Summary of key findings

Our aim in this review was to improve our understanding of how, why and in what contexts nurses, midwives and paramedics experience work-related psychological ill-health; and determine which high-quality interventions can be implemented to minimise psychological ill-health in nurses, midwives and paramedics. Through our analysis (see Chapters 4, 5 and 6) and discussions with our stakeholder group we realised there were some fundamental questions our work needed to answer. We therefore generated the following questions, which we have sought to address in Chapter 6, through our realist synthesis and our 14 tensions:

• Why is psychological ill-health in healthcare professionals still a huge and growing problem which has become entrenched in some settings?
• Why despite having interventions (some of which have an ‘evidence base’), does the problem persist?
• How can we optimise existing interventions, by analysing when and where they work suboptimally, as well as innovating and building upon what already exists?

Our overall review findings are summarised in Box 1.

BOX 1

Summary of key overall findings

We now summarise our main findings in relation to the three specific aims of our study.

Aim 1: understand when and why nurses, midwives and paramedics develop psychological ill-health at work, and provide examples of where and how it is most experienced

Our findings suggest that staff come into health care with high ideals, strong values and the desire to do a good job every day, yet many develop psychological ill-health as a result of their work. Our study suggests this is highly prevalent, and should be anticipated and prepared for, given the emotional, social and ethical aspects of the work. High degrees of empathy can also cause vicarious or secondary trauma. In short, delivering excellent care to patients can often come at a high price for staff in terms of their own psychological ill-health.

Our synthesis reveals that psychological ill-health in nurses, midwives and paramedics results from complex interactions between the individual, their professional role and values, the desire to deliver high-quality care and current working structures and conditions. These complex interactions – and resulting risk – are further exacerbated by intersectionality factors, such as gender, ethnicity, disability. It has always been challenging to provide prompt, high-quality, empathic care for patients, particularly when there are staff shortages, and this has been exacerbated by the COVID-19 pandemic. It is now even more challenging to provide excellent care and to ensure that the psychological health or nurses, midwives and paramedics is maintained and does not deteriorate further. Staff have gradually adapted their work behaviours and norms as conditions have got progressively worse. For example, gradually increasing overtime or work intensity can erode some protective mechanisms (such as job satisfaction and engagement and time with colleagues and family), resulting in harm to psychological health. This gradual worsening of working conditions has been compared to the ‘boiling a frog’ analogy (see Chapter 6; Tension 2) and can cause significant work dissatisfaction. Such work conditions Herzberg²³⁴ calls hygiene factors (see Figure 2). These are contextual extrinsic factors, including those deemed essential in Maslow’s hierarchy of needs, which we identified in the literature as causes of psychological ill-health.

FIGURE 2

Herzberg’s two factor theory diagram (adaptation based on our data).

Our review has highlighted (see Box 1 for summary):

1. that trauma is not only acute, but can be chronic and cumulative, with seemingly benign events triggering psychological ill-health;
2. that collective blame is often attributed to individual staff and that there are double standards in accountability; and
3. that fitness-to-practice processes can be psychologically harmful and when staff do speak-up they can encounter a ‘deaf effect’ with no action.

Nurses, midwives and paramedics are often exhorted to ‘put patients first’ within a culture of giving 100%, which can send a message that their own psychological and physical needs come second, thus the needs of the system override staff wellness. There is also a professional culture that promotes a ‘serve and sacrifice’ ethos, which in the context of staff shortages can result in MD. Furthermore, excessive requirements for emotional labour can be at odds with messaging that well-being matters and exhortations to improve workplace psychological ill-health. This can cause some staff to feel they are failing at maintaining their own psychological health because they are not resilient enough, blaming themselves and individuals rather than work conditions and systemic forces identified above.

Secondary interventions that target individual nurses, midwives and paramedics to modify their response to stressors (and thereby prevent their psychological health deteriorating further) may address aspects of the causes identified in our review. However, our review has highlighted that these are fragmented (not part of a wider ‘well-being’ strategy that also focusses on primary prevention) and typically ignore the wider context. In doing so, offering such interventions can unintentionally backfire by sending a message that blames staff for their own psychological ill-health.

