Qualitative study of candidacy and access to secondary mental health services during the COVID-19 pandemic

Abstract

Candidacy, a construct describing how people's eligibility for care is negotiated between themselves and services, has received limited attention in the context of mental health care. In addition, candidacy research has only rarely studied the views of carers and health professionals. In this article, we use concepts relating to candidacy to enable a theoretically informed examination of experiences of access to secondary mental health services during the first wave of the COVID-19 pandemic in England. We report a qualitative study of the views and experiences of service users, carers, and healthcare professionals. Analysis of 65 in-depth interviews was based on the constant comparative method. We found that wide-ranging service changes designed to address the imperatives of the pandemic were highly consequential for people's candidacy. Macro-level changes, including increased emphasis on crisis and risk management and adapted risk assessment systems, produced effects that went far beyond restrictions in the availability of services: they profoundly re-structured service users' identification of their own candidacy, including perceptions of what counted as a problem worthy of attention and whether they as individuals needed, deserved, and were entitled to care. Services became less permeable, such that finding a point of entry to those services that remained open required more work of service users and carers. Healthcare professionals were routinely confronted by complex decisions and ethical dilemmas about provision of care, and their implicit judgements about access may have important implications for equity. Many of the challenges of access exposed by the pandemic related to pre-existing resource deficits and institutional weaknesses in care for people living with mental health difficulties. Overall, these findings affirm the value of the construct of candidacy for explaining access to mental healthcare, but also enable deepened understanding of the specific features of candidacy, offering enduring learning and implications for policy and practice.

Highlights

• •Candidacy is a framework for understanding influences on access to healthcare.
• •We studied access to mental healthcare services during the COVID-19 pandemic.
• •Candidacy was profoundly affected by service changes.
• •Understandings of entitlement and thresholds for access were reconfigured.
• •The value of the candidacy framework for mental health contexts is affirmed.

1. Introduction

Challenges in ensuring that people with mental health difficulties have care available to them at the right place and time, and in a form that addresses their needs, are not new (Lamb et al., 2015). Nor is the social patterning of mental health difficulties, which may particularly affect those most socio-economically disadvantaged and those most at risk of marginalisation (Ribeiro et al., 2017). Among the many adverse consequences of the COVID-19 pandemic has been its intensification of these challenges in provision. In the UK and globally, people living with, or at risk of, severe mental health difficulties, who are normally cared for in secondary mental healthcare settings (e.g. inpatient and community mental health services), have faced restrictions in availability of care (Chen et al., 2020; NHS Reset, 2020; Patel et al., 2020; Rethink Mental Illness, 2020), at the same time as the pandemic and associated control measures have increased need. Given these challenges, explaining access to mental healthcare in a theoretically informed way is an important goal.

Previous work has suggested that thinking of access solely in terms of availability, supply, and use of services is too limiting to offer depth of understanding. Offering an alternative approach, the construct of candidacy emphasises:

The ways in which people's eligibility for medical attention and intervention is jointly negotiated between individuals and health services. … [It] is a dynamic and contingent process, constantly being defined and redefined through interactions between individuals and professionals, including how ‘cases’ are constructed. Accomplishing access to healthcare requires considerable work on the part of users, and the amount, difficulty, and complexity of that work may operate as barriers to receipt of care. The social patterning of perceptions of health and health services, and a lack of alignment between the priorities and competencies of disadvantaged people and the organization of health services, conspire to create vulnerabilities. (Dixon-Woods et al., 2006)

Structured analytically around seven features – identification of candidacy, navigation, permeability of services, appearances at health services, adjudications, offers and resistance, and operating conditions (Table 1) – candidacy can be understood as a continually negotiated property of individuals that is subject to multiple influences. Candidacy is influenced, for example, by individuals themselves and their socio-economic contexts, macro-level structures and allocation of resources, how services seek to constitute and define the appropriate objects of healthcare attention and intervention, and the decisions and actions of those at the sharp end of providing care (Dixon-Woods et al., 2006).

Table 1.

Understanding candidacy for mental health.

Feature	Original version	Mental healthcare (updated in light of study findings)
Identification of candidacy	How people recognise their symptoms as needing medical attention or intervention is important to how they assert a claim to candidacy.	How people make judgements about seeking help for their mental health from services (which need not be exclusively medical) is important to whether they see themselves as candidates who need, deserve, and are entitled to attention and care. These judgements can be strongly impacted by macro-level changes that mean people internalise views of what constitute “worthy” objects of care and attention.
Navigation of services	Using services requires knowledge of the available services and depends on having the practical resources to avail of them.	Using services demands that service users and their carers have knowledge of services and how they operate, the social capital necessary to navigate the system, and access to the material and practical resources needed to find a point of entry to services.
Permeability of services	The ease with which people can use services depends on how many and what kinds of criteria people have to meet to avail of them, and on cultural alignment between services and individuals.	The ease with which service users and their carers can use mental healthcare services depends on how well configured services are to meet the specific needs of those with mental health difficulties, the criteria being used, the degree of cultural alignment, and number and type of barriers to access.
Appearance at services	Appearing at services involves people making a claim to candidacy. It requires a set of competencies and social/cultural alignments.	Appearing at services involves people making a claim to candidacy. It demands a set of competencies and socio-cultural alignments that may be particularly challenging for those with mental health difficulties.
Adjudications by professionals	Professional judgements about candidacy strongly influence access to attention and interventions, and depend in part on a repertoire of judgements but also on operating conditions and resource constraints.	Professional judgements about candidacy strongly influence access to attention and interventions, and depend in part on a repertoire of judgements. Operating conditions and resource constraints may powerfully constrain these judgements and may at times result in moral injury.
Offers and resistance	Offers of care may be made that may be accepted or refused by individuals, sometimes because they wish to resist the nature of the care on offer.	Offers of care may be made that may be accepted or refused by individuals, sometimes because the care offered is not seen as helpful or acceptable. Recursivity – previous experiences of poor care – may be an important influence.
Operating conditions and the local production of candidacy	The perceived or actual availability and suitability of resources has a major impact on the local production of candidacy, as do other relevant operating conditions.	Macro-structural influences have major impacts on people's candidacy for care, and reconfigurations of systems may be particularly consequential not only in restricting availability of supply, but also in refiguring how individual service users, carers and staff construct people's eligibility care.

