The Resilience Hub approach for addressing mental health of health and social care workers during the COVID-19 pandemic: a mixed-methods evaluation

Efforts to diversify the staff consultation group

After the first meeting, a short, anonymous demographics form was sent to group members. While the group was diverse in terms of some characteristics (e.g. job role, sector, age and disability), all group members identified as being white women, although not all White British. This was most likely a result of initially approaching those accessing a Hub for support, amongst whom men and people from ethnic minority communities were under-represented (see Chapter 4).

The staff consultation group and the research team worked to diversify the group. Actions taken included targeted emails to Hub clients in under-represented groups, specifically highlighting that we sought to involve people who identified with a range of different demographic groups. RAs from each site also contacted local NHS Trusts’ ‘BAME Networks’ and equality, diversity and inclusion (EDI) leads, who provided advice around the wording of invitation e-mails. RAs attended network meetings for consultation, advice and promotion of the study.

For example, as part of this work, the RA at Site B also attended local BAME Network and EDI network meetings. At this meeting the RA introduced the Hub offer, the research study, and the need to diversify the Staff Consultation Group. These networks sent out information and invitations to be a part of the group via network meetings during online sessions and through the network mailing list. Other engagement included informal virtual meetings and e-mail correspondence with colleagues in EDI positions and NHS staff with a passion for reducing health inequality. The RA attended five BAME Network meetings and met informally with others outside the BAME Network (three monthly meetings with a smaller group of three stakeholders from December 2021 to February 2022; additional individual meetings with potential stakeholders who were interested in the research; meeting with champions in the community who worked with voluntary, community and social enterprise (VCSE) organisations to seek advice around engaging with under-represented communities).

Following these engagement efforts, two new regular members joined the Staff Consultation Group, one being from a white non-British background, and other from a minority ethnic group. However, the group nevertheless remained all women. Much of the involvement from people from minority ethnic communities was conducted in an informal manner due to the barriers with research engagement discussed in Equality, diversity and inclusion.

Outreach

Each of the Hubs used outreach to promote the services and increase uptake of individual and team support. The Hubs offered presentations to staff teams to promote Hub support, how to self-refer and invite staff to ask questions. The Hubs’ team-based interventions were also used to disseminate to staff teams, regular meetings with acute trusts, stakeholders, social care and local authority, and through e-mail contacts between organisations. Common methods of outreach included distribution of flyers, banners and business cards, information packs distributed throughout care homes and meetings with leads of partner organisations. Other methods of promotion included social media (Site B), promotional items (e.g. branded stationery, well-being pack; Site B and D) and slots on local media platforms (all sites).

Site D used a locality system whereby Hub clinicians were assigned to specific workforces/areas within the region. Site A launched a branding and communications campaign across a wide range of partnership organisations.

Site B employed a full-time Strategic Engagement Lead to direct and manage communications and outreach across the region. Stakeholders were invited to a monthly Partnership Engagement Forum to gain stakeholder engagement and support targeted outreach. A monthly newsletter was disseminated across the region’s networks at the local and strategic levels to reach as many staff as possible. Site D and B’s critical care leads participated in close planning with the regions critical care network to direct critical care specific outreach and promotion.

Outreach activities were conducted by each Hub to target specific staffing and demographic groups that had lower uptake of Hub support. Targeted outreach included visiting and providing materials to care homes (Sites B, C and D), producing information for care homes (Sites A and D), gathering direct e-mail addresses for care home staff, ambulance staff and hospices to promote the offer (Site C; Site D), developing bespoke social media graphics for different groups (Site B; Site D), promoting workshops or facilitated peer support (FPS) sessions for care home staff (Sites B, C and D) and men’s mental health (Site C; Site D). Meetings with NHS Trusts Equality Leads took place to promote the Hubs and target health and care staff from minority ethnic backgrounds (Site C; Site D), as well as creating links with race equality networks (Sites B, C and D). A race equality campaign was implemented to attempt to reach more staff within this group (Site B). Webinars were also conducted with the local Council of Mosques (Site A). Face-to-face check-ins were delivered with critical care staff within the region (Site D) and links made with the ambulance service Suicide Prevention Lead to attempt to increase the uptake from this group (Site B). Meetings with representatives from emergency services also took place (Site A), and further meetings were held with representatives from critical care units to identify how best to spend specific funding to support critical care staff (Sites A and D).

Website resources

Each Hub had a website, providing information about services, eligibility, psychoeducation materials and downloadable self-help resources and short webinars. The websites also provided details of other mental health services/charities available to staff, as well as crisis helplines for emergencies. Individuals could self-refer on each Hub’s website. Site A’s website also included testimonials from Hub clients to promote and normalise the service. Staff accessing Site B’s website could book onto peer support groups and well-being workshops, as well as sending feedback, testimonials and suggestions to the Hub.

Self-referral and mental health screening

At each Hub, prospective clients were encouraged to self-refer online, although other options were available. Mental health screening formed a part of the self-referral process, although this process varied across Hubs. Screening at each Hub included measures of post-traumatic stress [International Trauma questionnaire (ITQ) or PTSD Checklist for the DSM-5 (PCL-5)],45,46 depression Patient Health Questionnaire-9 items (PHQ-9),47 anxiety Generalised Anxiety Disorder-7 (GAD-7)48 questionnaire and social and occupational functioning [work and social adjustment scale (WSAS)]. 49 Sites A, C and D also included a measure of problematic alcohol use alcohol use disorders identification test (AUDIT). 50 Each Hub collected demographic and occupational information from prospective clients, and on occasions additional screening measures not covered at other Hubs. The use of specific screening measures across Hubs is summarised in Table 5, and the measures are described in detail in Routinely collected measures.

At Site A, online mental health questionnaires were presented as a tool for keyworkers to assess their own mental health and receive immediate feedback on their score, with the option to self-refer to the Hub after feedback. In contrast, completion of mental health screening within Sites C and D acted as a self-referral into the service, with immediate feedback provided via e-mail. Within Site B, individuals completed a brief self-referral form, and following acceptance of the referral by the Hub, clients were sent the mental health screening measures to complete prior to clinical assessment.

Assessment

Self-referral and mental health screening data were used to inform subsequent clinical assessment of keyworkers’ difficulties, to determine support and interventions offered by the Hubs.

At the time of data collection, Sites C and D offered an assessment phone call following the completion of screening questionnaires. Clients at Site C were offered either a rapid or a full assessment, dependent on the outcome of an initial review of screening scores. Within Site D, dependent on the initial presentation of the individual’s needs, the assessment was provided by any member of the clinical team, typically lasting between 30 and 60 minutes. The aim of this assessment was to gather information about the individual’s difficulties, construction of a collaborative formulation and quickly ascertain how the service could offer evidence-based support to manage those difficulties. Clients at Sites A and B were offered a full 60–90-minute formulation-based clinical assessment from a qualified clinician [a cognitive–behavioural therapy (CBT), therapist or CP] via video consultation as standard.

All Hubs offered an in-depth formulation-led assessment. Where Hubs offered shorter assessments as standard, in-depth assessments were offered dependent on clinical need. The aim of the in-depth assessment (60–90 minutes, with the potential to extend to a further session) was to develop a psychological formulation-led characterisation of individuals’ difficulties, and to inform support and treatment planning. For example, at Site D, a panel was used to decide whether a full assessment was clinically necessary or would expedite access into therapy. Site D also provided facilitated assessments into services where an agreement had been made, to prevent clients from re-telling their story. These assessments follow a pre-determined clinical framework developed by the Hub and are conducted by senior clinicians. If risk was a concern (e.g. a score of 2 or 3 on item nine of PHQ-9), duty clinicians at the Hubs ascertained the level of distress and facilitated support and/or access to appropriate services.

Onward referrals

There was variation across sites in the extent to which onward referrals were made following screening and assessment, versus direct provision of therapy within the Hubs. As one of the earlier sites to set up, the aim of Site D’s staff well-being work was to be proactive and preventative. As such, there was a focus on team and system support to reduce the number of clients requiring individual intervention. If deemed clinically necessary, onward facilitated referrals were typically made to maximise usage of already commissioned services, and fewer clients seen within the Hub for formal therapeutic work. Site D therefore facilitated a high number of outward referrals into other appropriate services, in addition to psychosocial support. When mainstream commissioned services were exhausted, or there was a clinical rational the Hub utilised its own therapy resource for those with complexity and clinical risk otherwise unmet. The site used an outreach and clinical advocacy model to ensure individuals received appropriate levels of support and offered evidence-based psychologically informed advice, self-help support and psychoeducation. Site C adopted these same principles and at the time of data collection facilitated approximately 60% of referrals to other services. Site B adopted the same principles, utilising regional mapping and waiting time check-ins to establish whether locally commissioned offers could be considered ‘timely’ to ascertain the most appropriate source of psychological support for that client, taking into account timeliness of available local support against risk of deterioration. As a result, Site B had an approximately even split between in-house intervention and onward referral.

