‘The WOW factors’: comparing workforce organization and well-being for doctors, nurses, midwives and paramedics in England

Abstract Background High rates of poor mental health in healthcare staff threatens the quality and sustainability of healthcare delivery. Multi-factorial causes include the nature and structure of work. We conducted a critical review of UK NHS (England) data pertaining to: doctors, nurses, midwives and paramedics. Sources of data Key demographic, service architecture (structural features of work) and well-being indicators were identified and reviewed by a stakeholder group. Data searching prioritized NHS whole workforce sources (focusing on hospital and community health services staff), which were rated according to strength of evidence. Findings Key differences between professions were: (i) demographics: gender (nursing and midwifery female-dominated, doctors and paramedics more balanced); age (professions other than doctors had ageing workforces); ethnicity (greater diversity among doctors and nurses); (ii) service architecture: despite net staffing growth, turnover and retention were problematic in all professions; 41.5% doctors were consultants but smaller proportions held high grade/band roles in other professions; salaries were higher for doctors; (iii) well-being: all reported high job stress, particularly midwives and paramedics; sickness absence rates for nurses, midwives and paramedics were three times those of doctors, and presenteeism nearly double. Growing points Sociocultural factors known to increase risk of poor mental health may explain some of the differences reported between professions. These factors and differences in service architecture are vital considerations when designing strategies to improve well-being. Areas timely for developing research Multi-level systems approaches to well-being are required that consider intersectionality and structural differences between professions; together with inter-professional national databases to facilitate monitoring.


Introduction
The well-being and mental health of healthcare professionals has been gaining increasing attention as a major public health concern and threat to the quality and sustainability of healthcare delivery-in the UK and globally. This has been spotlighted and further exacerbated by the COVID-19 pandemic with the added pressure on healthcare staff of delivering care in extreme circumstances. 1 The National Health Service (NHS), one of the world's biggest employers (and the biggest in the UK), employs nearly 1.6 million people 2 and needs healthy, motivated staff to provide high quality patient care. However, increasing workload due to societal demand for healthcare services, combined with increasing external scrutiny of their work, has been associated with a high prevalence of mental illhealth amongst staff. Due to budget constraints and staff shortages, pressure is building in the healthcare system and this is taking its toll on staff as well as patients. 3,4 Some commentators have described staff as the 'shock absorbers in a system lacking [the] resources to meet rising demands', and suggest the current situation is not sustainable. 5 Neglecting the well-being of healthcare staff has significant implications for staff and patients. Although the NHS as an employer has a duty of care to staff, staff well-being also affects patient care, safety and delivery. High levels of stress and burnout among NHS staff affect their ability to provide high quality care. [6][7][8] In the UK, the mental health of the NHS workforce is a major issue, leading to presenteeism (working while unwell), absenteeism and loss of staff from the workforce. 4,9 Stress among healthcare staff is greater than in the general working population and explains >25% of staff absence, 10 and depression, anxiety, loss of idealism and empathy are also reported by nurses and doctors. [11][12][13][14] NHS staff sickness absence rates are double the national average 15 and are estimated to cost £1.1 billion. 4 Multiple government and industry reports and publications have highlighted the need to reduce stress and improve mental health in NHS staff, e.g. 4,[16][17][18][19] Staff well-being is a pressing and complex problem influenced by many factors at individual, organizational, inter-professional and broader societal level. Research highlights the need for workplace policies and interventions to be informed by an indepth understanding of such factors, and for more engagement with healthcare workers, in order to develop effective policies and interventions. 1,20 Multiple professions and specialities are involved in the delivery of healthcare, and often share the same work environment, but they also have very different roles and responsibilities, and potentially different structural contributors to staff well-being and poor mental health.
In the NHS in England, the types of services and treatments available is determined regionally by clinical commissioning groups (CCGs). In 2020, there were 135 CCGs. NHS Trusts provide the services/treatments commissioned by the CCGs and include hospital, ambulance, mental health, social care and community services. Primary care is delivered in GP practices who work within primary care networks (PCNs). There are ∼1300 PCNs currently in England, each covering a population of 30-50 000 people. The most robust and accurate workforce data available for NHS staff are the NHS Workforce statistics produced by NHS Digital (validated data extracted from the NHS Human Resource and Payroll System). Although these provide extensive data for hospital and community health service workers (covering all the types of Trust listed previously), reporting of data for primary care NHS staff is currently limited in scope.
This paper therefore focuses on hospital and community NHS staff from four professions; doctors, nurses (registered nurses only), midwives and paramedics, comparing features of these professions and how that profession's work is structured that may be pertinent to understanding their well-being, which we have conceptualized as the 'service architecture'. This work builds on previous work focussed on doctors 20 Care Under Pressure (completed in 2019) and a current study focussed on nurses, midwives and paramedics: Care Under Pressure 2 (ongoing to July 2022) 1 .
A key recommendation of Care Under Pressure is that policies that aim to secure the future of the NHS workforce should foster a supportive work culture in which individuals can thrive. Policies and interventions that target the individual in the absence of a supportive work culture are unlikely to succeed. 20 As part of the ongoing work on Care Under Pressure 2 we realized the importance 1 Care Under Pressure 2: Caring for the Carers a realist review of interventions to minimize the incidence of mental ill-health in nurses, midwives and paramedics National Institute for Health Research Award ID: NIHR129528 https://fundingawards.nihr.ac.uk/award/NI HR129528 of investigating whether and how organizational factors-service architecture-that may differ within and between these professional groups may be important contributors to mental ill health.
We have selected these groups, because together doctors, nurses, midwives and paramedics comprise around 60% of the clinical workforce in the UK NHS. All have high rates of illness, and pressing recruitment and retention issues, but each profession also has distinct structural features. To our knowledge, this is the first time that this type of multi-professional comparative work has been undertaken. Given the evidence of poor mental health and challenges to staff well-being in the UK NHS and the current problems with recruitment and retention, it is important to gain an understanding of which contextual factors have resulted in these (unintended) impacts and to equip NHS managers, policy makers, leaders, staff, researchers and other stakeholders with this understanding. A necessary first step is to extract and collate such detail to enable comparison.