A large literature on candidacy has now developed, mostly affirming its value in systematically accounting for access to healthcare and how vulnerabilities may arise in relation to it (Koehn, 2009; Mackenzie, Conway, Hastings, Munro and O'Donnell, 2013), while also deepening and refining understanding of the original construct. One important addition, for example, is the concept of recursivity, which describes the interdependency between people's experiences of health services and their future actions in relation to help-seeking (Hunter et al., 2013). Also important is the concept of identity, which describes individuals' sense of self and how it is maintained in interaction with others (Macdonald et al., 2016). Threats to identity can arise when people feel stereotyped or disempowered, when their subjective experience is ignored, or when moralised judgements are made about them. However, some important gaps in the literature are evident.

One of these gaps arises in relation to carers and healthcare professionals. They remain typically under-represented in candidacy research thus far, which has mostly focused on the experiences of candidate help-seekers themselves (van der Boor and White, 2020). A second gap relates to the range of health conditions addressed by the literature on candidacy. For mental health, an area with a number of distinctive features, it has remained small (Gask et al., 2012; Chinn and Abraham, 2016; Kovandžić et al., 2011) and has largely neglected services for those with enduring severe mental illness. In the UK, these services are typically provided by secondary mental health services (inpatient hospital and community care) (Box 1). Third, the COVID-19 pandemic has had significant impacts on access to healthcare, including restrictions on availability of services and shifts from in-person to remote forms of care (Rethink Mental Illness, 2020), which may have also impacted on candidacy.

Box 1. Secondary mental health services in the English NHS.

Care provided by the English National Healthcare Service (NHS) is free at the point of delivery to UK residents and covers certain types of mental healthcare services. Secondary mental healthcare services require an initial referral from a general practitioner and are normally provided by a specialist NHS trust. These services include community mental health teams, crisis and home treatment teams, assertive outreach teams, early intervention for psychosis services, and mental health inpatient units. Secondary mental health services usually provide support to people with enduring, moderate to severe mental difficulties.

Alt-text: Box 1

In this article, we address these gaps by using the candidacy construct to enable a theoretically informed examination of access to secondary mental health services during the COVID pandemic through a qualitative study of the views and experiences of service users, carers, and healthcare professionals.

1.1. Methods

Between June and August 2020, we undertook a study using semi-structured interviews as the data collection method and the constant comparative method as the analytic strategy (Charmaz, 2006). The study was designed in conjunction with six experts-by-experience advisors (three service users and three carers) and a peer researcher from the McPin Foundation, a mental health research charity.

We interviewed three groups of people: individuals with mental health difficulties who either accessed secondary mental healthcare during the pandemic or needed these services but did not access them; informal carers of people with mental health difficulties; and point-of-care staff working in NHS secondary mental health services. We recruited in England only. We did not include individuals seeking to access mental healthcare for the first time through primary care, nor staff working in primary mental health services.

Our purposive sampling strategy (Palinkas et al., 2015) aimed to represent diversity. We used a combination of network-based and snowball approaches across multiple channels. People interested in taking part were asked to complete a form which included information about their ethnicity, gender, location in England, and job role (for staff). To maximise diversity, we prioritised invitations based on this information. As data collection and analysis progressed in parallel, the size of the sample was adapted in line with the principle of information power (Malterud et al., 2016).

Potential participants were contacted via phone or email, depending on people's preferred contact method. They were given a link to register and consent on Thiscovery, a secure online research platform developed by THIS Institute according to the AA Web Content Accessibility Guidelines. Four experienced researchers (EL, JP, NR and JW) conducted qualitative interviews with service users, carers and mental healthcare staff. To comply with the UK lockdown regulations in place at the time, all interviews were conducted remotely. Participants could choose to be interviewed over the phone or through Thisovery, using a video-supported secure software. Interviews lasted an average of 40 min (range 22–95 min). A researcher with lived experience (Sweeney et al., 2009) of accessing secondary mental health services carried out all interviews with service users, as well as contributing to analysing and writing up the findings. Service user and carers participants were compensated £25 each.

Analysis of the transcripts by a five member-team (EL, NR, JP, NB, DS) was based on the constant comparative method (Charmaz, 2006). We combined a deductive approach using the seven features of the candidacy framework as sensitising concepts (Charmaz, 2006) with an inductive approach to generate open codes. We recursively interrogated the data to identify similarities and differences, patterns and relationships, and points of departure and convergence. The final organising themes were used to process the data, facilitated by NVIVO software.