By contrast, upon set-up of Site A, a scoping exercise of local services concluded that during the start of the COVID-19 pandemic, waiting times were significant; therefore, their intervention model was designed to provide in-house therapy provided directly by the Hub. Therefore, a small number of onward referrals were made to other services where appropriate.

Site B also provided a pastoral care pathway involving engagement with community, charity and third-sector organisations to offer an alternative model of care for interest-based support. A dedicated pathways advisor signposted keyworkers to appropriate support groups and activities, such as therapeutic singing, martial arts or music groups.

Interventions for individuals

Support for teams

A wide range of team-based interventions was offered across the Hubs, developed to support to the needs of managers, team leaders and help support the psychological safety of the health and social care system. Clinical leads, CPs or other psychological practitioners delivered team-based support across Hubs. Team-based interventions were bespoke to teams’ needs as determined through an initial consultation and formulation, and could be made up of a combination of different interventions.

While there were variations across sites, all sites provided consultation with managers and team leaders to help identify difficulties faced by teams and provide bespoke solutions. Team-based work incorporated trauma-informed approaches, signposting to supportive resources, reflective and resilience-based sessions, self-care workshops, training for teams and organisational strategy support, as well as direct support for managers as needed.

Workshops were provided virtually and face-to-face where appropriate across the Hubs. These were delivered by qualified clinicians with experience in supporting teams and organisations, and focused on different emotional well-being, self-care, psychological first aid and validation of the team’s difficulties. Examples included sessions on burnout, ‘containing the containers’, moral injury and psychological self-care. Furthermore, FPS was also offered by some Hubs to provide a safe reflection space for teams to come together to discuss relevant topics/difficulties.

Teams could engage with multiple aspects of team support provided by the Hubs, determined through initial and ongoing consultation with managers and teams. An example of this is outlined in Box 1.

Additional support for critical care staff

Critical care departments experienced high demand and were exposed to high numbers of patient deaths in the pandemic. Critical care staffs were therefore identified as a group likely to have been significantly impacted by the COVID-19 pandemic and all Hubs offered support by senior clinicians to this client group. Three of the Hubs utilised additional available funding to provide and co-ordinate support for this group, determined according to the number of ICU beds taken up with COVID-19 patients. The Hubs co-ordinated to provide broadly similar support across the services, differences in their methods of input are described below.

Within Site D, enhanced support was offered, including team and individual support, consultation, training and support for Professional Nurse Advocates to develop their skills, promotion of the general offer and more enhanced resources and ‘check-ins’ for Band 7 and Band 6 staff to help reflect and resolve difficulties. Each critical care unit had a senior clinician as an account manager and had the opportunity for onsite presence.

Site A similarly offered team well-being sessions for ICU staff as well as contributing to training consultations and supervision for key staff members. Site A also planned to assist these staff groups to enhance their well-being packages for example, team rooms and garden units.

Site B developed bespoke well-being resources for critical care staff and focused on promoting the offer to ICU staff. Drop-in sessions were also provided to some hospital sites to provide space to identify difficulties and access initial support with signposting.

Site B and Site D also joined a new critical care forum, which aimed to share learning and updates from supporting critical care staff to further improve the level of support.

Depression

Symptoms of depression were measured using the 9-item version of the PHQ-9. 47 This questionnaire asks participants to rate on a four-point Likert scale (0 = ‘not at all’ to 3 = ‘nearly every day’) the extent to which they have struggled with common symptoms of depression in the previous 2 weeks. PHQ-9 scores are added to compute a total score ranging from 0 to 27, with scores between 0 and 4 indicating ‘no depression’, 5–9 indicating ‘mild depression’, 10–14 ‘moderate depression’, 15–19 ‘moderate-to-severe depression’ and 20–27 ‘severe depression’. For the purposes of the present analyses, ‘caseness’ on the PHQ-9 was defined as scores suggestive of at least moderate depression.

Anxiety

Symptoms of anxiety were measured using the brief GAD-7,48 a self-administered tool that is frequently asking participants to rate on a four-point Likert scale (0 = ‘not at all’ to 3 = ‘nearly every day’) the extent to which they were bothered by common symptoms of anxiety in the previous 2 weeks. The total score can range between 0 and 21, with scores between 0 and 4 indicating ‘no anxiety’, 5–9 ‘mild anxiety’, 10–14 ‘moderate anxiety’ and scores ≥ 15 indicating ‘severe anxiety’. GAD-7 ‘caseness’ in the present analyses was defined as scores suggestive of at least moderate anxiety.

Post-traumatic stress symptoms

At two of the Hubs (Sites A and B), post-traumatic symptoms were assessed using the PTSD Checklist for the DSM-5 (PCL-5),46 a self-report questionnaire comprising 20 items corresponding to the DSM-5 symptom criteria for PTSD. Each item is rated on a five-point Likert scale (0 = ‘not at all’; 4 = ‘extremely’). A total PTSD severity score (ranging between 0 and 80) can be obtained by summing the items’ scores, and used to identify individuals with probable diagnosis of PTSD (i.e. individuals with severity scores ≥ 31). At the remaining Hubs (Sites C and D), post-traumatic symptoms were assessed using the ITQ. 45 The ITQ is a self-administered questionnaire assessing PTSD symptoms and additional features of complex PTSD (CPTSD) known as ‘disturbances in self-organisation’ (DSO) experienced by respondents in the previous month. It comprises 18 items rated on a five-point Likert scale (0 = not at all; 4 = extremely). The ITQ was scored to provide both a dimensional PTSD symptom score (i.e. a total post-traumatic stress score computed using six items assessing core PTSD symptoms including avoidance, hyperarousal and intrusions) and according to the standard ITQ diagnostic algorithm to identify probable International Classification of Disease – 11th Revision diagnosis of PTSD or CPTSD (via endorsement of specific PTSD and DSO items in conjunction with self-reported impact of symptoms on the person’s functioning). 45 For the purposes of the reported caseness analyses, any individual meeting ITQ criteria for PTSD on the PTSD subscale, irrespective of their CPTSD status, were regarded as having a probable diagnosis of PTSD.

Problematic alcohol use

Three Hubs (Sites A, C and D) collected data on harmful alcohol use using the AUDIT,50 a self-report questionnaire comprising 10 items rated on a five-point Likert scale (0–4). The AUDIT total score can range between 0 and 40, with scores between 0 and 7 indicating ‘low risk’, 8–15 indicating ‘hazardous drinking’, 16–19 ‘harmful drinking’ and scores ≥ 20 indicating ‘possible dependence’. In our analyses, AUDIT ‘caseness’ was defined as scores suggestive of at least hazardous drinking.

Social and occupational functioning

The WSAS49 is a brief measure comprising five items rated on a nine-point Likert scale (0 = ‘not at all’; 8 = ‘very severely’) to assess impact of mental health difficulties across multiple day-to-day tasks/domains. The total score can range between 0 and 40, with scores between 0 and 10 indicating functioning levels expected in non-clinical populations, between 11 and 20 indicating ‘significant’ impairment in functioning, and scores > 20 indicating ‘moderately severe or worse’ functional impairment. For the present analyses, WSAS ‘caseness’ was defined as scores suggestive of at least significant functional impairment.

The Hubs also collected data on a range of Hub clients’ characteristics relevant to the planned analyses, including the following.

Demographic data

Information was available on Hub clients’ age, gender, ethnicity, disability status (i.e. whether they self-reported having a disability) and sexual orientation.

Occupational and work environment characteristics

Although there was considerable variance across sites, the Hubs routinely collected information pertaining to Hub clients’ work setting, for example, whether the person worked in hospital settings (ICU/Critical care, Nightingale, A and E, Other Ward/Service, Across Hospital Site) or other setting (Primary care including GP Practices, Education, Emergency Services, Residential Care, Community Care, Local Authority, Voluntary/Charitable Sector, Other), and in what role.

Pre-pandemic mental health concerns

All Hubs collected information on whether the person was concerned about their emotional well-being/mental health before COVID-19.

Impacts of COVID-19

All Hubs collected information on common impacts of COVID during the acute phase of the pandemic. These covered whether the person had been impacted by COVID-19 in any of the following ways: (1) seconded to a different post; (2) moved to work in a different location; (3) undertaking new tasks within usual role; (4) been ill with confirmed COVID-19 (recovered at home); (5) been ill with confirmed COVID-19 (including being in hospital); (6) family member been ill with confirmed COVID-19 (recovered at home); (7) family member been ill with confirmed COVID-19 (included being in hospital); (8) experienced family/close friend bereavement; (9) suffered financial loss within the household.

Health status

The EQ-5D, a generic measure of self-reported health status, was completed as part of the SUQ survey. The EQ-5D is used globally and allows for the comparison of health status across different disease areas and population groups. 67 The 5-L version of EQ-5D used in the present study captures five domains of health status: mobility, self-care, usual activity, pain/distress and anxiety/depression. Each domain is rated on a five-point Likert scale (no problems, slight problems, moderate problems, severe problems and extreme problems). Also included is the EuroQol visual analogue scale (EQ-VAS), a visual analogue scale in which participants rate their health on a scale between 0 (worst health imaginable) and 100 (best health imaginable).

