The acute phases of the coronavirus disease 2019 (COVID-19) pandemic altered most aspects of daily life, impacting the physical and mental health of populations. Attempts to slow the spread of COVID-19 through lockdown measures caused disruptions to employment, education and health service use. Notably, a study of 10 UK mental health service sites reported reduced daily referrals, in-patient admissions and caseloads during the first UK lockdown period in spring 2020, compared with a pre-lockdown period; and this observation was also found elsewhere. Reference Bakolis, Stewart, Baldwin, Beenstock, Bibby and Broadbent1,Reference Chen, Jones, Underwood, Moore, Bullmore and Banerjee2 Similarly in the US, mental health-related emergency department visits declined from March 2020 relative to a pre-pandemic period, and some evidence suggests there was no significant change in psychiatric admissions. Reference Anderson, Radhakrishnan, Lane, Sheppard, DeVies and Azondekon3,Reference Kim, Rao, Collins, Eboh, Chugh and Sheladia4
In the UK, Black and other racially minoritised groups are more likely to come into mental health services through compulsory pathways, frequently involving contact with police and criminal justice, Reference Halvorsrud, Nazroo, Otis, Brown Hajdukova and Bhui5,Reference Devonport, Ward, Morrissey, Burt, Harris and Burt6 despite no consistent evidence of greater violence or substance misuse. Reference Bhui, Ullrich, Kallis and Coid7 This has also been observed in other countries. Reference Barnett, Mackay, Matthews, Gate, Greenwood and Ariyo8 During the early stages of the COVID-19 pandemic, police in the UK were given greater powers to enforce pandemic mitigation measures, such as social distancing. Reference Harris, Joseph-Salisbury, Williams and White9 It has become evident that there was significant racialised inequality in the way these measures were used. Reference Harris, Joseph-Salisbury, Williams and White9
Further, access to mental health care was not equally distributed during the pandemic, with ethnic inequalities reported in some countries. For example, in the US, compared to White Americans, Black, Hispanic, and Asian Americans were significantly less likely to receive mental healthcare services or telemedicine appointments during the pandemic. Reference Lee10,Reference Adepoju, Chae, Ojinnaka, Shetty and Angelocci11 Even prior to the pandemic, ethnic inequalities in mental health access and outcomes were known to be a major concern in the UK and internationally. Reference Bansal, Karlsen, Sashidharan, Cohen, Chew-Graham and Malpass12
Although pre-existing ethnic inequalities in compulsory psychiatric hospital detentions are well-documented, a review of the literature (see Supplementary Material 1 available at https://doi.org/10.1192/bjp.2025.10346) suggests no previous studies have yet investigated how COVID-19 lockdowns affected pre-existing ethnic inequalities in mental health-related admissions and detentions. Therefore, using a large population-level sample of people in contact with secondary mental health services in South East London, the aim of the current study was to investigate (a) how the first and second COVID-19 lockdowns, which occurred in 2020, affected overall mental health hospital admissions (both voluntary and compulsory combined, referred to as ‘admissions’ from this point) and compulsory psychiatric hospital detentions (referred to as ‘compulsory detentions’ from this point), and (b) whether this varied between racialised minority groups, compared to White British mental health service users. The present analysis forms part of a larger mixed methods study which investigated the impact of the COVID-19 pandemic on health and service use in racially minoritised people. Reference Impara, Bakolis, Bécares, Dasch, Dregan and Dyer13
Method
Participants and setting
Patient-level demographic and hospital admissions data in this study were extracted from mental health medical records from one of Europe’s largest secondary mental healthcare providers, the South London and Maudsley NHS Foundation Trust (SLaM), using its Clinical Record Interactive Search (CRIS) platform. Reference Perera, Broadbent, Callard, Chang, Downs and Dutta14 The construction of the dataset is detailed in Fig. 1. In the UK, healthcare is free at the point of contact, and, within this context, SLaM is the near-sole-provider of secondary mental health services for 1.3 million residents in a defined geographic area in South East London. Like other urban areas in the UK, the catchment area for this study is characterised by high ethnic and socioeconomic diversity (see Supplementary Materials 1, Table 1 for further details).
Fig. 1 Flow diagram showing the construction of the dataset for analysis. SLaM, South London and Maudsley NHS Foundation Trust; CRIS, Clinical Record Interactive Search.