JP and NB analysed service user interviews, EL analysed carer interviews, and NR, DS, NB and EL analysed staff interviews. We maintained rigour through a range of techniques, including regular cross-checking between researchers and team debriefings at every stage of the research (data collection, open coding, organising themes). Ethical approval for the study was obtained from the University of Cambridge Psychology Research Ethics Committee. All participants were provided with information about the study and gave consent.

1.2. Results

We had a strong response to recruitment efforts (Table 2), receiving 220 expressions of interest. A total of 139 individuals were invited to take part in the study; 70 either did not respond or declined. A final number of 69 people were interviewed (24 service users, 10 carers, and 35 members of mental healthcare staff). Four interviews were excluded from analysis because further information emerged during the interview indicating they did not meet the inclusion criteria.

Table 2.

Interview sample.

	Service users	Carers	Staff	Total
Individuals interviewed	24 (4 excluded from analysis)	10	35	69
Expressions of interest received	60	18	142	220
Gender	•8 Women •7 Men •2 Non-binary •3 unknown	•6 Women •2 Men •2 unknown	•19 Women •11 Men •5 unknown
Ethnicity	•8 White •3 Black •2 Asian •4 Mixed ethnicity •3 unknown	•7 White •1 Asian •2 unknown	•24 White •3 Asian •2 Mixed ethnicity •1 from ‘any other ethnic group’ •5 unknown
Region	•1 North West •5 East Midlands •6 Greater London •2 East of England •1 South East •2 South West •3 unknown	•1 West Midlands •4 East of England •2 South East •1 South West •2 unknown	•4 North East •7 North West •2 East Midlands •5 West Midlands •4 Greater London •2 East of England •3 South East •3 South West •5 unknown
Additional information	The most common diagnostic categories were psychotic, bipolar, and personality disorder. Less frequent diagnostic categories included depression, anxiety, agoraphobia, obsessive-compulsive disorder, autism, complex Post-Traumatic Stress Disorder (PTSD), COVID-related PTSD. Many participants had comorbidities/reported more than one mental health diagnosis.		•17 Psychiatrists (including 13 trainees and 4 consultants) •10 Mental health nurses (including care coordinators, matrons, non-clinical prescribers) •8 Clinical psychologists (including CBT therapists and systemic family therapists) Services covered: •Community Mental Health Teams (CMHT) •Early intervention for psychosis (EIP) •Crisis Teams •Acute hospital wards •Secure Forensic services

Our analysis affirmed the overall salience of the construct of candidacy in explaining access to mental health care during the pandemic. It confirmed several known features of candidacy in this context, showing that access to mental health services is the outcome of iterative and reinforcing processes, and is subject to dynamic, shifting and constantly renegotiated interplay between care-seeking behaviours and practices, the macro-level structures and systems of care provision, and micro-level interactions with services and professionals. A distinctive impact of the pandemic, however, was the impacts on candidacy of macro-level reconfigurations of mental health services introduced in response to COVID-19. These changes not only restricted availability of supply, they also crucially remade how service users, carers, and staff constructed people's eligibility for secondary mental healthcare services.

To accommodate the findings of our analysis in relation to the specifics of mental health services, the inclusion of the perspectives of informal carers and staff, and novel insights arising from the impact of the pandemic on candidacy, we updated the original seven features of the candidacy framework (Table 1). In the sections that follow, we explain each feature of the framework.

1.3. Operating conditions

In the original account, candidacy was seen to be influenced by operating conditions, which include, though are not limited to, the availability of resources, local pressures and policy imperatives (Dixon-Woods et al., 2006). In our study, it was clear that changes in the operating conditions of secondary mental healthcare brought about by the pandemic were rapid and far-reaching, with multiple implications for candidacy. Despite the large-scale character of the changes of services (including, for example, new criteria for resource allocation), particularly in the early stages of the pandemic response, it was notable that restructuring was undertaken without discussion or consultation with service users themselves or with staff. Below, we offer a summary of these changes based on the accounts of participants. In this section only, and in the interests of space, illustrative quotations are not supplied.

Our analysis identified four macro-level strategies that were deployed to address the problem that services could no longer be provided in pre-pandemic mode. One involved withdrawing some services – particularly those officially deemed ‘non-essential’. Participants reported that these services variously included family therapy, complex trauma work, the at-risk mental state pathway (which offers preventive interventions for people at risk of psychosis), day facilities, outpatient departments, assessment services for speech and language therapy, and memory clinics. In some inpatient services, therapeutic interventions such as art therapy were restricted or stopped entirely, and social and outdoor activities were scaled back. Many forms of psychological intervention, occupational therapy, and employment advice were either withdrawn completely or were paused. A second strategy involved offering services in a different form – for example, remotely (e.g. by telephone or online video), or by offering much briefer, monitoring-style contacts rather than therapeutic interventions.