The data from the surveyed keyworkers were analysed according to recommended methods. 67 The level of missing data was summarised, including: (1) missing altogether (i.e. participants failed to return any part of the EQ-5D questionnaire) and (2) missing in part (i.e. participants partially completed the EQ-5D questionnaire). Participant responses across the domains and levels of the EQ-5D were reported (number and percentage reporting at each level, and summarised for no problems vs. some or more problems by level). EQ-5D values (also known as utility values), which are summary scores traditionally anchored between 0 (dead) and 1 (full health), were calculated using value sets which reflect preferences for health states. In the absence of a value set for the EQ-5D-5L, the crosswalk algorithm which was recommended by NICE until February 2022, was used. 68–70 NICE recently revised their recommendation to the mapping function developed by the Decision Support Unit (DSU) using the ‘The National Institute for Health and Care Research’s Policy Research Unit in Economic Methods of Evaluation of Health and Care Interventions (EEPRU) data set’ for future appraisals. 71,72 Subsequently, health status profiles were converted to EQ-5D values using the methods recommended by NICE at the time of the study initiation and data collection (i.e. the crosswalk algorithm). However, the results using the new DSU algorithm are also reported. EQ-5D values are reported using simple descriptive summary statistics. A preliminary linear regression was conducted to assess whether participant characteristics were associated with participants EQ-5D values. A significance level of 0.05 was used to interpret the results. Variables for inclusion in the regression were first identified using stepwise linear regression (forward and backward) and were supplemented with common characteristics associated with the EQ-5D and characteristics that were important in this population. Participant responses to the EQ-VAS were summarised using simple descriptive statistics (Ms, SDs, minimum, maximum and frequencies).

The EQ-5D values were compared to population norms to assess how the health status of our participants compared to the norms for a similar age group. Population norms were taken from other published data, though it should be noted that this source used the three-level version of the EQ-5D. 73 A targeted review was conducted to identify publications focusing on utility and/or health and social care use among health and social care staff (keyworkers). However, no publications were identified for comparison. A comparison was made to Health Survey for England (HSE) 2018 data, as the EQ-5D(-5L) and occupation variables were collected. 74 HSE 2018 interviewed 8178 adults, capturing demographic and health data. The data and methodology are described elsewhere. 75 HSE reported two occupation categories which are highly relevant to our study; health professionals (n = 311) and health and social care associate professionals (n = 72). Data from these participants were combined and used to provide a sample of EQ-5D data for comparison, which reflected health and social care workers (keyworkers) before the COVID-19 pandemic. The HSE interview weight was applied to account for selection and non-response biases. 75

Hub set-up costs

Hubs provided one-off costs associated with the set-up (or expansion) of Hub support for keyworkers. These are reported in Table 14. Site D, an existing service which expanded its provisions to include keyworkers, reported no one-off set-up costs as the service was already established with staff and equipment, and all staff and equipment were ongoing costs throughout the project. Therefore, the reported costs represent the set-up of new Hubs.

The average set-up cost was £111,713 (SD £53,471). It should be noted that set-up costs are challenging to collect retrospectively, and the teams noted some issues around this (e.g. a need to estimate costs). None of the Hubs reported any temporary staff set-up costs (e.g. staff hired on a temporary basis solely to set-up the Hub). Costs associated with staff recruitment were not considered, therefore potentially underestimating set-up costs. However, some Hubs reported that a percentage of their overheads would account for some of the associated recruitment cost. Across the remaining sites, the reported other costs were variable. All newly established Hubs reported set-up costs relating to IT infrastructure, equipment and staff training. Estates and overheads, reported for two sites, accounted for most cost and were also the most variable. These were noted by Hub teams to vary according to COVID-19 regulations (i.e. for a period one Hub needed to rent more space as per social distancing guidelines). One Hub did not report estate/overheads as they were virtual at the time of questionnaire completion. Training costs showed the least variability. Some ad hoc costs, such as printing, were not reported and estimates have been used where exact cost data are unavailable. The set-up costs are sunk costs (already spent and non-recoverable) and therefore any variability or uncertainty in the reported estimates will not impact the ongoing intervention cost estimates which follow in the next section.

Ongoing costs

The categories of ongoing resource use/costs related to providing Hub support and level of information reported by Hubs are reported in Table 15.

Staff costs were well reported across sites and accounted for the majority (86%) of ongoing costs. Management and administrative costs showed the least variation, while clinical staff and clinical leadership showed the most. Variation in clinical staff costs was expected due to capacity differences in the services provided directly by Hubs. Some variability was observed with how job roles were categorised. For instance, RAs were listed under ‘other’ by two sites and ‘clinical staff’ by one, contributing to greater differences between categories. However, due to variability in the level of detail provided (e.g. role description), it was not possible to standardise job roles across categories. While Site C reported two temporary staff members, employed for 6 months, these were excluded from ongoing costs as the funding for these roles was not continuous.

Other costs accounted for 14% of ongoing costs and were variable across Hubs. Estates/overheads accounted for the greatest share of non-staff costs. Promotion/advertisement (e.g. producing merchandise and promotional materials) and equipment costs demonstrated a large variation. However, this can largely be attributed to differences in how costs were reported. For instance, one site reported all ‘non-pay’ costs (including travel, supplies and consumables) as a single equipment estimate, as opposed to reporting these separately across categories.

Costs per keyworker reached

Across all four Hubs, the cost per keyworker given individual support ranged from £577 to £1481, with an average of £1011 (SD £409). The observed variability across Hubs was largely due to differences in the number of keyworkers supported, and given the unpredictable nature of the pandemic, it is likely that numbers of keyworkers requiring support will be challenging to predict. Additionally, some of the variation in cost per keyworker is very likely due to differences in service delivery (e.g. some Hubs provide complex interventions in house, while others facilitate more outward referrals for complex interventions). For instance, Hubs who make more outward referrals, as opposed to delivering support directly, would be capable of supporting (outwardly referring) more individuals, thereby reducing the cost per keyworker, but increasing the caseload and costs for other services. Producing an average cost per keyworker masks variability between keyworkers according to their need for services/support and engagement with the Hub. For instance, costs will be higher for those with greater mental health needs, or those with greater engagement, compared to those with lesser needs, or those who do not engage with support (e.g. those who are screened but do not access any support).

Team support estimates were based on data from three sites (A, C and D). The costs per keyworker accessing team support were less than for individual support, as would be expected, and ranged from £99 to £429, with an average of £272 (SD £165) across Hubs. Data on the number of whole team referrals were more variable (e.g. some based on actual figures, and some based on estimation etc.) and were therefore not used to produce a cost estimate as this may have impacted the accuracy of results. Furthermore, as noted earlier, the number of people who might benefit from team support is likely to be an underestimate. Therefore, these figures only provide an indication of costs.

Health status

EQ VAS scores

The mean VAS score was 66.41 (SD 19.22, 95% CI 64.11 to 69.71) and the median was 70. Figure 2 displays the frequency of VAS scores, which shows the most commonly selected VAS scores were 50 (n = 32/270), 70 (n = 20/270) and 75 (n = 19/270).

FIGURE 2.

EuroQol visual analogue scale scores by frequency.

Service use and associated costs

Logic model

To aid future research, a logic model was developed. The aim was to describe how the implementation of Hub support for keyworkers will lead to various outcomes of interest (including key economic outcomes). Given the design constraints of the current project, the logic model can help to explain potential outcomes that cannot be analysed due to data limitations (e.g. productivity losses) and provides a starting point for future research in terms of highlighting the range of outcomes available. This is particularly helpful for future economic evaluation, as it demonstrates the complexity between activities and outcomes, and delays related to time as economic outcomes often take longer to occur (e.g. as health status is likely to only affected only once mental health symptoms change).

The logic model (Figure 3) summarises the key resources needed to deliver Hub support, the activities performed as part of Hub support and subsequently, the potential short, medium, and long-term outcomes. The resource section describes what is needed for Hub delivery. These were captured in the Hub support costing section where possible, but this gives an overview for future Hubs drafting business cases. The resources are used for three key activities (outreach and promotion, referral/screening for individual support and team support). There are also interactions between these, for example, a keyworker identified through outreach may self-refer, while similarly, a keyworker reached by team support may self-refer for further individual support. The relationship between team and individual support is likely to be bidirectional (e.g. individuals may be less likely to require individual support if their needs are sufficiently addressed through team initiatives). These interactions/relationships are demonstrated by arrows on the model diagram. Individual and team support can take several forms, which are presented on the diagram. It should be noted that some Hubs have very specific offerings (e.g. on pastoral care or pharmacological intervention), but as these are unique to Hubs, they have not been included in the logic model which summarises typical Hub support only. Following activities, keyworkers reached by Hub support will experience related outcomes, which are split into immediate, short term, and medium/long term. As previously discussed, the evidence base is limited, and more research would help to define the time period in which these outcomes would be achieved. External factors which may influence any part of the logic model are included. For example, the impact of stigma, which may affect resources (i.e. as Hubs think about how to effectively use social media to target groups likely more impacted by stigma), activities (e.g. if fewer keyworkers self-refer for support) and outcomes.