Methods
Aim: to extract, synthesize, critically review and compare workforce demographic, service architecture and well-being data for doctors, nurses, midwives and paramedics working in hospital and community health service settings in England, in order to enhance understanding of shared and distinct contextual factors that may contribute to their poor mental health at work for the benefit of managers, policy makers, researchers, staff and other stakeholders.

Objectives
• Identify the key workforce demographic and service architecture features that may differ within and between professional groups and be important contributors to mental ill health. • Source and extract data regarding these workforce features and measures of well-being/mental illhealth, including assessment of the data in relation to (i) strength/accuracy of evidence; (ii) comparability across professions. • Produce a summary of the key features and how they compare and contrast across and within the four professional groups, and describe their potential relationship to well-being/mental ill-health.

Design
A critical review aims to go beyond description of the included sources and include a degree of analysis and conceptual innovation, resulting in a model or new interpretation of existing data. 21

Identification of key contextual features and stakeholder involvement
Key contextual features that may be important contributors to mental ill-health for each profession (doctors, nurses, midwives, paramedics) were brainstormed by the author team and expanded further through sharing drafts with two separate stakeholder groups formed to support wider projects on the causes of poor mental health in nurses, midwives and paramedics 2 (Maben et al., 2020b), and doctors 20 (Care Under Pressure, and Care under Pressure 2). The stakeholders comprised doctors, nurses, midwives and paramedics-including those with self-disclosed lived experience of work-related poor mental health; representatives from relevant regulatory bodies and professional organizations; and patient/public representation. Stakeholders were asked to comment on an initial draft of the demographic, service architecture and wellbeing features felt to be important to capture and compare across (and within) professional groups, in particular to state if there were any omissions. Feedback suggested that our identified factors and features provided a useful summary of key statistics 2 Care Under Pressure 2: Caring for the Carers a realist review of interventions to minimize the incidence of mental ill-health in nurses, midwives and paramedics National Institute for Health Research Award ID: NIHR129528 https://fundingawards.nihr.ac.uk/award/NI HR129528 that could inform attempts to improve workforce well-being. Limitations in relation to lack of data specifically for the primary care workforce was noted, and we agreed that it would be beneficial to include types of settings in which different health professionals work (e.g. community, primary care, acute settings) if such data were available. Unfortunately, we have been unable to find such data in reliable sources and consistent formats, hence our decision to focus on hospital and community health service settings in England only.