In terms of profession-specific issues, we identified more similarities than differences between nurses, midwives and paramedics. Most profession-specific causes we identified in the literature are likely to be applicable to other subspecialties (see Chapter 4, Table 6). The staff groups within our three professions that were most at risk included groups that were subject to discrimination at societal, structural or interpersonal levels, including ethnicities except White British, LGBTQ⁺, disabled staff and women. Structural features of work and working conditions may place certain staff at greater risk of psychological ill-health, including those with increased exposure to trauma, blurred role boundaries and geographic isolation. These features were mentioned for specific types of nurses (e.g. critical care; end of life; mental health inpatient setting; ENPs; district and community nurses; prison nurses) and are common to some midwives and paramedics. Leaders and managers were also identified as a group that may be particularly lacking support. In terms of stages of professional life where risk might be increased, this included staff in transition (e.g. newly qualified or new in post) or crises points (e.g. after trauma exposure or subject to investigation or complaints).

Aim 2: identify which strategies/interventions to reduce psychological ill-health work best for nurses, midwives and paramedics, find out how they work and in what circumstances these are most helpful

Our second aim focused on the strategies and interventions, which may operate differently in different contexts and for different staff groups. Our review found that the complex interactions that lead to psychological ill-health mean a reductionist ‘individual intervention’ approach would be inappropriate, and that identifying the ‘right’ intervention for the ‘right’ circumstances at the ‘right’ time is challenging.

Overall, our literature synthesis noted that individual-level interventions were unlikely to be enough to support staff due to systemic problems, and in the absence of a wider contextual lens could do more harm than good (potentially blaming staff for their own poor psychological health instead of intervening at the system level). However, individually focused interventions aimed at modifying response to stressors, such as mindfulness, are likely to be useful to some staff in the moment, and there is good evidence for some interventions aimed at wider organisational culture change such as Schwartz Rounds⁴⁸ and the Blue Light Programme.¹⁷⁸ This key finding concurs with previous research arguing for systems approaches to psychological health at work.^65,235

Our review recommends avoiding implementing interventions into organisations without first understanding the service architecture, culture and work conditions that would impede or facilitate implementation (e.g. lone and/or community working or working night shifts affects access; trust, psychological safety, and compassionate leadership affects speaking up and disclosure of emotional support needs). We noted tensions in organisational priorities between

1. quick fixes and longer-term interventions, which may require patience and sustained support to produce their effects, and require non-quantifiable ways of showing benefit (e.g. Schwartz Rounds);
2. interventions focused on organisational learning versus staff healing; and
3. the need to act and offer support versus:
   1. providing interventions that are ineffective because they are too soon, reactive and/or provided at a single time point;
   2. protective strategies appearing as ‘lip service’ and managers perceived as ‘out of touch’ for recommending approaches when staff are not given time to access and participate in interventions.

Our synthesis suggests there are no easy ‘plug and play’ interventions that would result in significant change. The (interrelated) root causes of psychological ill-health that we have identified in this review are where interventions would likely have most benefit – with a system rather than individual cause lens. Building on organisational psychology approaches²³⁶ organisational interventions are most likely to be effective if there is first a focus on minimising demands and development resources – at individual, group, leader and organisational levels. Our research suggests this includes addressing staffing shortages, and hygiene factors to reduce job dissatisfaction, and changing culture to one that encourages and supports speaking up and listening, recognises the inevitable challenges of healthcare work and seeks to recognise staff psychological ill-health as the norm.

CUP-2 programme theory

The 14 tensions highlight that healthcare delivery is a complex and dynamic balancing act. It is challenging to provide prompt, high-quality, empathic care for patients, especially in a context of staffing shortages, in a way that also concurrently maintains and even improves the health of the workforce. When we consider these tensions in relation to the five key findings, organisational ‘balancing acts’ came to the fore (see Figure 3); there are things that we need ‘more’ or ‘less’ of within our healthcare organisations and systems, based on the literature synthesised. It is worth noting that even too much of a ‘good’ intervention or strategy may be problematic (especially if not balanced with other elements). We deliberately focus the priorities for change at the organisational or systems level, rather than individual, given our learning from this review.