Table 1 Descriptive characteristics (median and interquartile range (IQR)) of the sample, with average number of daily new admissions and detentions by ethnicity
Measures
Hospital admissions data were used to provide daily counts of admissions and all classifications of compulsory detentions under the Mental Health Act 1983 (MHA), subdivided by ethnicity. The outcomes of interest were any admission to in-patient psychiatric services over the observation period, either as voluntary patients or as compulsory detentions. The units for analysis used were: (a) the number of new admissions and/or detentions per day (incidence, i.e. the number of patients added to the patient load on any given day) and (b) the total number of in-patients detained and/or all in-patients within in-patient services per day (prevalence, i.e. the total patient load compulsorily detained on each given day).
Demographic data, specifically ethnicity, were linked to the full list of admissions to SLaM from 1 January 2016 to 8 October 2021 using pseudonymised unique patient identifiers. Ethnicity was initially mapped on to the 2011 Census ethnicity criteria and grouped as: White British, Irish, White Other, Black Caribbean, Black African, Indian, Pakistani, Bangladeshi, Chinese, Other Ethnicity, and Mixed Ethnicity. Due to small sample sizes, a decision was taken to aggregate groups, resulting in: White British, White Other (including Irish), Black Caribbean, Black African, South Asian (a combined category including Indian, Pakistani and Bangladeshi groups) and Other Ethnicity (including other mixed ethnicities, Chinese ethnicity and all other Asian ethnicity). Participants of mixed ethnicity were grouped with individuals with their minority ethnic identity; for example, participants who are mixed Black Caribbean and White British were included in the Black Caribbean category.
Time periods for analysis
England first introduced lockdown regulations on 23 March 2020, which involved closing all ‘non-essential’ businesses, recommending that people work from home if possible and only allowing people to leave their home for essential purposes (including seeking medical care). These regulations started to be lifted from 13 May 2020, with most national restrictions lifted on 4 July 2020 and replaced with regional regulations. A modified national lockdown policy was reintroduced on 5 November 2020 and, while this was replaced by regional regulations on 2 December 2020, the London regulations remained strict (enforcing limited contact between households, recommendations for people to work from home if possible and closure of ‘non-essential’ businesses). Shortly after this, on 6 January 2021, a third national lockdown was introduced, with a phased exit from lockdown starting on 8 March 2021 and ending on 18 July 2021. Reference Brown, Kirk-Wade, Baker and Barber15 These measures were supported by expansions to policing powers, as described above.
For the purpose of this study, we therefore chose to treat the UK lockdown measures as two distinct periods, as the regional restrictions placed on London between the second and third national lockdowns were of the highest level, and similar to those introduced during the national lockdown periods. A detailed timeline of the UK coronavirus mitigation measures can be found in the Institute for Government report. 16 For the purpose of this analysis, the first UK lockdown was defined from 23 March 2020 to 13 May 2020, and the second from 2 November 2020 to 8 March 2021. The cut points used for the lockdown periods in the analysis and control periods from 1 January 2016 to 8 October 2017, and 1 January 2018 to 8 October 2019, are illustrated in Fig. 2.
Fig. 2 Timelines of the study with cut points and comparison periods. Lockdown periods are sections of time that include regulations relating to the closure of businesses and limited household mixing. * The lockdown 2 period includes the lockdown period in November 2020, regional restrictions in London in December 2020 and the resumption of national lockdown measures from January 2021 to March 2021.
Statistical methods
First, we calculated the median and interquartile range (IQR) of the number of new and total admissions and detentions for each study period. Next, we used a fuzzy regression discontinuity in time with a difference in differences strategy to determine the effect of the four intervention cut points (i.e. the start and lift of the two described lockdown periods). Reference Bakolis, Stewart, Baldwin, Beenstock, Bibby and Broadbent1,Reference Hausman and Rapson17,Reference Hildersley, Easter, Bakolis, Carson and Howard18 We included two parallel pre-pandemic comparison periods in the models (1 January 2016 to 8 October 2017, and 1 January 2018 to 8 October 2019) to account for any changes associated with the intervention time points that were not due to the COVID-19 lockdown policies, shown in Fig. 2. We used Poisson regression models with robust standard errors, to assess the change associated with the cut points, accounting for trends associated with the month and day of the week. Where needed, we ran zero-inflated Poisson models to account for the high frequency of zero counts in the models analysing the daily rate of new admissions and compulsory detentions. Diagnostic tests, including checking for goodness of fit, autocorrelation, normality of the residuals and robustness of the linear predictions in the model to ensure assumptions, were not breached. First, we assessed the overall change in in the total and newly admitted or detained cases per day associated with the cut points, then we assessed the changes associated with the UK lockdown for the ethnic subgroups. Further details of the statistical methods, including assumptions, can be found in Supplementary Material 2. All analyses were conducted using Stata version 17 for Windows.
