction working with ND detainees) which highlights the importance of training, particularly for practitioners who lack personal experience. It should be noted that while confidence and satisfaction working with ND detainees was not significantly higher if participants had personal experience of neurodiversity, the difference was in the expected direction with a small-to-medium effect size; our study may have been underpowered to detect these smaller effects.

Our conceptualisation of internal (e.g. knowledge, skills, confidence) and external (e.g. training, environmental, systemic) resources for the demands of custody work with ND detainees was produced inductively through qualitative analysis of practitioner accounts. However, this has strong parallels with the Job Demands-Resources (JD-R) model (see Bakker et al., 2023), which we therefore use to further contextualise our findings. JD-R proposes that job performance predictors such as burnout and work engagement are related to work demands, which in turn are moderated through the availability of job resources (e.g. aspects of work that support goal achievement) and personal resources (e.g. individual characteristics that promote self-determination) (Bakker et al., 2023). For example, workers who are higher in confidence or self-efficacy (personal resource) are more robust to work demands and less likely to experience burnout (Ventura et al., 2015).

Stress and burnout are key psychosocial factors for police custody staff in England and Wales, related to organisational culture, workload, the physical work environment, and access to resources (Houdmont, 2013; Werner-de-Sondberg et al., 2018, 2021). Consistent with this, some participants in our study described limitations in external resources and how these constrained their practice with ND detained people. In particular, limited neurodiversity knowledge (an internal/personal resource), the fixed nature of the physical custody environment and a lack of time (external/work resources) to adequately assess and understand ND detainees were highlighted as barriers to effective work (as per Holloway et al., 2020).

National inspections similarly note that police often lack resources required for high-quality service delivery (HMICFRS, 2023), and broader occupational health psychology research demonstrates the importance of work resources for the psychological wellbeing of police staff (Oliver et al., 2023). Concerns are also increasingly raised about staff retention in the police service more broadly (see Stanislas, 2025). This pattern is particularly problematic in police custody contexts. As Dehaghani and Newman (2017, p. 1216) describe, vulnerabilities at institutional, practitioner and detainee levels “work in tandem”. Accordingly, resources such as knowledge and confidence (internal/personal), alongside training, staffing and time (external/work), support custody practitioners' resilience to vulnerability, which in turn supports the wellbeing of detained persons.

By synthesising our findings with previous work on the relationship between vulnerability and resources in custody (Dehaghani and Newman, 2017) and the principles of JD-R (Bakker et al., 2023), we propose a Resource-Demand Model for ND detainee engagement (see Figure 1). This model is intended as a conceptual framework to illustrate our discussion and guide future research, rather than a definitive causal theory. A set of professional practice recommendations are presented in Table 8.

Our sample comprised custody staff from the South West of England, and as such, our findings are not necessarily representative of England and Wales. For example, police funding is not equally distributed on the national level (Ogden et al., 2023), directed by factors such as crime rates, socio-economic status and population density (Johnston and Politowski, 2016). Regional variations will bring different custodial challenges and training needs that should be examined on the local and national level. Our quantitative and qualitative datasets also had certain limitations. Only a small proportion of participants completed the six-month follow-up survey, and within this sample the majority could not be linked to the original survey responses. While our six-month follow-up responses provide promising indication of training durability, longitudinal designs are needed to assess longer-term impacts (see Railey et al., 2020). We also do not know whether custody practitioners had received previous neurodiversity training, nor the true extent to which the current training has informed their knowledge and practice. Studies which could track training outcomes by the quality of specific interactions with ND detainees would take this further (Sreckovic et al., 2023). Our qualitative data were primarily short written responses in which participants did not often refer to specific ND conditions – instead ND detainees were described as a group. Furthermore, we cannot say for certain whether participants always knew whether the detainees to which they referred were ND, or whether participants sometimes inferred this based upon detainee presentation. Future research should instruct custody officers to specify the conditions with which detainees self-identify or are diagnosed when discussing their work with ND people.

Practitioners' interactions with ND people detained in police custody can be challenging, especially given the high-stress, time-limited environment in which they work. In this exploratory study, we highlight how internal resources (knowledge, skills, confidence) can be gained from external resources such as training delivery, and the importance of broader environmental and system factors. Applying a Resource-Demand framework may therefore be helpful for the identification and allocation of resources for working with ND people in police custody. Future research should empirically test the proposed Resource–Demand Model to examine the mechanisms underpinning these relationships. Regardless, it is vital that police custody contexts are allocated the resources needed to equip practitioners to work with vulnerable people at the “gateway” (Skinns et al., 2017) to the criminal justice system.

The anonymised quantitative data for this study are available on the Open Science Framework: https://osf.io/4auj6/. Qualitative data are not shared because participant consent and ethical approval did not include sharing of qualitative materials.

We are very grateful to the participants who took part in this research. Furthermore, we give special thanks to Creased Puddle and Dr Chloe Holloway-George for the training materials that underpin the project. Finally, we thank the Economic and Social Research Council (ES/W007789/1) for funding this research.

The supplementary material for this article can be found online.