FIGURE 3

Imbalances based on the tensions embedded within each of our five key findings.

Figure 4 is based on our key findings (see above and Box 1), which has highlighted that a greater emphasis on self-care is a shared responsibility between individual staff, teams, managers, organisation, governing bodies, while being careful that this does not erode the high standards of patient care. The emphasis on patient care needs to be matched with an emphasis on staff psychological well-being; professional accountability needs to be matched by listening and responding to staff, with transparency of how staff input has translated into tangible changes and results. The emphasis on reacting and responding to events now needs to be balanced by more emphasis on prevention and the cumulative build-up of smaller stressors over a longer time period. And, perhaps most importantly, the predominant individual-focus of interventions, which can be perceived to place blame on the individual, must be balanced by interventions, which focus on organisational and system-wide change.

FIGURE 4

Key focus areas to restore the balance. Each element on one side of the scale is in tension and needs to be balanced against the one on the opposite side.

Aim 3: design and develop resources for NHS managers/leaders so that they can understand how work affects the psychological health of nurses, midwives and paramedics; and what they can do to improve their psychological health in the workplace

The design and development of resources for NHS managers and leaders is in progress, due for completion and delivery to NIHR in Spring 2023. This section describes our progress to date towards meeting this aim.

The evidence and analyses presented in this review has been translated through an iterative process with stakeholder group, advisory group and policy-makers (as outlined in our protocol, see also Chapter 2), to produce eight overarching recommendations (see Box 2 below). These are targeted at Wellbeing Guardians and executive boards, those responsible for leading teams and/or those refining/designing interventional strategies to tackle nurses’ midwives’ and paramedics’ psychological ill-health. We have also begun to translate these recommendations for different audiences (noting that these are interdependent) (see Table 11).

BOX 2

Eight overarching recommendations to tackle nurses’, midwives’ and paramedics’ psychological ill-health

TABLE 11

Key recommendations for framing narratives and refining/developing strategies to reduce psychological ill-health to reduce mental

To translate our project findings and recommendations (see Box 2 and Table 11) into solutions and resources we have also started working with our stakeholder and advisory groups to determine what could further support our various recommendations and develop our guide for managers (see Appendix 10). Our approach to designing and developing the resources is underpinned by the CFIR⁶⁸ as shown in Table 12 (below).

TABLE 12

Application of the CFIR to the design, development and implementation of resources from this study

CUP-2 strengths and limitations

There are several key strengths of this study. The use of realist methodology was beneficial for uncovering insights that went beyond the surface-level, well-established understanding of workplace psychological ill-health for three professional groups. It also helped us identify both causes and solutions comparing nurses, midwives and paramedics to identify important contextual factors, as well as subsets of individuals who were (more) affected within professional groups, and when. Our realist synthesis included different papers to other reviews such as commentaries and editorials, which offered rich insights that would usually be excluded in other review methodologies.

Realist methodology also allowed us to take an iterative flexible approach to searching and analysis, which meant that we were able to accommodate recent relevant data on COVID in a way that did not overwhelm the core study. We used the RAMESES quality standards for realist synthesis to ensure that the study design was rigorous and in line with principles of the realist approach.⁴⁰