A third strategy involved re-prioritisation of individual service users for intensity of contact. For example, staff working in community health services were asked to reassign each service user to a risk category corresponding to how urgently, frequently, and proactively they were to be contacted in the new circumstances. A red, amber, green (“RAG”) rating system, corresponding respectively to high, moderate, or low risk, was sometimes used, based on factors such as risks to self and others, current or past mental health difficulties, medication-related factors, and wider social and family factors. This re-categorisation work was carried out either individually, by care coordinators considering their own caseloads, or collectively, during multidisciplinary team meetings. Individuals classified as high or moderate risk were offered relatively proactive and frequent contact, sometimes face-to-face, and were encouraged to contact services if they felt they needed support. However, those classified as low risk were expected to initiate contact themselves and were not contacted proactively.

The final strategy was to change the thresholds for admission to and discharge from inpatient services to address reduced capacity arising both from infection control measures and from diminished staff availability. Risk management for infection led to an increased emphasis on progressing individuals to be discharged rather than remaining in hospital. The option of ‘transitional’ discharge (where service users could leave the hospital for short periods of time before full discharge) was removed, making discharge decisions more final and abrupt.

More broadly, efforts to control infection transmission risks for staff, service users, and the wider community resulted in severe restrictions on face-to-face contact. Staff availability was reduced because of self-isolation, shielding, and sickness associated with COVID-19. Some staff had to be redeployed across the service to cover prioritised services, including inpatient wards. In settings where care was still being provided in-person (e.g. specialist inpatient services), infection control measures cut capacity and resulted in physical changes to space, as well as limiting outdoor facilities and therapeutic interventions.

These changes, taken together, decreased the availability of services and increased the selectivity of those on offer. At the same time, service user needs for care grew, exacerbated by the conditions of the pandemic and measures taken in response. As we show in the analysis that follows, the combined effect was to reconfigure candidacy.

1.4. Service users’ identification of their own candidacy

The original account of candidacy proposed that whether and how people recognise their symptoms as likely to require or benefit from professional attention is key to understanding how they assert a claim to services (Dixon-Woods et al., 2006). We found that service changes brought about by the pandemic powerfully impacted on service users’ identification of their own candidacy, including their sense of whether they deserved or warranted care.

Some service users reported that, in contrast to previous practice, they were asked to initiate contact themselves if they felt they needed support, usually by calling an emergency hub rather than a known healthcare professional. The practical effect of this was to place the initial onus for determining eligibility on to people living with mental health difficulties. The individuals we interviewed often lacked guidance on what might be legitimate grounds for seeking care, did not always feel able to advocate for themselves, were influenced by public health messaging to ‘protect the NHS’, or felt that their needs were invalidated by media reporting.

Services did not discuss the exact meaning of being ‘in need’ with service users, leaving them uncertain about what level of psychological distress could be seen as needing support. For example, service users sometimes perceived that events that would normally be seen as life crises (such as a bereavement or a relationship breakdown) were likely to be seen as insignificant in comparison to the global COVID-19 crisis, leaving them reticent to contact services even as their mental health worsened. Some people with great need imagined that they would be deemed ineligible by others, and their expectations led to their concealing their candidacy.

The psychiatrist … you feel like you contact her if you're in crisis, which actually I am, and I've not even told her about it … My [spouse] has said we're going to get divorced … So it is a crisis, but … there is something worse happening everywhere else. … You feel like, in perspective, your crisis is not that big. (Service user)

A striking feature of service users' accounts of their candidacy was a strong emphasis on their perceptions of their “deservingness”. Some mental health difficulties left service users particularly hesitant to claim candidacy: for example, a low sense of self-worth combined with repeated assertions about the strains on services and the risks of contact left some service users reluctant to see themselves as warranting care. Some reported feeling that others were worse off and consequently more deserving, to the extent that even people classified as ‘vulnerable’ during the pandemic struggled to describe themselves as a priority. Feelings of being burdensome (the belief that one is an inconvenience to others and society), often associated with hopelessness and increased risk of suicide (Van Orden et al., 2010), were heightened. People found that making decisions about whether to ask for support, and from whom, was particularly difficult when they were unwell or becoming unwell – ironically, when they most needed help. Lacking the support of familiar professionals or access to informal carers compounded the challenges.

Maybe I just I feel like I'm bothering [staff]. (Service user)

It's a bit weird because when I'm really, really depressed I can't see myself as a priority, and I see that other people are needing more support. (Service user)

Staff themselves recognised this phenomenon, reporting concerns about what they called the ‘silent’ individuals, including those with low self-esteem might not see themselves worthy of help and might therefore avoid contacting services proactively.

If you are from a family that are quiet and just get on with it, even though you're struggling, your voice might not be heard in amongst all of that, so they might get missed. (Family therapist)

One [service-user] contacted me, and he was like, ‘Oh, I really am so sorry to bother you, I know you're so busy.’ And I was thinking I'm paid to be here for you, so you don't need to apologise for ringing. But you could tell that he felt like a pain for ringing, which was really awful. And yet I was saying, ‘Oh no, don't worry about it’ … but knowing, at the same time, we had given the message that we were too busy to speak to people. (CBT therapist)

Besides an internalised sense of eligibility for care, service users’ understandings of what was likely to be on offer if they did seek help, or how they were now expected to make contact, also impacted on their help-seeking behaviours. Some, for example, were reluctant to seek help because of fear of speaking to a professional unknown to them.