FIGURE 3.

Logic model. a, The impact of Hub support is not only likely to affect the Hub user, but also friends/family, colleagues and the healthcare system.

Hub clients

Hub clients were invited by e-mail or telephone. Interviews explored in-depth the acceptability, enablers and barriers of the Hubs’ outreach approach; the acceptability of screening and self-referral processes; and reasons for engaging or disengaging. Interviews also explored (1) barriers and facilitators of support access and (2) whether Hub clients were considering or would have considered accessing other services in the absence of the Hub. This information helped to consider how keyworkers’ service use and health may have varied had the Hub support not been available, and inform the health economic analyses.

Keyworkers who did not engage with a Hub

Interviews explored whether participants had heard of their regional Hub, and if so, the reasons for choosing not to engage with the service. Participants were asked about whether they received any support from other services, and the types of difficulties for which they sought support. Participants were asked to complete a modified version of the SUQ used in Chapters 4 and 5, which avoided explicit reference to the Hubs’ involvement in the processes of accessing services as this is not relevant for this group. SUQ responses were used to elicit further qualitative information on how participants negotiated access to services, whether they met their needs, and whether they considered them to be helpful.

Hub staff and wider stakeholders

Interviews explored Hub clinicians’ perceptions and experiences of working at the Hubs, the feasibility and appropriateness of the model for supporting keyworkers, and how the Hub model was embedded into routine practice. Interviews explored clinicians’ perceptions of the enablers and barriers to outreach, supporting keyworkers, and helping them to access onward care via facilitated referrals and clinical advocacy. Contextual factors and implementation aspects were explored, including clinicians’ experiences of the set-up and structure of the Hubs, such as staffing and integration with other services.

Understanding the ways the Hubs differ from other types of staff support (differentiation)

Hub staff across sites clearly distinguished the Hub model from other mental health services across several aspects including proactive outreach; timely assessment and intervention; team-based work; and responsiveness and flexibility according to need. They saw the Hubs as different from routine staff support in terms of confidentiality and separation from keyworkers’ employing organisations (differentiation):

the Hub is something that’s separate from all the Trusts and separate from all the local authorities … it’s neutral ground and it’s a place where the clients can come and feel a little bit more reassured of confidentiality.

Hub staff 05

Staff from one Hub described the Hub approach as having fewer constraints on session numbers compared with other services, whereby clinicians could work more flexibly, providing more supportive calls and containment prior to formal therapy, enabling a more comprehensive, holistic approach.

Wider stakeholders saw the Hubs as having more specialist mental health staff with the ability to see more complex presentations than standard organisational staff support offers:

[the Hub] was providing access to more specialist, higher-end support, which wouldn’t normally be provided to a normal organisational health and well-being service.

Wider stakeholder 04

lots of Trusts didn’t have psychology for staff health and well-being in existence. While there was a pathway and an offer through to talking therapies, some of this stuff was so much more specialist than that.

Wider stakeholder 03

Stakeholders from NHS Trusts that already provided different staff support offers, including high-intensity psychological therapies, felt that the Hubs were not significantly different from their own offer:

there’s a fair amount of overlap, we do offer some of the similar interventions, such as EMDR. We don’t offer the team interventions in the same way the Resilience Hub do … But a lot of it … is whether someone would like to be seen in house, or whether that’s uncomfortable for them.

Wider stakeholder 07

Keyworkers also positively differentiated the Hubs from occupational health in terms of simplicity of self-referral without involvement of managers, and speed of access. Keyworkers who had turned to private therapy valued the Hubs being free of charge:

the process of actually getting into the service can put you off sometimes, but the Resilience Hub seemed a lot more straightforward, not lots of paper filling in and lots of line manager involvement.

Hub client 12

Other keyworkers did not want to access occupational health due to potential negative consequences within work:

it’s all supportive and then it turns into ‘well is it going to become a performance issue here? Is this person going to be able to manage?’ … all they’re focused on is statistics and getting through the cases … there’s no room … for allowing real support for a staff member that’s unwell.

Hub client 07

Hub staffs were concerned that keyworkers may not always understand how Hub support differs from other initiatives being advertised:

they might perceive it as duplication of role like ‘why would I need the [Hub] because we’ve already got a staff well-being service’.

Hub staff 13

This concern was borne out amongst some keyworkers:

There were so many things about where you could go for help, you could ring this number, you could use that app … I almost felt I couldn’t see the wood for the trees … because I wasn’t feeling well, there was almost too much choice.

Non-Hub keyworker 10

Where keyworkers did not differentiate Hub support from occupational health, some were mistrustful of the Hubs (relational integration) and concerned that their managers would find out they had accessed support. Hub staff described having to work to build up this trust:

Initially, it was quite difficult to engage with the staff because there was a mistrust. There was a feeling that, maybe, anything that was said in confidence would be carried back to management or HR. So, it took a long time, a lot of listening, a lot of patience and a lot of cooperation to gain the trust.

Hub staff 10

Individual and system understandings of the Hubs (individual and communal specification)

Hub staff felt that amongst Hub teams, there was a consistent, shared understanding of the model, and a shared sense of the purpose of the Hubs as supporting and containing individuals and teams directly providing or supporting access to timely intervention; promoting well-being; and supporting staff in a way that prevented the need for individual psychological intervention. Hub staff and wider stakeholders acknowledged challenges in getting this message out to keyworkers:

I would say there is no consistent understanding … I don’t think that we have effectively been able to reach everybody within the Integrated Care System, to, first of all, help them have an awareness of the service, and then specifically what the service is about.

Hub staff 02

I think [communication is] an ongoing issue … I think for desk-based staff not so much, because it’s easy to communicate by e-mail and communicate about the different options; when we’re trying to reach ward staff who don’t spend a lot of time at the computers it’s difficult to convey complex messages, so it’s definitely not clear to everyone what the different options are, and that’s definitely an ongoing challenge.

Wider stakeholder 07

Keyworkers who had not accessed mental health support or the Hubs before, and had not accessed the Hubs, at times did not know where to start when looking for support. Where keyworkers were aware of their local Hub, they did not always understand the full extent of what was on offer (individual specification):

I just thought it was more of a resources website where people can get a little bit of general information about how to avoid burn-out and how to keep well when you’re stressed. So I feel like I didn’t really know the potential that it has to offer.

Non-Hub keyworker 13

Others were not sure how to access the Hubs, or thought they were only for front-line clinical staff:

I didn’t really know how to access it and then my supervisor wasn’t sure either how to access that and it’s just a case of when you get so busy, things get pushed aside, and you think I’ll look at that later.

Non-Hub keyworker 09

Misunderstandings around eligibility may have been exacerbated by limitations of other staff support offers:

I did the online triage [for a non-Hub service], and then I got an e-mail back saying that … they were unable to help me at that time, ‘cause they were prioritising critical care workers … it made me feel like … other people obviously needed help and services more than I did, and I thought, oh well, you know, I’m alright, I’ll manage … So, I didn’t actually go anywhere else at that point.

Non-Hub keyworker 10

Creating value for the Hubs and support in general (internalisation)

Hub staff strongly agreed about the value of the Resilience Hubs and the support they provided (internalisation), for example, in providing a separate, confidential service:

staff who don’t want to seek support from their colleagues … want to seek support from an outside service that’s not linked at all to occupational health, not linked to the team they work within or their management structure, and that’s very much a role the Hub fills.

Hub staff 09

Several keyworkers spoke of valuing mental health, and feeling comfortable about seeking support:

there was an element of that self-acknowledgement that, yeah, I’m struggling here and whatever I’m doing isn’t helping me, so I need somebody else to help me.

Hub client 19

However, prioritisation of mental health support was not universal amongst keyworkers.

The role of managers in influencing engagement with Hubs (initiation and enrolment)

Within the Hubs, staff believed that the right people were involved in setting up and driving forward the services (initiation); key determinants were said to be extensive experience in mental health and trauma, seniority, contacts and positions of influence. However, Hub staff felt that senior managers in the care system were not always spreading the message of the Resilience Hubs to help promote the service to staff:

[organisations have] had challenges in terms of the senior managers, [maybe they] have been dealing with a crisis … perhaps they’ve had other priorities in the service, and if you don’t get in at a high level then it doesn’t happen.

Hub staff 13

Wider stakeholders felt that generally appropriate routes had been taken in promoting the Hubs at a senior level within organisations, although it was suggested that early involvement of Chief Medical Officers and Chief Nurses could have helped to establish further buy-in.