The multidisciplinary skills and expertise of the core team (including experienced PIs and co-applicants, PIs immersed in the relevant literature, realist methods expertise, and an information specialist) and the linkage within the team with CUP-1 is a key strength of this project. This core team strength was supplemented by the expertise in our project advisory group, having both subject and methodological expertise, further strengthening the study, ensuring a robust approach and the inclusion of expert literature suggestions. We also had a diverse stakeholder group, which included staff experts by experience, providing diverse perspectives (multiprofessional and cross-professional and thereby ensuring discussions were not siloed). We ensured a supportive, safe and confidential environment by implementing many of the ground rules used in Schwartz Rounds (both JM and CT are experienced Schwartz Rounds facilitators), including clarity around confidentiality within the group, and enabling contribution in anonymous ways (using an online whiteboard [Padlet]). We also ensured the safety of members by providing the offer of psychological support (via Dr Diana Bass) to any members that may need it. Advisory and Stakeholder members critiqued and helped us make sense of the findings, which has strengthened the outputs and ensured relevance of the findings to the real-world problems faced in health care by nurses, midwives and paramedics and enabled translation of these findings into recommendations for practice.

Using retroduction (identification of hidden causal forces that lie behind identified patterns or changes in those pattern) we were able to develop complex findings around ‘tensions’ in healthcare architecture that help explain psychological ill-health in our staff groups. These tensions are underexamined in the literature and hold much potential for development in thinking about how to improve work conditions for the psychological health and well-being of healthcare staff. While this approach re-establishes and deepens our understanding of this topic it does not produce black and white answers. Furthermore, the cross-professional analysis (comparing nurses, midwives and paramedics, and also to doctors through comparison with CUP1) has significant benefits. Most previous research has tended to focus on whole healthcare workforces or one professional group (or a subset of that professional group), though a systems focus is essential to solutions to psychological ill-health as the healthcare system is inherently multiprofessional.

In term of limitations, due to the broad scope of the review, we analysed the data for nurses, midwives and paramedics separately and in some instances extrapolated these findings to apply across the professional groups, in discussion with our advisory and stakeholder groups. More research is needed regarding the role and service architecture features that may be distinct and place staff at greater risk. While there may have been generally patterned distinctions in the findings, there were always exceptions to the rule thus, further research will be necessary to build on this study.

The database searching for this review involved three separate searches for the three professions under study. The paramedic search terms were designed through iterative searches, in consultation with our stakeholder group, as the initial search retrieved a very limited set of papers. To ensure our review was as relevant to a UK context as possible, we applied database limiters, which in the CINAHL database were not as accurate as we would have liked, that is, some UK papers were not identified by the filter. However, we were still able to identify a focused sample of relevant UK papers and we were not aiming to search the literature comprehensively. We did not carry out citation searches, which are commonly used in realist reviews, as hand searching, and stakeholder/expert suggestions was an efficient way to identify papers that the database searches had failed to retrieve papers to approximate the quotas.

To ensure our review was up to date, and to manage the large literature (particularly in nursing) we used a reverse chronology screening. This enabled us to initially exclude the COVID-19 literature, and then subsequently return to this, and to stop searching at a pre-determined number of papers for each profession, for our initial analysis. The limitations of this approach means we might have missed significant literature, however, our subsequent inclusion of systematic and other reviews and use of key reports together with the subject expertise in the core team, advisory and stakeholder groups make this less likely but it remains a possibility. That said, a realist review is not intended to be exhaustive but takes a sample of literature for deeper insights going beyond the surface-level. The descriptive analyses of causes and interventions required an element of subjectivity for the categorisation into themes (HSE domains, informal/formal interventions and primary, secondary, tertiary target levels), and as such should be interpreted with caution. Furthermore, our review is limited in generalisability due to the key gaps in the literature (see p. 75) including lack of focus on intersectionality and how gender, ethnicity, disability and other factors (and their inter-relationships) may place individuals at greater risk of psychological ill-health.