Post-hoc analysis
We conducted further post-hoc analyses to assess if observed changes in admissions were due to new admissions of people coming into secondary mental health services during the COVID-19 period for the first time. This post-hoc analysis assessed whether potential lower levels of early and/or preventative access to community care during the COVID-19 pandemic was associated with higher levels of in-patient admissions. To assess this, we re-ran the analyses restricted to people making contact for the first time during the pandemic (i.e. those flagged as first admissions only) in the clinical data.
Sensitivity analysis
We ran the following pre-planned sensitivity analyses, as per Bakolis et al: Reference Bakolis, Stewart, Baldwin, Beenstock, Bibby and Broadbent1
-
(a) To account for anticipatory effects of both lockdown periods potentially impacting people’s behaviours, and UK social distancing measures which were introduced earlier, on 16 March 2020, we re-ran the analysis with the start of lockdown dates set to one week before the actual introduction of the policies (i.e. the start of lockdown 1 on 16 March 2020, and the start of lockdown 2 on 26 October 2020).
-
(b) To assess the sensitivity of changes associated with the start of the lockdown lifts, we re-ran the analysis with the lockdown lift indicators coded as the time periods that had the start of the least restrictive lockdown measures (i.e. the lift of lockdown 1 set as 4 July 2020, and the lift of lockdown 2 set as 18 July 2021).
-
(c) To check that the findings were not an artefact of the pre-lockdown time periods, we altered the time window for the lockdown periods to include the period from 30 January to 23 March 2020, and 28 June 2020 to 2 November 2020 (i.e. the same number of days as the lockdown period but different specific dates).
Patient and public involvement
Members of our core team had lived experience of the issues under investigation; J. Dyer has lived experience of caring for siblings with severe mental health issues. The study was part of the COVID-19 Ethnic Inequalities in Mental health and Multimorbidities (COVE-IMM) project which involved an ethnically diverse steering group that included university researchers and people with lived experience of severe mental health problems. Reference Impara, Bakolis, Bécares, Dasch, Dregan and Dyer13
Results
Daily distributions (medians and IQRs) for the number of new admissions and compulsory detentions for the full sample, grouped by ethnicity, are shown in Table 1. The distributions for total number of in-patients and compulsory detentions are displayed in Supplementary Table 3.1. The median number of new admissions per day was 10, with 6 new compulsory detentions per day. The median number of total in-patients was 697 per day, with 503 compulsorily detained (Supplementary Table 3.1). Overall, total psychiatric admissions and compulsory detentions dropped during the first lockdown period and remained lower than the pre-lockdown period throughout the study period. In contrast, the daily rate of new admissions detained increased over lockdowns compared to the pre-lockdown period and remained higher throughout the study period.
During the first lockdown (23 March 2020 to 12 May 2020), the total number of in-patients admitted and compulsory detentions fell compared to pre-lockdown trends, as illustrated in Fig. 3(a). This trend was consistent for all racially minoritised groups, as illustrated in Fig. 4(a). As displayed in Table 2, in adjusted models, the overall number of new admissions reduced slightly (incidence rate ratio (IRR) 0.87 (95% CI: 0.75–1.00)) during the first lockdown period and returned to the same rate following the lift of the first lockdown. However, the number of new compulsory detentions rose during the first lockdown (IRR 1.25 (1.02–1.54)). This trend was not noted across all racially minoritised groups, and, in fact, was mostly driven by the Black Caribbean group, who experienced a significant rise in daily new detentions compared to the pre-lockdown period (IRR 1.54 (1.08–2.20)) despite not showing a corresponding increase in total new admissions (i.e. comprising both voluntary admissions and compulsorily detentions, IRR 1.17 (0.86–1.58)). Further, there was a consistent decrease in the total number of Black Caribbean patients who were in-patients (IRR 0.81 (0.80-0.28), Supplementary Table 3.2) and total number of Black Caribbean in-patients who were detained (IRR 0.78 (0.77–0.80), suggesting that Black Caribbean patients were being detained more frequently, however were not being kept as in-patients for as long as they had previously.