Even though they do ask ‘Do you need a call?‘, because you don't know the individual, I feel it's quite hard to say ‘Actually it would be quite nice to have someone to give me a call’. It's just harder to ask for support when it's just some random person. (Service user)

Expectations of being ‘rejected’ by services, deemed ineligible, or stigmatised, or of not being able to obtain the help needed, further impacted on willingness to seek support. These expectations and fears were often recursive, building on cumulative negative experiences.

When I'm at my lowest point I feel completely suicidal, and I feel like my world has ended, and then it takes a lot to build up to asking for help. And then when you actually do ask for help and then you're not taken seriously then it can have a huge impact that I myself have to pick up those pieces and I have to be responsible for that. (Service user)

S/he said s/he's not doing any therapy with me. So I thought, if it's just a check-in and we have nothing to say, what's the point? So I was kind of angry. I wanted help but didn't know how to ask. (Service user)

Service users and carers did recognise the exigencies that had forced the need for prioritisation, but commented on its complex, multi-dimensional, and ‘tricky’ nature. They argued for a more holistic and individualised approach – one that would take into account people's wider social context, rather than solely their mental health symptoms, drawing on staff's personal knowledge of service users and the factors that might trigger mental health symptoms. Among the groups they felt should be prioritised for support, they frequently mentioned people from ethnic minority communities, those affected by wider societal disadvantages, those at risk of gender-based and domestic violence, and those with complex healthcare needs who had to isolate from families or social support.

I feel they've got to talk to people. You can't prioritise if you don't know what's going on … You can make decisions that you think are right, but actually, who have you asked? (Carer)

1.5. Navigation and permeability

Once a need for help had been identified by service users (or in some cases, their carers), they had to gain a point of entry to services. The original formulation of candidacy and subsequent analyses have shown that significant work may be needed to find a way to and through the healthcare system (Dixon-Woods et al., 2006; Koehn, 2009). Successful navigation may depend on resources, commitment, perseverance, and competence in negotiating complex and often disjointed and frustrating systems that may be unevenly and unfairly distributed. Services with low permeability (hard to gain entry) may require meeting certain criteria (such as having a referral), and perhaps require a higher degree of cultural alignment, for example in relation to how far people feel comfortable with particular organisational values and modes of provision. More porous services, on the other hand, require fewer qualifications of candidacy to use them, and may require the mobilisation of fewer resources.

Participants in our study reported that service changes in response to the pandemic reduced permeability, meaning that finding a point of entry to those mental health services that remained functioning might be much more demanding and require far more effort and resource to negotiate. With the shift to remote care, for example, people with limited access to phones or other communications technology, and those who lacked secure accommodation or who lived in material deprivation, were particularly challenged:

Some of my patients only have phones with credit on, and unless I call them, they can't even call me. So even if they have a problem, it's dependent on when I have the time to call them. (Care co-ordinator)

Participants reported serious difficulties in working out what was (still) available and how to access it. No service user or carer reported being consulted or involved in the reorganisation of mental health services during the pandemic, and many reported that communication was poorly managed.

The problem was that as my [family member] was very unwell, we had heard nothing from Community Mental Health. And the Government, and every message you got was, ‘Stay at home, don't do anything, just shut yourselves in’. So it may seem strange in hindsight, but you sort of felt, ‘Well, we can't do anything. We've just got to sit this out.’ … […] It took them at least three weeks to get a letter out saying ‘We are functioning, these are our phone numbers’ etc. (Carer)

Some carers and service users assumed that no mental health support was available, or could not readily identify how to get support. This left them managing high levels of uncertainty about how to get help, including for crises and for medication.

I am a bit worried that I have absolutely no idea how services would respond if [family member] is in crisis, if she takes an overdose. I mean, I'm presuming that, you know, we'd still be able to dial 999 and get somebody out. (Carer)

When service users did attempt to contact services, some experienced severe problems linked to decreased permeability arising from the requirement to go through crisis teams or other triaging services, or simply not being able to get through.

One day, during this period, I thought I really needed to go … into psychiatric hospital again, and phoned them and couldn't get through – and gave up, really. (Service user)

I was really feeling suicidal and I rang 111 [NHS non-urgent medical helpline] and they said, oh, we'll ring you back when we've got someone. And then they took about an hour to ring back, but by then I had already overdosed because I was feeling so bad and I didn't know when they were going to ring back. (Service user)

Consistent with the concept of recursivity, these kinds of barriers compounded some service users’ previous negative experiences of accessing services. It resulted in some losing trust in services and disengaging altogether. Carers similarly highlighted the severity of harm caused when service users could not mobilise support and worried about increases in risk-taking behaviours:

If they're desperate to get care, and they realise they don't get it, they do just take more and more extreme [actions]. You can see that pattern happening all the time, but [COVID] was like an extra turn of a screw – it was like, it went up another notch. (Carer)

1.6. Appearing at services, adjudications, and offers

In the original account of candidacy, appearing at health services involves people asserting a claim to candidacy for healthcare attention or intervention. Making such claims involves work and resources, and requires a set of qualifications or competencies (defined from the perspective of the system), including the ability to formulate and articulate the issue for which help is being sought and to present credibly. Service users in our study reported multiple challenges in relation to making appearances at health services. One challenge was that their interpretations of what constituted a crisis worthy of immediate action were not always the same as those of services, and were sometimes complicated further by absence of face-to-face contact or continuity of care. Accordingly, though many individuals and their carers were desperate for care, they found that offers of care that aligned with their needs were not made.