Future research

Future research examining psychological ill-health in nurses, midwives and paramedics should build on our synthesis and

1. seek to evaluate and refine systemic interventional strategies already implemented that take account where possible of well-being bundles and/or systems approaches to intervention (primary to tertiary levels, and informal and formal approaches);
2. build future research programmes and design, implement and evaluate new interventional strategies, where possible tailoring to local organisational and workforce needs and co-designed with front-line staff and staff experts by experience;
3. seek funding that prioritises complex evaluations of whole systems approaches rather than only individual-focused secondary interventions;
4. identify and develop more sophisticated outcomes rather than those just easy to measure;
5. investigate further the tensions identified in this study, for example, cumulative impact of everyday stressors not only acute or one-off traumatic incidents;
6. explore the role of undergraduate education in preparing for psychological wellness throughout their career, including development and evaluation of anticipatory socialisation programmes or reflective spaces that are known to be beneficial;
7. investigate further the impact of equality diversity and inclusion issues, particularly the role of intersectionality on psychological ill-health;
8. continue to include health economic analysis in studies and investigate the cost-benefits on investing in staff psychological health.

Equality, diversity and inclusion

We have tried hard to address issues of equality, diversity and inclusion (EDI) in our study, but are limited by what has been published on these issues related to psychological ill-health in nurses, midwives and paramedics. We have used the lens of gender and ethnicity to explore both causes and to a lesser extent intervention. We were not able to retrieve relevant material directly relating to LGBTQ+ and disability issues in our sample for staff psychological well-being. We have included this as a focus for more research in this area and future studies may wish to use specific search terms in later cycles of realist synthesis to specifically search for EDI issues related to psychological ill-health. Our research team and stakeholder groups does include those from groups generally under-represented and there was a range of experience and expertise across the research team. Stakeholder participants and members of the public were recruited from across the country and involved representation of nurses, midwives and paramedics, some of whom had lived experience of psychological ill-health. We repeatedly discussed the diversity of the stakeholder groups with them and asked for their help to increase diversity, which resulted in more inclusion. Development opportunities and training were provided for research team members and psychological support was provided by Dr Diana Bass for experts by experienced members and as needed members of the team.

Patient and public involvement

Our engagement with the experts by experience (nurses, midwives and paramedics) and members of the public commenced before our study started with interviews with nurses, midwives and paramedics (n = 10) and this engagement continued throughout the study, as outlined in Chapter 2 (Stakeholder group contributions to analysis). We have held four project stakeholder group meetings and two project advisory group meetings. In each of these there has been public representation and representation of experts by experience (nurses, midwives and paramedics who have experienced psychological ill-health at work). This involvement had significant impact during the study; helping refine the project team’s ideas and thinking, providing a real-world perspective and challenging some of our suggestions (see Chapter 2, Table 5 in Stakeholder group contributions to analysis). In summary, our stakeholder PPIE members commented on and helped revise our theories, tensions and findings as the study progressed. Their views are embedded in the realist chapter and this report. They have also offered guidance on the dissemination of findings and how these can best have impact and we will be holding a dissemination event in December 2022/January 2023.

Conclusions

Our realist synthesis unequivocally suggests the need to improve the systemic working conditions and the working lives of nurses, midwives and paramedics to improve their psychological well-being. Individual, one-off psychological interventions are unlikely to succeed alone. Psychological ill-health is highly prevalent in these staff groups (and can be chronic and cumulative as well as acute) and should be anticipated and prepared for, indeed normalised and expected.

We expected to find variation between our three staff groups but found more similarities than differences. It is the way jobs are designed (service architecture) that can increase risk rather than the profession itself and these risks can be further exacerbated by intersectionality influences. Organisational and team cultures matter, and it is difficult to promote staff psychological wellness where there is a blame culture, and where the needs of the system override staff psychological health. We found that interventions are fragmented and individual-focused with an insufficient focus on systemic and hygiene issues (work dissatisfiers).

Synthesising the literature using a realist approach has allowed a nuanced and richer understanding of context and has enabled us to provide new insights into the body of evidence and make recommendations for practice and for policy. We had excellent engagement from our stakeholder groups, including nurse, midwife and paramedic experts by experience whose wise advice we have sought to heed. In the future, we anticipate the need for more research, particularly evaluation of system-wide, multipronged interventions. Due to their complexity these are difficult to evaluate well, yet vitally important for the systemic changes our study suggests are required if we are serious about supporting staff to care well as well as avoid psychological harms for themselves.