Fig. 3 Line graph showing the unadjusted changes in the total number of in-patients and number of in-patients detained within mental health services during (a) the first lockdown period from 23 March 2020 to 13 May 2020 and (b) the second lockdown period from 2 November 2020 to 8 March 2021. Dashed lines show the trends during the 2016/2017 and 2018/2019 control periods.
Fig. 4 Line graph showing the unadjusted changes in the total number of in-patients and number of in-patients detained, by ethnicity, within South London and Maudsley NHS Foundation Trust services. Dashed lines show the trends during the 2016–2017 and 2018–2019 control periods. (a) First lockdown period from 23 March 2020 to 13 May 2020. (b) Second lockdown period from 2 November 2020 to 8 March 2021.
Table 2 Incidence rate ratios (IRR) indicating changes in daily admissions (informal admissions and compulsory detentions) and compulsory detentions associated with lockdown compared to the pre-lockdown periods, stratified by ethnicity and adjusted for seasonal and weekly trends. Results obtained using zero-inflated Poisson regression models accounting for lower admission rates on different days of the week and robust standard errors
There were fewer changes associated with the second lockdown (2 November 2020 to 8 March 2021), as shown in Fig. 3(b) and Fig. 4(b). After adjusting for seasonal and weekly trends, accounting for the control period, there was a slight drop in the total in-patients (IRR 0.87 (0.86–0.27), Supplementary Table 3.2) and compulsory detentions (IRR 0.92 (0.91–0.93), Supplementary Table 3.2) during lockdown 2, and this continued past the lift of the lockdown on 8 March 2021. However, this trend was not universally observed in all racially minoritised groups: the total number of in-patients and compulsory detentions rose for both the White Other and Black African groups in lockdown 2 (Supplementary Table 3.2). Table 2 shows no overall change in the total number of daily new admissions associated with the second lockdown period; however, there was an increase in the Black Caribbean group (IRR 1.31 (1.02–1.67). Further, there was an increase in the number of new daily compulsory detentions in the total sample associated with the second lockdown (IRR 1.28 (1.11–1.49), Table 2). This appeared to be mainly driven by an increase in new compulsory detentions in the Black Caribbean group (IRR 1.53 (1.14–2.06)) and the Black African group (IRR 1.57 (1.06–2.34)).
In post-hoc analyses, we restricted the sample to those who were admitted for the first time to psychiatric units during the pandemic. The restriction was undertaken to assess the possibility that potential changes to community mental health service provision at the time of the pandemic (i.e. reduced access and/or transition to remote provision over face-to-face contacts) had impacted associations. Despite the sample restriction, which resulted in a smaller overall sample size (initial sample n = 15 869 reduced to n = 11 033 with restrictions), we found a higher effect size for detentions at all time points in the total sample, but with no changes to estimates for daily new admissions overall. Additionally, rates of compulsory admissions were higher in the Black Caribbean group during the second lockdown (IRR 2.07 (1.11–3.84) in the sample restricted to first contacts versus IRR 1.53 (1.14–2.06) in the unrestricted sample) (Supplementary Table 5).
In sensitivity analyses we observed evidence of an anticipatory effect in the week leading up to the first UK lockdown, with similar effect sizes to those observed in the main analyses but not as large (see Supplementary Materials 4 Table 4.1). Our findings were robust to further sensitivity analyses which assessed the sensitivity of changes associated with the start of the lockdown lifts (see Supplementary Table 4.) and sensitivity analyses to check that the findings were not an artefact of pre-lockdown time periods (see Supplementary Table 4.2).
Discussion
There have been longstanding concerns that individuals from ethnic minority groups in the UK and across international settings experience inequalities in higher levels of compulsory psychiatric detentions; Reference Barnett, Mackay, Matthews, Gate, Greenwood and Ariyo8 however, it was unknown how lockdown measures implemented during the COVID-19 pandemic affected this. In line with mental health service trends reported during the early acute stages of the pandemic, Reference Bakolis, Stewart, Baldwin, Beenstock, Bibby and Broadbent1,Reference Anderson, Radhakrishnan, Lane, Sheppard, DeVies and Azondekon3 we found that admissions to mental health units dropped significantly during lockdown periods for mental health service users within our sample, irrespective of ethnicity. In contrast, there were more detentions across the study sample during lockdown periods, which were driven by higher rates of detentions in Black Caribbean and Black African people relative to pre-lockdown reference periods.