Some keyworkers agreed that managers and leaders within their organisations could be doing more to promote mental health and well-being support for staff, and encourage staff to access services including the Resilience Hubs. Some participants felt that management should be more strongly advocating support access, so that it became an expected or standard thing for staff to do:

it’s not coming from the powers that be that everybody has to … It needs to be said that, you know, everybody needs to access the support … I think everyone should, not be made to access it, but it needs to be really drilled down to people that, you know, this service is here for you.

Hub client 06

Hub staff described doing considerable groundwork such as establishing and maintaining new relationships with organisations and teams, and developing communication strategies to try to engage the right people within organisations. Keyworkers agreed that it was essential to make in-roads with senior management:

We try and find the right person in the system to engage with, who actually can make things happen … That’s often the workforce lead or the well-being lead or the staff well-being psychologist.

Hub staff 14

Hub staff and keyworkers both acknowledged that managers were under their own considerable pressures, and that this impacted the extent to which they promoted mental health and well-being services to their staff:

We talk about well-being and things. There isn’t an awful lot that is done about it though … And I do wonder whether, sometimes, managers are disinclined because it might reflect badly on them if their staff are needing to seek support. And then of course it’s going to affect their numbers [if I go off sick].

Non-Hub keyworker 08

The above challenges of understanding the Hub offer, how to access the Hubs, and barriers to prioritising mental health and well-being support all contributed to keyworkers’ decisions around whether or not to access support (enrolment). Managers also had an important role in encouraging keyworkers’ engagement with Hubs when they were involved and actively promoting the Hubs or mental health and well-being support in general (enrolment):

we so desperately want to engage these people and we so desperately want to support them and it almost feels that this manager level almost feels more key than the individual staff.

Hub staff 13

when I first asked for help in this job, it was as a direct result of my line manager saying … ‘I think it would be helpful for you to talk to Occupational Health and get some counselling’ … And she sort of pushed that quite hard, in a nice way, which was good.

Non-Hub keyworker 10

leaders in the organisation talking about their vulnerabilities I think has really helped more people come forward with stuff.

Wider stakeholder 03

Wider stakeholders echoed these views; one described how managers may not always proactively promote staff access to well-being support:

I think [managers promote well-being services] almost when it’s got to bubbling point rather than going to that proactive kind of thing … we should be looking after our staff slightly better than we are and … I’m not bashing the managers because they also need looking after as well.

Wider stakeholder 10

Hub staff and keyworkers found that managers sharing their own experiences of accessing support played an important role in personally promoting the Hub to others and creating safe spaces to talk about well-being (enrolment):

If managers are really well engaged in our service and then they manage somebody who’s struggling, they’re far more likely to actively encourage them to, you know, seek support.

Hub staff 13

if a senior leader said, oh, I accessed the Resilience Hub to get some support with A, B and C and they supported me in C, D and F, that would almost give people permission to say … ‘so it’s okay if he has done it, so it must be okay if I can access it as well’.

Non-Hub keyworker 20

The extent to which Hubs were seen as the ‘right’ way of supporting staff (legitimation)

Wider stakeholders generally agreed that the Hub model had utility and legitimacy as a service to support staff (legitimation). Particular aspects of the model were highlighted, including team-based and preventative work, and the Hub being part of an offer within the health and care system. It was felt that existing in-house support within organisations were vital, but that the Hub had an important external role:

[The Hub is] part of an offer and I don’t think it would be the right thing if it was the only offer … I think for us it’s good because it’s the end of a process, it’s the highest level of support and, therefore, it has to be accessed in a different way.

Wider stakeholder 02

One wider stakeholder from a VCSE organisation, however, felt that the Hubs should not have been hosted by NHS Trusts as they were not separated from the NHS system. They also felt that by virtue of being led by clinicians, the Hubs over-medicalised distress:

The Resilience Hub does pathologise because it’s a clinical system run by the NHS … by NHS clinicians … I think The Resilience Hub is far too clinical … it defaults into that person needs to have treatment with a clinical thing, whether that be CBT, whether that be counselling.

Wider stakeholder 06

Others, however, felt that the Hubs were less medicalised or diagnostic than other NHS mental health services.

However, Hub staff felt that at times they struggled to persuade management that the Hubs were legitimate, commissioned, services (legitimation). One Hub staff member felt that this would have been an easier task if organisations were encouraged or mandated at a system level to engage with the Resilience Hubs:

I think that some systems need to be told they have to engage with us. So we have got some areas that have not engaged [but] we know full well they’re not alright.

Hub staff 14

One wider stakeholder, however, felt that top-down national approaches to engagement would not work, and that the impetus to engage with the Hubs needed to come from the organisations themselves:

I think the evidence over a number of years, whether it’s been through strategic health authorities or whatever form the higher-level NHS took it, is that the NHS has at various points tried to push that kind of model onto organisations and make it happen. And actually, that’s never really worked. Because, if organisations are not willing to support it then it doesn’t happen. That’s why I attach so much importance to the information and the story telling, because I think that just starts to build that groundswell of bottom-up opinion that actually this is a really good thing and we need to support it.

Wider stakeholder 04

Sustaining engagement in the Hubs: flexibility and availability of services (activation)

Keyworkers who had accessed the Hubs found that Hub staff were flexible in terms of rearranging appointments. This flexibility helped some keyworkers to stay engaged in the process (activation):

If I’ve had to move anything around, there’s been a couple of times when I’ve been called into work and had to cancel last minute, but … [Hub clinician]’s been understanding and just rescheduled. So, it’s been quite noticeable that they’ve, sort of, worked around me and been really helpful as well.

Hub client 01

However, Hubs’ ability to offer out of hours or face-to-face appointments was limited, and having to fit in sessions while on site at work had caused one participant to stop attending. Other participants also described this as a barrier, or had had to disclose the support to their managers to arrange time off. Other keyworkers felt fortunate to be working from home due to COVID, as they would not otherwise have been able to engage with the Hub’s support.

Keyworkers described having a consistent clinician to speak to for each appointment as a strength of the Hubs. Where this had not been possible, inconsistency of Hub staff was described as a limitation. One participant had discontinued their support from the Hub in part due to speaking to multiple different Hub staff members, in favour of face-to-face support accessed through work.

Hub staff and wider stakeholders felt that the flexibility of the Hub model was an important strength that sustained the Hubs (activation), responding to the needs of keyworkers and the health and care system as the pandemic evolved.

[it] feels like there’s enough flex to move and adapt … So, the fact that it’s something that’s been allowed to evolve naturally, I think it’s doing its job quite nicely really.

Hub staff 01

I like that it’s flexible … I think it’s great that there’s a kind of responsiveness to what the need is and there’s ongoing change.

Wider stakeholder 07

Integration of Hubs within the wider health and social care system (interactional workability)

Wider stakeholders identified some early tensions between the Hubs and organisations whose staff they supported, in part due to overlapping offers (interactional workability):

probably we could have done better at the outset in terms of getting a common view and a common understanding and a sign-up to that across all partners to what the Hub would offer … for quite a while I thought there was a little bit of tension playing out between organisational employee health and well-being services and what they were offering staff and what the Hub was offering.

Wider stakeholder 04

Where organisations such as NHS Trusts did not particularly differentiate the Hubs’ support from their own in-house staff support because they already provided high-intensity psychological interventions, several wider stakeholders reported not needing the Hubs’ services to a significant extent, particularly where waiting lists were longer than in local services:

I think we have probably not used the Hub as much as some organisations because we have our own staff psychology service and Clinical Psychologists. So, whilst we have referred some of our most difficult cases where they really need significant psychological support, we have dealt with a lot of it ourselves.

Wider stakeholder 02

I think my only reservation … is that if somebody comes through to the Hub, and it’s for something quite low level … they’ve had to wait quite a while for that [at the Hub], but actually the Trust has a commissioner provider service with a local talking therapies service where we can get them in within a week.

Wider stakeholder 03

Hub clinicians acknowledged that variations in the support offered by organisations’ occupational health or employee assistance programme (EAPs) created challenges for the Hubs in terms of identifying support available and referring on Hub clients (interactional workability).

Wider stakeholders suggested that Hubs and organisations could work together more closely in terms of information sharing so that Hubs could identify where to prioritise their efforts:

I think the Hub needs to work out where it needs to spend its time most, I think it needs to see some information from organisations to make those prioritisation decisions.

Wider stakeholder 04

Building trust and confidence in the Hubs (relational integration)

Work was required across the system to build confidence and trust in the Resilience Hubs and the support they provided (relational integration). Subsequent to the tension described above, Hub staff described anxiety and even suspicion of the Hubs at a senior level within some organisations, particularly from existing staff support services. Staff across the Hubs described having to be careful to complement other staff well-being offers to avoid duplication particularly where there was some overlap between the Hub offer and other staff well-being initiatives:

there was a lot of anxiety to start with [within some organisations], we’ve, kind of, dispelled that and helped people realise that we’re here to help them and here to, you know, to fill in the gaps in provision rather than to nick the stuff that they’re doing … by us engaging with their staff and seeing their staff and getting good feedback from those staff groups, that will, hopefully, give them more confidence to be able to engage with us at a strategic level when the time is right.