Data sources
For each key feature, searches were conducted for relevant data using a stepped approach, ordered according to the credibility and comparability of data. This began with attempts to find relevant data using NHS Digital (NHS Workforce Statistics), and/or NHS England-related sources based on the whole NHS hospital or community services workforce in England, prioritizing those sources where the data could be broken down by the four professions of interest. The most recent sources were used where possible, in order to provide the most relevant up-todate data, but with priority given to using a slightly older source if it meant better comparability across professions. This included data from 2016 to 2021 (the majority of comparable NHS Digital data was from 2018, and NHS Staff Survey data were taken from the latest published survey results, 2020). If these searches were unsuccessful, the next step was to search profession-specific national (or UKwide) sources such as the relevant regulatory bodies (General Medical Council for doctors; Nursing and Midwifery Council for nurses and midwives; and the Health and Care Professions Council for paramedics), or professional bodies/membership organizations (e.g. Royal Colleges for doctors, nurses and midwives and the College of Paramedics). We also asked our stakeholders to suggest data sources/contacts relevant to specific professional groups if we were struggling to access data. Following these attempts, other sources were examined such as charitable organizations/trade unions (e.g. the Kings Fund), university and other relevant websites, internet searches (e.g. via google); and searches for empirical research. For some variables the data for a profession includes other related staff, most notably for paramedics where data are often reported by NHS Digital for Ambulance Staff as a group, comprising: managers, emergency care practitioners, paramedics and ambulance technicians; and data for doctors from the NHS Staff Survey are only available for medical and dental staff combined. Moreover, NHS Digital data for Hospital and Community Health Services (HCHS) doctors include a small number of Hospital Practitioner/Clinical Assistant, who may not be medically qualified.

Data extraction
Data for demographic features, service architecture features and workforce well-being outcomes were extracted from the cited sources and are presented in Tables 2-4, respectively. Since data were presented in varying ways in different sources, for different professional groups and different features, it was necessary to transform some of the data to enable comparability across features and across professional groups. This was the case for any figures that had been reported as total numbers, which have been transformed into percentages (using a defined denominator) to enable comparability between staff groups.

Appraisal, synthesis and analysis
Data were evaluated according to the overall strength of evidence they provided 'within' the professional group. This was based upon an assessment of their representativeness and/or completeness in relation to the whole population of doctors, nurses, midwives or paramedics in hospital or community service settings in England; and in relation to the validity of the measure, i.e. how the data was collected (see Table 1). After appraising the data's quality and strength 'within' each professional group, the data were rated in relation to the validity of comparing 'across' groups (Table 1). Using this approach, each row of data in Tables 2-4 has a rating (of high, moderate or low) for within group and between group comparisons. The rating tool was developed specifically for this review as there were no available tools that would allow both strength of evidence within and between professional groups. CT and AC lead appraisal process, though all ratings were reviewed and confirmed by all other authors.

Results
Tables 2-4 provide comparative data for four key professions within the NHS hospital and community services workforce in England. The tables facilitate comparison across the different professional groups and draw attention to the key features of the professional contexts that may contribute to well-being or mental ill-health of these critical NHS staff. In the narrative summary below, we present the information relating to three categories: Demographics, Service Architecture and Workforce Well-being.