This study is the first to highlight how COVID-19 lockdown measures exacerbated pre-existing ethnic inequalities in compulsory psychiatric detentions. The findings may be attributed to pre-existing social, economic and health inequalities which were amplified in the pandemic context, alongside concerns that structurally mediated racism was potentially magnified during the pandemic. Reference Harris, Joseph-Salisbury, Williams and White9
People from some ethnic minority backgrounds in Western countries experience disadvantage relating to social and economic determinants of mental health, Reference Mude, Oguoma, Nyanhanda, Mwanri and Njue19 and policies and laws enacted during the pandemic reinforced and widened these inequalities, having a disproportionate impact on the health and well-being of racialised groups. Reference Harris, Joseph-Salisbury, Williams and White9,Reference Perry, Aronson and Pescosolido20 For example, research in the US demonstrated that individuals from ethnic minority groups were more likely to be laid-off or lose employment income during the pandemic Reference Perry, Aronson and Pescosolido20,Reference Park21 and experience housing instability Reference Chun, Roll, Miller, Lee, Larimore and Grinstein-Weiss22 and food insecurity due to financial difficulties. Reference Perry, Aronson and Pescosolido20,Reference Morales, Morales and Beltran23 Individuals from ethnic minority groups were also more likely to reside in crowded housing, rely on public transportation and work in low-wage essential jobs that lacked paid sick leave and the ability to work remotely, and increased regular contact with strangers, directly and indirectly increasing the risk of COVID-19 infection, as well as possible stress and anxiety associated with this. Reference Khazanchi, Evans and Marcelin24 Similar inequalities have been reported in the UK. Reference Finney, Nazroo, Bécares, Kapadia and Shlomo25,Reference Ellingworth, Bécares, Šťastná, Nazroo, Finney, Nazroo, Bécares, Kapadia and Shlomo26 During lockdown periods, restrictions made it difficult for families to cope with COVID-19-related losses, as they could not be involved in the normal processes of grieving (e.g. saying goodbye or holding a funeral) and access to usual social support systems were reduced. Reference Torrens-Burton, Goss, Sutton, Barawi, Longo and Seddon27,Reference Ocloo, Stuart, Dasch, Dyer, Choudhury and McAnuff28
Outside of the pandemic context, studies from both the UK and US have argued that ethnic minority groups often fear seeking treatment for early symptoms of mental illness due to experiences of racism and discrimination, stigma associated with religious and/or spiritual practices, mistrust of services, lack of knowledge, fears of confidentiality and language and cultural barriers. Reference Devonport, Ward, Morrissey, Burt, Harris and Burt6,Reference Bansal, Karlsen, Sashidharan, Cohen, Chew-Graham and Malpass12,Reference Bruce and Smith29–Reference Winsper, Bhattacharya, Bhui, Currie, Edge and Ellard31 This underrepresentation in primary mental healthcare can lead to more chronic episodes of mental illness, greater functional limitations and overrepresentation in crisis pathways. During the pandemic, there were challenges in reaching some ethnic minority groups in the UK following closure of community organisations and the transition to remote service delivery for usual check-ups and providing resources. Reference Ocloo, Stuart, Dasch, Dyer, Choudhury and McAnuff28 Community service closures during lockdowns may have made people who were already isolated more vulnerable. Reference Ocloo, Stuart, Dasch, Dyer, Choudhury and McAnuff28 The provision of remote healthcare would have further disadvantaged those due to the digital divide (the gap between those who have access to digital technologies and those who do not). Reference Winsper, Bhattacharya, Bhui, Currie, Edge and Ellard31 A previous qualitative study of racially minoritised service users from the same catchment of this study specifically highlighted the adverse impacts of the abrupt withdrawal and/or rapid transition to telephone and video contacts for the delivery of mental healthcare, hospital appointments, social service and community and voluntary organisation contacts on the mental health of people living with severe mental health conditions and on their carers. Reference Ocloo, Stuart, Dasch, Dyer, Choudhury and McAnuff28 Participants in this study reported how the withdrawal of support adversely impacted their health and ability to access care, with resultant detrimental impacts on pathways into care, which were then delayed or disrupted. Reference Ocloo, Stuart, Dasch, Dyer, Choudhury and McAnuff28 The findings in the study highlighted at least one account of police intervention, which was attributed to disruptions in access to home treatment teams during the pandemic. Reference Ocloo, Stuart, Dasch, Dyer, Choudhury and McAnuff28 In addition, our post-hoc analyses suggested that when the sample was restricted to first admissions (i.e. excluding those previously known to mental health services), rates of compulsory detentions were even higher for the total sample. In addition, rates of compulsory detentions were higher for Black Caribbean service users during the second lockdown, in the sample restricted to first contacts. This could imply that changes to UK community health service provision during the pandemic were potentially implicated in higher compulsory admissions as early and/or timely access may have been disrupted; however, this latter observation will need further exploration.