They deemed it not a crisis … They said ‘Well, you know, like a change of medication or resuming a medication that you're already on … that's not really a crisis, is it? So, we'll give you the number for the secretary and they'll be able to do it.’ (Service user)

Once someone has appeared at a service, the candidacy framework suggests that judgements or adjudications are made by services that strongly influence subsequent access to attention and interventions. These adjudications may draw on repertoires of routine judgements or typifications about how to characterise the candidacy of individuals and determine which offers of help can be made (Dixon-Woods et al., 2006). Our interviews suggested that professionals’ adjudications about secondary mental healthcare – and the offers they made as a result – were powerfully impacted by the operating conditions of the pandemic, including their sensitivity to resource constraints, service capacity and infection risk. For example, only service users with very severe mental illness were seen as qualifying for inpatient admission, altering the previous basis of adjudications.

Before [the pandemic] we would have [admitted] people who were acutely unwell, but now that seems to have upped a little bit more, there has definitely got to be that increased risk to themselves or to others. (Nurse)

Though access was restricted, uncertainties and ambiguities over thresholds for inpatient admission and discharge for people who were acutely unwell with mental health difficulties were also reported by service users, carers, and staff.

It was a bit chaotic … So, for example, when [my family member] got in … you would get people swearing blind that [s/he] was going to be taken in to inpatient care, and then s/he wouldn't … People often seemed to be quite confused. (Carer)

From the perspective of staff, a major influence on decisions about admission was their awareness that the care that was being offered in inpatient services during the pandemic was less than ideal (e.g. because of restrictions on physical exercise and social activities) and could potentially have negative effects on service users’ mental health. These considerations seemed to have led staff, on one hand, to have higher thresholds for hospitalisation, and led some service users, on the other hand, to decline offers of admission (unless admitted on a compulsory basis under the Mental Health Act). Some service users reported that they evaded some offers of help because of the conditions in which assessments and decisions were being made.

It's not the same kind of therapeutic experience that people used to have … I think that changed our perspective in terms of admitting or not admitting, because sometimes it was not beneficial … [Service users] would say ‘I just want to go home because I feel like I'm going backwards by staying here’ … So you start [thinking], okay, do they really need to be in a restricted contained environment? (Trainee psychiatrist)

I had [contact with] a new [healthcare professional] who didn't know me as well. S/he was saying things to me like, ‘Oh I think you're a bit racy [hyper]’, but I was like, ‘No, no I'm fine, I'm fine.’ Whereas I think … face-to-face it would have been different. (Service user)

Changes in decisions about hospital admissions and discharge, along with staffing and capacity issues and the move to remote working, created increased pressures and dilemmas in community care. For example, faster discharges from hospitals meant that many service users presenting to community services were more acutely unwell; often the transitions from inpatient to community care were more fractured.

People [are] coming in much more poorly … The crisis teams and the wards have been under pressure, so we've been getting discharges happening very, very quickly … so the patients and their families are feeling quite bewildered. (CBT therapist)

They always do try and discharge people as soon as they can. I really appreciate why – I think it's really important. But I did get the feeling that this time they did discharge [my family member] much more rapidly … And quite honestly … I felt that it was extremely obvious that [family member] wasn't going to cope. So, I think they were clearly trying to keep people out. (Carer)

Several participants described community services entering into a state of emergency functioning; the thresholds for accessing psychological services for both new and existing service users were reported to be higher or less flexibly applied.

We didn't have the capacity to assess like we might have before. I think we've always had that really low threshold, and we've tried to work with the values of EIP [Early Intervention for Psychosis] and kind of making sure that there's nothing there, and through the pandemic we've certainly tightened our criteria a bit. (CBT Therapist)

Many professionals reported that their work with service users became focused more on containing risks than achieving improvement in their mental wellbeing, with some reporting that the therapeutic function of their role was severely compromised. They also reported that lack of face-to-face contact made it difficult to identify and address significant changes in service users' mental health before they reached crisis point. Lack of consensus on criteria, and how judgements should be made, led to ambiguities and uncertainties. Professionals were often in the unwelcome position of having to make adjudications and offers of care that they felt were poorly aligned with their professional values and judgements and their understanding of the service users’ needs.

Any routine assessments, such as autism assessment … any routine therapy got stopped … Appointments have generally been checking appointments really … kind of trying to help them tread water rather than make any gains or benefit. (Trainee psychiatrist)

Many found the decisions and actions they had to take as professionals deeply troubling, reporting their experiences as forms of moral injury (Liberati et al., 2021), understood as “perceived violation of one's own professional integrity and obligations and concurrent feeling of being constrained from taking the ethically appropriate action” (Lamiani et al., 2015). They reported that the rationale underlying the changes in service was not always clear and that they had not been consulted, yet they were acutely conscious of the impact of reduced services and of prioritisation systems on service users and carers. They expressed serious concerns about the potential for long-term damage, especially linked to withdrawal of psychological and social support services.

