Ethnicity

We found that Black and Asian patients reported higher satisfaction across several quality-of-life and treatment-experience domains compared with White patients. However, Asian patients reported lower physical health satisfaction in 2021–2022. This is consistent with reports of increased morbidity resulting from the COVID-19 pandemic reported among Asian people. ^{Reference Sapey, Gallier, Mainey, Nightingale, McNulty and Crothers26}

At discharge, no significant differences were observed between ethnic groups. This may be due to small sample sizes (only 240 discharge records across both years, of which 97 were White, 44 Asian and 57 Black). Alternatively, the effect of treatment may have mitigated earlier disparities.

Research consistently shows health inequalities among minoritised ethnic groups. However, most research focuses on poor public health outcomes, ^{Reference Raleigh27} physical health outcomes ^{Reference Shah and Kanaya28} and barriers to accessing care, ^{Reference Kapadia, Zhang, Salway, Nazroo and Booth29} alongside experience of primary care. ^{Reference Hayanga, Stafford and Bécares30,Reference O’Dowd31} NHS Digital data have shown a greater likelihood of Black ethnic groups receiving treatment following detention under the Mental Health Act. ³² On the other hand, data from primary care patient experience statistics (GP services) show that Black ethnic groups (both Black African and Black Caribbean) reported a positive overall experience (often similar to, or slightly below, the level of White British patients). For example, in the 2020–2021 UK GP Patient Survey, Black African patients had the highest reporting level of a positive overall GP service experience (at 86.2%), whereas Pakistani and Bangladeshi ethnicities reported the lowest rates of satisfaction. ³³

In other published research, worse outcomes for ethnic minorities are partly explained by associations between ethnicity and socioeconomic status: minoritised communities are more likely to suffer from socioeconomic deprivation which, in turn, is associated with poor health outcomes. Nevertheless, ethnicity-based health inequalities persist even after adjusting for deprivation. ^{Reference Evandrou, Falkingham, Feng and Vlachantoni34}

Challenges to accessing care for minority ethnic groups are due to a range of reasons (perceived and institutional), including health beliefs within the community, stigma and service configuration. ^{Reference Winsper, Bhattacharya, Bhui, Currie, Edge and Ellard35} When researching SMI populations, most research focuses on patient experiences in in-patient units (alongside coercion). ^{Reference Bennewith, Amos, Lewis, Katsakou, Wykes and Morriss36} Delay in help-seeking for mental health might further contribute to increased acuity and perpetuate coercion. Perceived lack of trust and poor experience have been identified in qualitative studies. ^{Reference Bhui, Mooney, Joseph, McCabe, Newbigging and McCrone37}

However, there is little research with the community on mental health patients with SMI. The Patient and Carer Race Equality Framework highlights the need to examine inequities in access, experience and outcomes. ³⁸ Unfortunately previous large-scale PROM data analyses did not explore differences between groups. ^{Reference Bull, Teede, Watson and Callander5–Reference Bull and Callander7,Reference Mosler, Priebe and Bird19,Reference Amat-Fernandez, Pardo, Ferrer, Bosch, Lizano-Barrantes and Briseño-Diaz39–Reference Coelho, de Bienassis, Klazinga, Santo, Frade and Costa41} We did not find any published research on variation in PROMs and PREMs scales in a community SMI cohort.

Minoritised ethnic groups are diverse, and the way in which a particular minority group interacts with the wider society is often unique; interpretation of findings is therefore complex. Our findings contrast with those from the USA of Kaur et al, who reported that, compared with White non-Hispanic patients, ethnoracial minorities had worse outcomes in regard to mental health after adjusting for age, gender, education, employment, marital status, comorbidities and insurance type. ^{Reference Kaur, Zeng, Malapati, McCleary, Meyers and Bryant42} There is a risk of oversight or misunderstanding if we group all minority ethnic groups’ data in one box. We need to be mindful that there are variations across countries and systems, and that these may not be generalisable across countries. For instance, the New Zealand Attitudes and Values Study (2019) found higher satisfaction among Māori and Pasifika patients after adjusting for cultural respect and ethnic concordance compared with the European-descended majority. ^{Reference Lee and Sibley43}

People struggling with SMI often face severe social disadvantages. There is some evidence that poor, marginalised White people, especially young men, face significant adversities. Health inequity is complex. Our findings do not disprove the health inequities faced by minoritised communities, especially in regard to accessing treatment, facing discrimination or having disproportionate experience of coercive treatment. Our findings, however, offer the perspective of minoritised communities’ experience of quality of life and its change through treatment and treatment experience. DIALOG scales are more likely to have been collected from those who engaged in overcoming the barriers to access. We found that, when patients with SMI engaged with care and treatment, those from minoritised communities had good outcome and experience. This offers hope that, when engaged, there is no reason to aspire and expect positive outcomes in patients with SMI from minoritised communities. Possible explanations of our findings may be related to local improvement in service provisions in east London, with several decades of research and work focusing on cultural sensitivity ^{Reference Bhui, Owiti, Palinski, Ascoli, De Jongh and Archer44} and co-production. We believe that mental health services should continuously monitor and analyse outcomes to investigate how those facing the most adversities are being served.

Gender

Men reported lower mental and physical health satisfaction in 2021–2022 versus 2018–2019, but were more satisfied than women in multiple domains including health, social, safety and experience; women were more satisfied than men in one domain, job situation. Our findings of women reporting higher levels of dissatisfaction with their care is not consistent with primary care surveys. ⁴⁵

At discharge, men were less satisfied in 2021–2022 compared with 2018–2019, especially in regard to medication and their experiences with mental health professionals.

Socioeconomic deprivation

Patients from more deprived socioeconomic areas had lower satisfaction across several domains of quality of life; however, the impact on the mental health domain was not statistically significant.

We used IMD as a proxy measure of deprivation. IMD does not offer the most sensitive measure at the individual level, and is known to be a relatively insensitive and blunt tool in capturing true deprivation. ^{Reference McCartney, Hoggett, Walsh and Lee46} Future research should investigate improved measures to capture socioeconomic deprivation more accurately. The Patient and Carer Race Equality Framework highlights the need to examine inequities in access, experience and outcomes. ³⁸

Age

Our finding of older population reporting greater satisfaction is consistent with previous studies, e.g. in primary care where younger age is associated with lower levels of satisfaction. ^{Reference Kontopantelis, Roland and Reeves47}

Broader findings

DIALOG, as a self-reported outcome and experience measure collected from a subgroup of patients with SMI, may not be able to capture the full set of nuances and interacting factors that influence health inequalities. We are aware that one of the major barriers to healthcare resulting in health inequality is access. ^{Reference Winsper, Bhattacharya, Bhui, Currie, Edge and Ellard35} Routinely, DIALOG may have some limitations in detecting changes in satisfaction level if initial levels of satisfaction in a domain are high. Intersectionality and added disadvantages resulting from socioeconomic deprivation, immigration, language barriers and poor physical health outcome may not be completely adjusted through the logistic regression carried out.

Limitations

Although we recognise that differences between groups in a domain could represent a variation caused by a range of potential factors including perceived need and dissatisfaction, which can be a proxy for health inequity, we also recognise that the findings can be influenced by a range of factors and confounders. The proportion of patients for whom we had access to their DIALOG scores was low (15–16%).

It is also likely that those who were more satisfied with treatment and engaged with services introduced a positive bias into the pooled results. However, this is likely to hold true for all subgroups and therefore should not particularly bias the analysis between subgroups.