Some adults from Black ethnic backgrounds in England have described mental health services as disempowering Reference Lawrence, McCombie, Nikolakopoulos and Morgan32 and detention as a racist and racialised experience, inseparable from a wider context of systemic racism, inequality and stigma within families and communities. Reference Solanki, Wood and McPherson33 The numbers of people compulsorily detained for treatment in the UK has doubled since 1983, when the most recent legislation governing compulsory admissions to in-patient mental healthcare was introduced, despite the prevalence of psychosis remaining broadly stable over the same time period. Reference Hawkes34 Our observation that there may have been similar or even elevated rates of compulsory admission in people who had not previously been admitted to mental health services during the pandemic further highlights the potential increase in use of compulsory admissions. In recognition of this increase in detentions, and the persisting ethnic inequalities, there are current moves to reform UK legislation regarding compulsory detention of people suffering from mental health disorders. Reference Dyer35,36 The review included the development of the patient and carers race equality framework (PCREF) that aims to improve mental health services for people who belong to racialised minority groups. 37 While these developments are important in addressing the overrepresentation of ethnic minority individuals in mental health crisis pathways, the present study has shown that we need to ensure that systems are also responsive and effective during times of crisis, in which inequalities across a range of wider determinants of mental health are synergistically exacerbated for individuals from marginalised and ethnic minority groups. Reference Khazanchi, Evans and Marcelin24 In addition, our findings suggest a need to directly tackle systemic racism in criminal justice and healthcare systems. The methods to address this at systems level may vary across international contexts. An example of an approach includes the roll-out across England of the PCREF, following the 2018 Mental Health Act Review which highlighted high rates of coercive admissions in racialised minority groups. Reference Ocloo, Stuart, Dasch, Dyer, Choudhury and McAnuff28 PCREF in England aims to embed anti-racism within mental health service provision through working in partnership with racialised communities, leading to co-produced services; through interventions to tackle racism and implicit biases in the workforce; and through improving data collection, leading to better monitoring and delivery of equitable mental healthcare for racialised communities. Although this is an important initiative which will potentially improve secondary mental health care in England, previous research has indicated that experiences of racism and racial discrimination are pervasive and persistent across domains and throughout the life course. Reference Bécares, Taylor, Nazroo, Kapadia, Finney and Begum38 As such, anti-racism initiatives that target single systems are not sufficient to address structural racism; what is needed are multisector, equity-oriented initiatives and policy reform. Reference Bailey, Krieger, Agenor, Graves, Linos and Bassett39
This study’s strength lies in the utilisation of routine data from a large healthcare provider that covers an ethnically diverse population within South East London. In the UK, most care is provided freely under the National Health Service (NHS), with the mental health trust in the present study being the sole NHS secondary mental healthcare provider in this area. Therefore, we can be confident that the analyses would have captured most mental health hospital admissions within the catchment, although a few private providers do also operate in the area of the study. Additionally, the study was conducted over a 5-year period spanning a pre-pandemic period which enabled robust comparisons with a prior reference period, and we adjusted for seasonal and weekly trends across models.
Some limitations should be considered. First, the study reports the effect of UK lockdowns on one healthcare provider in one region of the UK. The trends observed in this study may differ from other regions in the UK; however, they are likely to generalise to other similar urban and ethnically diverse catchment areas elsewhere. There were also limitations in sample size, particularly for some minority ethnic groups, which may have led to some analyses being underpowered. We had to combine ethnic groups due to smaller sample sizes (e.g. the ‘South Asian’ group included Indian, Pakistani and Bangladeshi people), which would have masked heterogeneity.