Hub staff 05

Extensive work was done by Hub staff to ensure that those in the system had a clear, shared understanding of what support is offered by the Hubs (communal specification); they reported having conversations with key people (activation) to improve trust and confidence within organisations (relational integration):

we are continuously trying to ensure that we have a voice around certain tables, to tell people more about us, to help people have a clear understanding of the offer from the Resilience Hub, and about our transparency with regards to activity outcomes … I think if it wasn’t for that, things might have been a lot trickier with certain organisations and referral numbers coming from those organisations. What’s actually happening now is those organisations, those specific services are actually active referrers to our service who would signpost their own staff members to us, because there’s a clear understanding of our offer, where historically they might have held onto those people within their services, and those people might have been on extensive waiting lists.

Hub staff 02

At an individual level, Hub staff recognised that some health and social care keyworkers did not trust the Hubs (relational integration), perhaps due to the previously described confusion about whether the Hubs would share information with employers. Concerted efforts were required to help to build keyworkers’ trust in the Hubs and clarify these points.

Most keyworkers who had accessed the Hubs’ support felt they could trust the Hubs and, crucially, saw them as a safe space to discuss their difficulties without fear of repercussions or negative implications:

I think everyone in the NHS to some extent has got a fear of judgement, but once I got to know [Hub therapist] better … I quite quickly realised I’m not going to be judged here for what I’m saying. He’s there to help me … So I can say things that other people might be shocked at, where he’s not going to be, he’s going to understand that and try and help me deal with those things rather than saying well, perhaps you shouldn’t have done that. So I probably took one or two sessions to properly relax and I felt I could say anything.

Hub client 02

Many NHS staff who had accessed the Hubs felt strongly that Hub staff understood what they were going through, and valued this understanding:

they understood what it was like being on critical care … And then the secondary bullying issue that I had, they’re very familiar with dealing with … that as well … I immediately had confidence that they understood what I was talking about … I think that was the most important thing, that this wasn’t a generic service where you had to explain everything about why you were where you were. They’re like, right we know you’ve been working in critical care, we know what that’s like.

Hub client 13

By contrast, one emergency services keyworker preferred to access support from within their own organisation, as they felt that they would better understand the unique challenges of their role:

I don’t think your everyday person will understand what we go through … there’s domestics, there’s [road traffic collisions], there’s sudden deaths, hangings … unless you’ve done that and been to those incidents, I don’t think you can ever understand what that person’s going through … I think if I spoke to a GP about that … they don’t understand the stress of the job … Whereas I think more work-related, they get it, they’ve had that experience, they’ve spoken to number of other people who’ve probably experienced the same thing.

Non-Hub keyworker 18

Trusting that nothing would go on their occupational or GP records was vital for keyworkers; participants who had not accessed a Hub flagged this as a key concern:

if I came to you and said, I’ve got suicidal thoughts, for example, would that be on my NHS record? Would that appear on my GP records? … I know, as an organisation, we can access people’s medical records in certain circumstances.

Non-Hub keyworker 11

Participants’ concerns were not just about employers finding out; some keyworkers described damaging past experiences in their personal lives that affected their trust in services. One keyworker described their experiences of mental health records causing problems getting insurance, another told of their mental health records being revealed in court, impacting on custody for their child.

Skills and experience of Hub staff (skillset workability)

Hub staff felt that the Hubs had a good mix of clinical skills and experience (skillset workability). This was felt to be important for several reasons linked with this client group, including engaging and maintaining engagement of Hub clients, particularly given the reluctance around help-seeking already described (internalisation). Hub staff felt that senior clinicians with experience in managing complex mental health difficulties were important, owing to the range of difficulties with which keyworkers presented.

Furthermore, clinicians’ ability to effectively support staff teams using experience in organisational working was seen as an important part of the Hub’s support offer. Hub staff felt that recruitment of staff with a variety of clinical skills, combined with availability of additional training and continued professional development to support development of clinical skills for specific and emerging difficulties (e.g. Long COVID) had appropriately helped the Hubs manage these more complex support offers.

Likewise, keyworkers typically described Hub staff as knowledgeable and skilled in relevant areas, such as trauma and burnout:

this is what people are suffering from, we’ve all got trauma. So, it’s having … people focusing on that specialism for us, that burnout and then branching out in those little areas around that. It’s so needed … For us caring helping services … you’ve got to keep us going, because there’s never ever been real good welfare or help for us, generally.

Hub client 07

Hub resources and linking Hubs with other resources in the mental health system (contextual integration)

Wider stakeholders acknowledged that the Hubs were an effective way of making available the expertise of a limited number of specialist staff to a wide range of keyworkers and organisations (skillset workability/contextual integration):

what the Hub can offer is targeted support from trained specialists, who are thin on the ground. So, it makes sense to provide that across organisations rather than to do it within each organisation.

Wider stakeholder 04

In addition to making available their resource of specialist clinicians, the Hubs were also integrated within the wider mental health system, and supported clients to take up support provided by other services (contextual integration). Where clients were facilitated into other mental health services by the Hub, such as IAPT services, staff across the Hubs described making follow-up enquiries to ensure that these referrals were actioned by services, due to pressures and demands on IAPT:

it’s like a point of practice now … if I do a referral, I will be calling that service to check that they’ve got it, because otherwise it just falls through the gaps.

Hub staff 12

When referrals were accepted by other services, Hub staff expressed concern at the long waiting lists for therapy due to immense pressure within overwhelmed mental health services:

I think [services] are becoming increasingly overwhelmed … they often don’t offer any kind of priority for staff members … it’s probably a symptom of the demand on services throughout the country.

Hub staff 13

Keyworkers similarly described the value of the Hubs’ work following up referrals and ensuring they were taken on:

I contacted IAPT, … I got told you don’t qualify for help. Which is not easy to hear. I spoke to the mental health nurse at the Hub about it, really upset. And she intervened and wrote a report … So she amplified that … And while I was waiting for [IAPT treatment] she was really knowledgeable about [trauma] and she was sending me booklets and things … she sent me lots of information about it and exercises to do while I was waiting for therapy to start. So it was a lot more than just talking to someone.

Hub client 08

Regarding funding within the Hubs, Hub staff flagged concerns around the lack of certainty of funding for the Hubs going forward, both for staff recruitment and retention within the Hubs, and for keyworkers’ perceptions of the Hub:

I think there’s a fear amongst staff and amongst people who access the Hub around the longevity of it, so I think improvement might come with a bit of certainty around this model not just going as quickly as it arrived. I think that would help with people trusting in the service and it being here not for a short time.

Hub staff 09

One wider stakeholder flagged the enduring impact of trauma, and the time it may take keyworkers to come forward for support, being incongruent with the duration of Hub funding.

Means of evaluating the Hubs (systematisation)

Hub staff shared that within-Hub evaluation was predominantly through informal feedback, word of mouth and feedback questionnaires (systematisation). Engagement and Hub activity was monitored through key performance indicators, session activity, and engagement with the website and online screening tools by potential clients. Hub staff described very positive informal feedback from clients. Hub staff and wider stakeholders both described a challenge of more formally evaluating the Hubs’ work and producing evidence that justifies their funding, particularly regarding some of the higher-level goals of the Hubs, including reducing staff sickness absences:

I know we’re doing a good job in what we’re doing, but it’s how we capture that and prove that it’s worthwhile financially. So it’s really hard to prove that we’re stopping staff going off sick because it’s a really hard thing to measure.

Hub staff 11

Wider stakeholders universally felt that there were insufficient data being fed back from Hubs to organisations and the health and social care system. Some implied that this exacerbated some of the initial tensions between the Hubs and other organisations. Access, activity and outcome data, it was felt, would help to reassure organisations of the value of the Hubs:

I think we [HRDs] need to see some information about what’s been going on, to provide that reassurance about the levels of activity and how they’ve changed. I know it’s very personal but sometimes people consent to give their own story about how that’s helped. I think it’s useful to see all of that information because it just provides a bit of evidence.

Wider stakeholder 04

Some stakeholders from NHS Trusts were also keen to identify patterns of staff accessing the Hubs so that they could do more in-house to support staff:

I think, for me, in occupational health, we’ve got massive gaps around we don’t know who’s accessing it, so I don’t know whether we should be putting additional support in … because if we’re not getting any feedback, then we can’t be proactive.

Wider stakeholder 09

Individual, team and system appraisals of the Hubs (communal and individual appraisal)

Staff at each of the Hubs described coming together frequently as a team to discuss and feedback on Hub support and processes (communal appraisal). Hub staff across sites valued the ability to reflect on and fine-tune the ways in which the Hubs worked to support staff:

the beauty of the Hub and having a group of clinicians who are all enthusiastic and really invested in wanting to help staff, dilemmas like that, it’s helpful to have all of our heads together.