For me, that was a real ethical dilemma – because you're talking about patient care … What type of support would those people get if they weren't coming to a day centre? How would the carers be able to cope? (Clinical lead)

So, I'm thinking about people with mild learning disabilities, who don't meet the criteria for statutory services, but are quite heavily reliant on third sector social groups and such like, and who are quite psychotic … And I've been quite uncomfortable with that kind of work just being left. (CBT Therapist)

As time progressed, several point-of-care staff members became increasingly uneasy with adopting risk-categorisation systems solely using formalised criteria based on information available during the initial prioritisation exercises. They reported using their ‘gut feelings’ and previous knowledge of service users to decide when and how to make contact, suggesting that adjudications may also demonstrate recursivity. For example, when staff members were concerned that service users categorised as ‘green’ might deteriorate, some decided to upgrade these individuals' risk category to justify making proactive contact. Others invited service users to contact them directly, rather than phoning a centralised emergency hub. These deviations were not officially sanctioned but often came at a cost for staff in terms of time, energy, and emotional labour.

[My employer] wanted us to lose about 50% of our caseload [through prioritisation], which in a psychosis service is a difficult thing to do because people aren't in a psychosis service for no good reason. (…) It was quite difficult. I resisted for a while because I figured, ‘Well, I'm home just phoning people anyway.’ So I couldn't really see the harm in keeping contact. (Care co-ordinator)

2. Discussion

Our analysis confirms the value of the construct of candidacy in understanding access to secondary mental healthcare, including during a pandemic. By characterising more fully the effects of macro-level policies and operating conditions, this study shows that changes in mental health services in response to the pandemic did not simply reduce availability and supply, but also changed perceptions of what and who are candidates for care. Service users’ sense of eligibility, their right to access care, and their own “worthiness” (Chase et al., 2017) was remade by their sensitisation to what now counted as a problem deserving of attention and priority. The suppression of candidacy was, accordingly, not a simple function of the reduction in the supply and availability of services, but also a function of how people began to internalise what they saw as (often) moralised classifications of their need and worthiness. Further important learning concerns the significance of moral injury and moral distress as a feature of adjudications about allocation of mental healthcare. Though conducted during the COVID-19 pandemic, this work is of enduring relevance. The findings should not be regarded as an extraordinary blip in response to extreme conditions: they have more far-reaching implications both for the construct of candidacy and for practice and policy.

The findings affirm, consistent with some critiques of the original account of candidacy (Mackenzie et al., 2013), the need to attend to macro-structural influences that shape access to care. Most obviously, issues of navigation and permeability that create barriers to care even during mental health crises may increase risks and erode the therapeutic alliance and mutual trust, and may be exacerbated by failure to consider communication needs. More broadly, how services are designed, the systems used for prioritisation, the impacts on people's sense of their eligibility for care, and the actions required of staff are highly consequential for candidacy. A contribution of this study to understanding candidacy is in identifying the profoundly structuring effects of reconfigurations of services on people's help-seeking behaviours and the care they are offered. At a time when people were particularly vulnerable to poor mental health, withdrawal and pausing of services, changes in admission criteria, routing of requests for care through central hubs, and introduction of modified prioritisation systems, all without consultation or clear communication explaining these changes, powerfully impacted on people's agency as individuals. Important here was illness identity (Macdonald et al., 2016): service users and carers perceived a collective ambivalence about the extent to which mental ill-health was seen (officially) as an appropriate and deserving object of attention.

Navigation became more difficult as services became more selective and less permeable. The quantity and complexity of work required to find points of entry to the system increased, as did the assets and social capital needed to use the systems and make appearances. Professionals had to make adjudications about the care to be provided, which was often perceived as unsatisfactory and ethically discomfiting for them. Candidacy, in these conditions, was highly recursive: it became cyclically reinforced (and often suppressed) through negative cues and experiences. Our work shows that by exposing service users, carers, and professionals to cues about what (legitimate) need looks like and how it will be met, macro-level shifts shape interpretations, decisions, and actions at the micro level, and reconfigures the identities of service users and professionals alike.

These findings deepen understanding of adjudications and offers as a feature of candidacy. The changes in response to the pandemic meant that both service users and professionals were confronted with new and recalibrated criteria for determining eligibility for services, with important implications for help-seeking post-pandemic. Service users, though often desperate for care, were faced with a much more restricted range of offers of care and lack of clarity about how adjudications for their eligibility were made. The apparently low level of rejections of offers by service users, in contrast with some other studies (Pétrin et al., 2021) is perhaps suggestive of a problem of under-supply of care rather than enthusiastic endorsement of the offers that were made or an acceptance that the offers aligned with their needs. Our findings also offer important insights into how the adjudications were made. Though some previous accounts of treatment decisions made by professionals tend to see them as the outcome of biases, including discriminatory judgements of people's moral worth (Spencer and Grace, 2016), the staff we interviewed were often deeply troubled by inadequacies in services and highly sensitised to the possible consequences of their actions. Having to make adjudications on the basis of unclear criteria, or because they were aware that some forms of care could impose further harms, or because service constraints limited what they could offer, sometimes resulted in painful forms of moral injury. This finding aligns with persistent themes of other analytical approaches deployed to study how access to public services are governed. The quandaries faced by mental health service professionals about how to interpret newly developed eligibility criteria such as traffic light systems, for example, resonate with examinations of the work of street level bureaucrats (Lipsky, 1980) as gatekeepers to service provision. Our findings help to respond to Chase et al.‘s call to move beyond “simplistic representations of health professionals as extensions of formal regulations governing entitlement” (Chase et al., 2017).