Further, due to the study design, we assumed other characteristics, such as area deprivation, remained constant between the different comparison periods and cut points, and we were unable to account for the potential confounding effect of these and other factors within our analysis. Our study only followed the sample during the initial lockdown phases of the pandemic, and so it should not be extrapolated out of this context.
In conclusion, there is a need to address ethnic inequalities in compulsory psychiatric detentions and attend to the exacerbation of pre-existing social and economic inequalities during health emergencies like the COVID-19 pandemic, in the UK and internationally. This cannot be achieved without first ensuring primary mental health care is culturally safe and appropriate, reaching ethnic minority groups early. In England, Mental Health Act reform will play a role in potentially preventing these evident ethnic inequalities within in-patient mental health services. Importantly, there is a need to address systemic racism within criminal justice and healthcare systems to reduce the overrepresentation of ethnic minority individuals in crisis pathways.
Supplementary material
The supplementary material is available online at https://doi.org/10.1192/bjp.2025.10346
Data availability
The data are not publicly available because they are owned by a third-party South London and Maudsley National Health Service (NHS) Foundation Trust Biomedical Research Centre Clinical Record Interactive Search (SLaM BRC CRIS) tool which provides access to anonymised data derived from SLaM electronic medical records. The data can only be accessed by permitted individuals from within a secure firewall (i.e. remote access is not possible and the data cannot be sent elsewhere) in the same manner as the authors, following relevant project approvals. Analytic code may be available by contacting the lead author (R.H.) or corresponding author (J.D.-M.).
Author contributions
J.D.-M., I.B. and R.H. conceived and designed the study. J.D.-M. led the acquisition of the data for analysis. Analysis of the data was led by R.H. and supported by J.D.-M. and I.B. All authors (R.H., T.K.O., I.B., L.B., A.D., J.D., M.H., J.O., R.S. and J.D.-M.) contributed to the interpretation of data analyses. Drafting of the work was led by R.H. and T.K.O. and supported by J.D.-M. All authors reviewed the work critically for important intellectual content, approved the final version for it to be published and agree to be accountable for the work done.
Funding
This paper represents independent research funded by The Health Foundation’s COVID-19 research priorities programme (award reference: 2238180), which part supported J.O. and R.H. The Health Foundation is an independent charity committed to bringing about better health and healthcare for people in the UK. J.D.-M. is part supported by the ESRC Centre for Society and Mental Health at King’s College London (KCL) (ESRC reference: ES/S012567/1). J.O. is part supported by the National Institute for Health and Care Research (NIHR) Applied Research Collaboration South London (NIHR ARC South London) at King’s College Hospital NHS Foundation Trust. M.H., R.S., J.D.-M. and A.D. are part supported by the NIHR BRC at SLaM. L.B. is supported by the Nuffield Foundation (WEL/43881), the ESRC (ES/V013475/1; ES/W000849/1) and the Health Foundation (AIMS 1874695). R.H. is currently funded by a doctoral studentship granted by the ESRC managed by King’s College London (ESRC KCL LISS-DTP). J.D. has additionally received funding from the Health Foundation working together with the Academy of Medical Sciences, for a Clinician Scientist Fellowship. J.D. and T.K.O. are additionally funded by UK Research and Innovation funding for the Population Mental Health Consortium (grant no. MR/Y030788/1), which is a part of Population Health Improvement UK (PHI-UK), a national research network which works to transform health and reduce inequalities through change at the population level. R.S. is additionally part-funded by the NIHR ARC South London at King’s College Hospital NHS Foundation Trust, the UK Research and Innovation (UKRI) Medical Research Council through the DATAMIND Health Data Research UK Mental Health Data Hub (MRC reference: MR/W014386) and the UK Prevention Research Partnership (Violence, Health and Society; MR-VO49879/1), an initiative funded by UK Research and Innovation Councils, the Department of Health and Social Care (England) and the UK devolved administrations, and leading health research charities. The views expressed are those of the author(s) and not necessarily those of UKRI, NIHR or King’s College London.
Declaration of interest
R.S. declares research support received within the last 3 years from GSK and Takeda. All other authors declare no conflicts of interest.
Open access
eLetters
No eLetters have been published for this article.