Hub staff 03

Some keyworkers described sharing positive experiences of the Hubs with colleagues, or deciding to access their local Hub after they had heard positive feedback from others:

I think it’s been a really beneficial service. I know a few other people that have engaged with it, both short-term and a bit more on a long-term basis, so, I think it’s been a really valuable resource, especially to people within the NHS that have been, and still are, going through a pretty rough time in a lot of instances.

Hub client 01

From a systems perspective, however, some wider stakeholders were concerned about capacity issues at the Hubs as waiting lists for formal interventions grew:

I’m significantly concerned about the waiting times, which seem long for the amount of money that’s been invested.

Wider stakeholder 02

I’ve had conversations about … support for staff members with complex needs; … when the Resilience Hub was first set up it had capacity to support staff members in that kind of situation, and not surprisingly that’s changed over time; so I think that does leave unmet need in that area.

Wider stakeholder 07

Nevertheless, from the feedback they had available from Hub clients, wider stakeholders valued the positive impact of the services. Team-based interventions were particularly valued, despite some initial disappointment that not all Hubs were able to provide on-site support due to infection control:

I think the outreach has worked really well. Once they’ve come on-site they’ve been really engaging. The staff have been quite receptive.

Wider stakeholder 09

Hub staff viewed the Hubs as generally very effective (individual appraisal) for those who have accessed them, but acknowledged that promoting the Hubs widely and persuading staff to engage had been two of the biggest challenges:

I think it’s been as effective as it can be as a new service that has hit the ground running. I feel like probably we’re in our prime now, I think we’re probably most effective now, because the word is out there, and people are understanding what we’re for and are accessing us and realising that we are there. I think it took time for that message (a) to get out there and (b) to be received.

Hub staff 01

Keyworkers who had accessed the Hub services typically described very positive experiences:

I got more out of it than I expected, because I expected just to fix the short term. I didn’t expect it to give me some long-term strategies to help prevent it happening again … I feel it’s an invaluable service. I do hope it continues. I think it’s a hugely neglected area of the NHS and of workers in the NHS, because the sort of stresses we’ve all suffered in the last two years, they’ve always been there to a certain extent.

Hub client 02

Sometimes it got me through the week knowing that that phone call was coming. It got me out of bed … At the end of every session I felt enlightened. I might have had a good cry all the way through. I might feel purged of all the anxiety that I’d been clogging up all week. But I did feel inspired at the end of it. I felt like there was something I could do. I had … not just tasks to do, but I had ideas that would help me stay well or stay less ill.

Hub client 08

Once exception to this positive feedback was around the previously described limitations around EDI:

The gaps are there … I asked for a culturally … somebody who’s on my background, you didn’t have that person, it’s not available … Your staff representation is not there of what’s on the outside, the diversity. Your wording of the assessments you use is not open.

Hub client 07

Adapting and developing the Hub model (reconfiguration)

Staff across the Hubs and wider stakeholders agreed that the flexibility of their service models, and adapting these to the changing needs of keyworkers and the system as a whole were strongly valued aspects of the services (reconfiguration):

I think we always try and do things differently. So, the beauty of the hub is that we’re always reassessing and re-evaluating and reviewing what we do.

Hub staff 14

Key adaptations that had been made to the Hub models (reconfiguration) in response to arising needs included offering team-based interventions, and the provision of individual therapy within one of the Hubs to meet increased demand alongside extensive waiting lists in local services:

[we started] looking at the impact of COVID on the teams and how to build across the team, resilience, and well-being. So, that approach changed as well, flexing with the need because initially I think we just thought we were going to do one-to-one work and then it became apparent that whole teams were struggling.

Hub staff 01

Impact of Hub models

The cross-case analysis display in Table 23 demonstrates that despite differences to the service model across Hubs, the demographic and occupational groups accessing services were broadly consistent. The proportion of Hub clients allocated to the ‘other’ category at Site D was likely due to lack of sufficient information to categorise into more specific groups, therefore does not represent a true difference (see Chapter 8, Limitations). Differences in the types of interventions offered appear to have impacted on the proportion of keyworkers accessing support directly from the Hubs, as would be expected. Site A, for example, delivered a higher percentage of services directly, while Site D’s model was more based on facilitating client access to existing services, and therefore had a smaller percentage of keyworkers accessing services delivered directly by the Hub (24.6% at Site D compared with 81% at Site A) (see Chapter 3, Onward referrals).

Utilisation of Hubs as a resource for keyworkers with more severe or complex mental health difficulties

Each work package demonstrated that keyworkers accessing the Hubs tended to have severe and/or complex difficulties for which they needed support. Mental health screening data (see Chapter 4) suggested that most keyworkers accessing the Hubs had high mental health support needs, for example, 60% met criteria for clinically significant difficulties on at least three different screening measures, and approximately 80% had scores suggestive of clinically significant impairments in functioning. As noted in Chapter 4 (see Discussion), differences were seen in PTSD caseness between sites that used the ITQ compared with the PCL-5. Keyworkers reported lower health status in comparison with the general population and a pre-pandemic health and social care professionals’ sample (see Chapter 5, EQ VAS scores) suggesting unmet need and/or ongoing impact of the pandemic. Health status data also indicated that the most affected domain related to anxiety and depression, for which only a minority of participants reported experiencing no problems. Complexity of mental health presentations, in particular pre-pandemic mental health concerns and financial difficulties, predicted higher levels of distress across most measures, and across all Hub sites. Qualitative data may in part explain these findings. Wider stakeholders from NHS Trusts, including HR and occupational health leads, tended to view the Hubs as a means of providing specialist support to a small proportion of staff who had more severe or complex difficulties, rather than as an additional, parallel, pathway for lower-level support. It was suggested that NHS occupational health departments typically do not provide this level of specialist support, and that the Hubs meet this need. This may in part explain why keyworkers with these presentations tended to access the Hubs. The severity of presenting difficulties of keyworkers may also be explained by the qualitative data demonstrating that keyworkers typically waited until things were very difficult before seeking support (see Barriers to valuing support: keyworkers’ beliefs about self and job roles). The higher complexity of needs of keyworkers registering with the Hubs provide evidence for the Hub staffing model, whereby the Hubs were staffed by predominantly senior clinicians with the experience to work with this client group. This finding aligns with the cost of mental health services varying according to need, as is also the case in IAPT services (see Chapter 5, Discussion).

Barriers and enablers

Barriers and enablers to successful uptake and implementation of the Hubs were consistent across sites. Themes related to clarity of the Hub offer, and beliefs around accessing support. Valuing and taking up Hub offers were contingent on keyworkers understanding what was on offer and how to access it, as well as differentiating it from other offers (see Understanding the ways the Hubs differ from other types of staff support). Keyworkers who valued mental health support, and felt their managers valued and promoted well-being support, were more likely to access it. Keyworkers were inhibited from seeking support by beliefs about themselves and their role, which was exacerbated by negative workplace culture and perceived lack of management support.

Satisfaction with the Hubs

Across the sites, satisfaction with the Hubs reported on the SUQ demonstrated high median satisfaction scores, and a substantial proportion of clients reporting that the Hubs had met their needs (see Satisfaction with Hub support and onward referrals). Only a small proportion (4.4%) reported that the Hubs had not met their needs. These findings are supported by interview findings demonstrating that generally participants had very positive feedback regarding the support they had received from the Hubs (see Individual, team, and system appraisals of the Hubs).

Subgroup-level cross-case analyses

The following sections present cross-case analyses conducted at the level of keyworker subgroups found to have low uptake amongst those accessing Hub support. Table 24 presents demographic groups (keyworkers identifying as: from an ethnic minority community; men), and Table 25 presents occupational groups (care home and emergency services staff). Emergency services staff are included due to low uptake; however, this is also partly explained by not all Hubs being open to all emergency services staff, this exclusion is in line with NHSE guidance (see Target population and Chapter 4, Discussion). Furthermore, while the following sections atomise these keyworker characteristics, as demonstrated by some of the quotes in the tables below, intersectionality may well compound barriers to access for those who identify across multiple under-represented groups.

Under-represented demographic groups

Under-represented occupational groups

Continuation of mental health screening, outreach and provision of direct/indirect support

The current study highlighted the potential value of the Hub model of outreach, screening, support navigation and provision of direct support, as well as ability of the Hubs to work flexibly in terms of responding to clinical need within staff groups. The model has been shown to be sufficiently adaptable to different contexts, including a new target population, and across multiple regions, and allowed targeting of Hub resources and monitoring of service access. The delivery of high-intensity interventions was particularly supported by study findings demonstrating the heightened mental health needs of Hub clients, and associated levels of complexity, necessitating the Hubs’ staffing of senior clinicians with the experience to work with individuals with these difficulties, as well as providing team-based support. The satisfaction levels of Hub clients were high across both quantitative and qualitative data, but consideration of selection bias needs to be taken into account. Although the Hubs’ team-based work was not directly evaluated by this study, our interviews demonstrated the value placed on these interventions by the wider system. Team-based interventions and prevention of mental health difficulties, rather than a solely reactive approach, were seen as unique and valuable additions to staff support within the health and care system. Wider stakeholders acknowledged that Hubs may not be able to provide that level of support or training for every team/organisation, but that prioritisation decisions could be made if there were greater information flow between the two. This recommendation is also supported by other research and NICE guidance for mental well-being at work. 130,131 Further evaluation will be required to further understand the potential clinical benefit of the Hubs (see Research and evaluation recommendations).