Overall, these findings add to understanding of the dynamic and recursive nature of candidacy (Pétrin et al., 2021), among other things further challenging a view of candidacy as involving a linear series of “stages” – the seven features are each relevant in producing candidacy, but not necessarily in easily predictable ways. For policy and practice, our findings suggest that the pandemic exposed and amplified multiple challenges and structural inequalities in relation to mental healthcare provision and organisation of services that pre-dated COVID-19. Mental health services are often in short supply and decisions about who is going to receive (or not receive) care – and why – may be made in ways that are not explicit (Gask et al., 2012). The pandemic put a spotlight on some of the processes at work, suggesting that preventive and therapeutic functions are particularly at risk of being de-prioritised, while services that remain may become increasingly preoccupied with containing risk rather than achieving improvement in service users’ mental wellbeing. As well as ongoing attention to the resource available to mental health services and issues of “parity of esteem” with physical care services (Millard and Wessely, 2014), the rationale behind these decisions should be transparent and subject to deliberation. Further, our findings emphasise the need to include service users, carers and point-of-care staff in service redesign and evaluation to avoid opacity, identify what changes mean in practice, and secure legitimacy.

A second and linked policy implication is that attention is needed to the “street-level” decisions and actions (Lipsky, 1980) that staff take, particularly when they are seeking to compensate for deficiencies they perceive in care. These street-level behaviours may have helped services to better meet the emerging needs of some service users, and to some extent may have protected staff from moral injury. But they may also have had less positive consequences. Professional judgements about need and right to access care may be susceptible to various forms of individual bias. When combined with the phenomenon (noted by service users, carers, and staff alike) of ‘silent’ service users who are unwilling, unable, or inhibited in putting themselves forward, decisions of this kind may exacerbate some problems of access even while mitigating others. Renewed attention is needed to the importance of continuity of care and the role of familiar clinicians in reaching out to those less likely to assert their own candidacy for mental healthcare. More broadly, any situation where multiple competing demands on scarce resources must be resolved by staff at the sharp end should seek to anticipate and support the need for difficult decisions to be made and the consequences of making those decisions – for staff themselves, for individual service users and their carers, and for equity and inclusion. While such decisions have been the subject of extensive debate in areas such as allocation of ventilators (Maves et al., 2020), mental healthcare has not received the same level of attention as yet. As the field moves forward, literatures relating to social justice and relational ethics are likely to be of value in informing the development of theory and guidance.

Our study has a number of strengths and limitations. It is one of the few studies of candidacy to bring together the perspectives of service users, carers, and staff. It helps both in deepening and refining features of candidacy in the neglected area of mental health and in identifying practical implications and directions for future research. We managed to interview a wide range of participants from across England, though we did not seek a statistically representative picture, nor did we claim to cover all possible experiences of access. Our online-only recruitment approach was a choice that was made necessary by restrictions introduced during the COVID-19 pandemic. Though concerted effort went into mitigating the consequences of the “digital divide”, including giving people the opportunity to take part in interviews over the phone, accessing information about the study and completing the informed consent process still required access to the internet and thus may have biased our sample away from those lacking in means. Our study was conducted in the context of the English NHS, which is free at the point of use, and is limited in its ability to explore financial barriers to access.

3. Conclusions

Using a candidacy framework helps in explaining access to secondary mental healthcare services during the COVID-19 pandemic and beyond. It demonstrates that the fragile, recursive, dynamic, and contingent properties of candidacy as people's ideas of what constitutes need for services to respond are shaped by macro-level reconfigurations. It shows how people's interpretations and behaviours in relation to their symptoms are powerfully influenced by how problems are defined institutionally, as well as revealing the dilemmas that professionals may experience in seeking to reconcile officially sanctioned problem definitions with their values and sense of professional duty and responsibility towards service users. These findings have important implications for policy and practice in mental health services as well as for the construct of candidacy itself.

Credit statement

Elisa Liberati: Conceptualization, Methodology, Investigation, Formal Analysis, Project Administration, Writing - Original Draft, Writing - Review & Editing. Natalie Richards: Investigation, Formal Analysis, Project Administration, Writing - Review & Editing. Jennie Parker: Investigation, Methodology, Formal Analysis, Writing - Review & Editing. Janet Willars: Investigation, Writing - Review & Editing David Scott: Formal Analysis, Writing - Review & Editing Nicola Boydell: Formal Analysis, Writing - Review & Editing. Vanessa Pinfold: Conceptualization, Methodology, Writing - Review & Editing. Graham Martin: Writing - Review & Editing, Methodology, Supervision. Peter B Jones: Conceptualization, Methodology, Writing - Review & Editing, Supervision. Mary Dixon-Woods: Conceptualization, Methodology, Writing - Original Draft, Writing - Review & Editing, Funding acquisition.

Funding

This project was funded by THIS Institute's grant from the Health Foundation. The Health Foundation is an independent charity committed to bringing about better health and health care for people in the UK. All contracted parties contributed to the study under agreements through the same funding. PBJ is supported by the NIHR Applied Research Collaboration East of England and by RP-PG-0161-20003. Mary Dixon-Woods is an NIHR Senior Investigator (NF–SI-0617-10026). The views expressed in this article are those of the authors and not necessarily those of the NHS, the NIHR, or the Department of Health and Social Care.

Declaration of competing interest

The authors declare no competing interests.