Promotion of Hubs

Hubs involved in the study had variable methods of outreach and promotion the research did not identify whether any particular methods of promotion used by the Hubs were more or less effective (see Chapter 7). Nevertheless, methods of promotion valued by keyworkers and wider stakeholders included testimonials and feedback from those who have been supported by the Hubs, and presentation of research evidence of the utility of the Hubs’ support. A personal approach was particularly valued, including face-to-face contact, team presentations and visibility of Hub staff within teams. Clarity of what is offered by the Hubs, and how to access support would be beneficial. Promotion should cover key areas of uncertainty/concern identified in our work, such as eligibility of staff groups (e.g. non-clinical staff) and not needing to be at crisis point to access support. Confidentiality of Hub support, specifically around self-referral and the Hubs’ use of standalone clinical systems (i.e. with no reporting back to occupational health, but nevertheless within the parameters of usual NHS systems, e.g. response to risk) was strongly valued by keyworkers, and could be emphasised within promotional materials.

Access in relation to equality, diversity and inclusion

The study was hampered by the same issues of recruitment of keyworkers under-represented groups as experienced by clinical teams. As such, recommendations are limited, however it is clear that simply providing information to particular staff groups is insufficient. It is important to understand and resolve barriers to access, and to understand the specific needs and reservations of different staff groups, which may in turn lead to more meaningful engagement with under-represented groups. Improvements are suggested to facilitate greater access by certain groups. Earlier and more extensive engagement with stakeholders identifying as men and from minority ethnic communities is recommended to design appropriate support offers that meet the specific needs of these groups. Representation of these communities on Hub clinical teams was recommended, as well as staff training, experience and understanding of how to support keyworkers with the impact of racism and discrimination, and promotion of how the Hubs can meet these needs. Ability to provide out of hours appointments was suggested to better support shift workers. Options to complete fewer questionnaire measures, or complete measures by telephone were valued to increase access for those in severe distress and/or less able to access digital technology.

Collection of data on Hub access, mental health screening, Hub activities and outcomes

This study demonstrated the value of the use of a standard set of screening questionnaires for understanding and triaging individuals who self-referred to the Hubs. The primary parameters of the screening offer of the four Hubs included in this evaluation (e.g. mental health domains assessed) are broadly consistent with those employed by the Greater Manchester Resilience Hub following the Manchester Arena incident, and therefore demonstrate sufficient adaptability to different contexts, populations and regions. It is recommended that screening initiatives continue at the study sites and is extended to other Hubs across England. It is also recommended that Hubs explore the feasibility of implementing standardised means of routinely gathering follow-up or outcome data following clients’ use of the Hubs. The value of routinely collected screening and outcome data should be balanced against the recommendations made in Resilience Hub recommendations to facilitate ease of access to the Hubs, such as options to omit questionnaires or complete them via telephone. In addition to their clinical value, the collection of standardised screening and outcome data has significant potential for further service evaluation/research and to address concerns related to sharing data with other organisations (see Data sharing and reporting). We elaborate further on how these measures could be used for research and evaluation purposes in Research and evaluation recommendations.

While it was acknowledged that granular data sharing that may identify staff members would be inappropriate, wider stakeholders reported that their ability to evaluate the Hubs’ effectiveness was impeded by insufficient high-level data shared across organisations on access and clinical activities provided by the Hubs. Some implied that this exacerbated tensions between Hubs and other organisations. Both wider stakeholders and Hub staff acknowledged the difficulties in evaluating the impact of the Hubs on higher-level concerns, including staff sickness and staff retention. That said, this is a wider issue affecting many NHS mental health services; services such as IAPT and EIP services are the exception in publishing activity and outcomes data. The Hubs were not set up to provide specific outcome or data metrics (unlike IAPT services) and have not yet had the time to develop appropriate metrics (unlike EIP services). The need for data collection and sharing should be balanced against the potential burden on clients, the importance of collecting information that is clinically valuable for the support of Hub clients, and the ‘ethos’ of the Hubs (e.g. protection of keyworkers’ confidentiality in light of specific barriers to access affecting these occupational groups). In addition, the Hubs evaluated in the study flagged the extensive Hub systemic and team-based activities which, while recommended in Hub guidance,33 were not currently systematically reported to NHSE. These data may usefully provide further evidence of some of the core activities of the Hubs.

Management of Hubs’ specialist resources

Interviews with wider stakeholders demonstrated the variability across organisations in terms of their own in-house staff well-being offer. There is a potential for greater information exchange with organisations so that Hubs can prioritise those that have less staff support on offer. For example, some NHS Trusts already had psychology provision separate to occupational health that could for example, offer specialist interventions (e.g. EMDR) or see more complex presentations, whereas many did not. Detailed mapping of services within regions may be appropriate to better allocate Hub resources. This should, however, be balanced against the need for organisations to have an adequate staff support provision, with Hubs providing an external offer to allow clients choice of access.

Ensuring preparedness of the Hubs for responding to future crises

Participants from all groups interviewed felt strongly that the pandemic had shone a light on staff well-being, as well as highlighting and exacerbating pre-pandemic issues in this area. Evidence suggests that keyworkers may take years before seeking support, if at all (Chapter 6, Qualitative Analysis), and our research corroborates a need for ongoing mental health and well-being support for health and care staff, which could be fulfilled, at least in part, by Hub support. Hub staff and wider stakeholders also saw a valuable potential role of the Hubs in trauma-informed system responses, and responses to traumatic incidents and emergencies. The literature also supports these recommendations in terms of the positive impact of staff health and well-being on patient outcomes. 132–134 Data collected as part of this study clearly suggest that Hubs that were already operational at early stages of the pandemic managed to reach and support a larger number of affected individuals. Having an existing Hub infrastructure in readiness for future incidents is therefore a vital part of the country’s emergency preparedness and response planning.

Creation of psychologically safe environments

Some study participants felt strongly that their employing organisations were relying on individuals to seek support or ‘cope better’, rather than addressing systemic and workplace issues that keyworkers felt caused their difficulties. These findings demonstrate the importance of the psychological health and safety agenda in helping to prevent the development of staff mental health difficulties within health and social care organisations. Organisations should attend to aspects of job roles which may be psychologically damaging. Hubs have a role to play in helping organisations to do this (e.g. team-based recommendations; see Management of Hubs’ specialist resources); however, ultimately the responsibility lies with organisations themselves. Consistent with recent NICE guidance130 for supporting mental well-being in the workplace, individual support offers should be offered as an adjunct to, not a replacement for, resolutions to systemic challenges.

Actively addressing structural inequalities and workplace discrimination/racism would contribute to the creation of psychologically safe environments for staff from minority ethnic communities. Participants and PPIE consultants from these communities described often feeling unheard. These stakeholders had often been asked to do additional EDI work due to their ethnicity, but that this was typically unpaid. Many were passionate about this work, but felt it went unrewarded, particularly due to lack of change seen. Staff who worked in supportive roles (e.g. working with Trust BAME Networks) felt the burden of holding others’ distress and experiences of racism. Greater support offered to keyworkers from minority ethnic groups may be valuable within organisations.

Genuine promotion of mental health and well-being support

Health and social care organisations should share the responsibility of promoting the Hubs and other well-being or mental health support offers. The study demonstrated that involvement of managers and leaders was central to encouraging keyworkers to access support from the Hubs. The influence of workplace culture was also found to be important in encouraging staff to take up support offers. Well-being conversations between managers and staff to genuinely promote help-seeking were valued, particularly where managers promote and discuss support offers with teams and individual staff members, rather than simply forwarding e-mails or relying on organisation-level communications.

Data sharing and reporting

Further to the above Hub recommendation to share more data where possible, wider stakeholders also acknowledged that greater collaboration from both sides of the table would be beneficial. Hub leads suggested that greater clarity of where best to share Hub data, and who with, would strengthen this work, particularly within the context of changing Integrated Care Board structures. Organisations should share information with Hubs about their own support offers for staff, so that Hubs are facilitated to make prioritisation decisions about how to allocate time and resources. Information sharing across organisation should also play a role in evidencing the Hubs’ impact on staff sickness and retention; the research demonstrated that this is a key outcome of importance to Hubs and wider stakeholders that was not currently captured at the time of the research.