Demographics
The professions of nursing and midwifery are heavily female dominated, with only 11.6% and 0.4% male staff, respectively ( Table 2). The professions of medicine and paramedic science are more gender balanced with 54.8% and 59% male staff, respectively. In terms of ethnicity, there are striking differences. Very high proportions of midwives and paramedics (85.4% and 93.9%) report their ethnicity as White, compared to 49.1% of doctors and 70.6% of nurses. Over a quarter (27.5%) of doctors report their ethnicity as Asian, compared to 10.5% of nurses and 1-2% of midwives and paramedics. There also appear to be more nurses identifying as Black (8.4%) and more doctors identifying as Chinese (2.3%) than other professions. At least 70% of all four professions report their nationality as UK. The medical profession has the most members from the EU (9%) and from the rest of the world (16%) followed by nurses (7% EU and 9% rest of the world). In terms of age, there are quite different pictures, with the

Service architecture
Service architecture is our way of conceptualizing the structural features of a profession, including a focus on features that may be pertinent to understanding their well-being (Table 3).

Staff turnover, retention and retirement
Data suggest a positive trajectory in the size of the NHS workforce. Between February 2019 and February 2020, there was a net growth in number of doctors (+5.4%), nurses (+2.8%), midwives (+2.2%) and ambulance staff (+0.5%). Of those joining the

Salary and pay gaps
The average annual basic pay for doctors (£68 777) is nearly double that of the other three professions, with midwives (£36 059) earning slightly more than nurses (£34 275) and paramedics (£33 487). It is important to note that this figure only includes NHS earnings, and excludes any additional salary from private practice. Doctors are also more likely to receive additional payments for working on-call (34.3% vs 17.4% midwives, 7.9% ambulance staff and 4.1% nurses, data taken from same source as salary). Across all four professions, there is a gender pay gap with average pay for female staff less than the average pay for male staff, and this varies from 1% in nursing to 15% in medicine. Across all four professions, the average pay for staff who report as BAME (Black, Asian or Minority Ethnic) in terms of ethnicity is less than the average pay for staff who report as White, and this varies from 1% in midwifery to 10% in medicine.

Working hours
All

Education and training
All four professions now require a university degree for entry to the profession and all four are required to pass examinations to allow them to register as a professional with their respective registering body. As undergraduates, doctors spend a smaller proportion of their time in clinical practice (around 25% overall), whereas the other three professions spend 50% of their time as undergraduates on placements in clinical practice. Doctors also spend much longer in training, both as undergraduates and after graduation, compared to the other three professions. Medical training typically involves 5 years of undergraduate study and 5-9 years of postgraduate training, whereas the other three professions typically involve 3 years of undergraduate study and have no requirement for postgraduate training (though many opportunities exist, including advanced practice Masters and doctoral qualifications and specialist practitioner courses).
When we look at the proportion of sickness absence due to anxiety/stress/depression/other psychiatric illness, this ranges from 24.1% for doctors to 34.7% for midwives. Presenteeism also appears to be higher in nurses, midwives and paramedics (49.3%, 55.3% and 56.3%, respectively reporting working when unwell in the NHS Staff Survey 2020) compared to 30.3% of doctors. The majority of all four professions report having unrealistic time pressures (between 80.8% of doctors to 89.7% midwives), and high proportions in each profession reported feeling unwell as a result of work-related stress (from 39.8% medical and dental staff to 58.2% paramedics). Data were also extracted for 2018 and 2019 in case there was a 'pandemic' effect of using the 2020 NHS staff survey data, but we found no evidence of this with little change in these variables in any of the groups over this period.

Discussion
The mental health and well-being of healthcare workers has been a pressing concern for many years, and has been intensified by the ongoing COVID pandemic. 1,23 Poor mental health is the consequence of a complex interplay of bio-psycho-social-cultural factors, among these, the nature and structures of healthcare work may be major contributors.
Although some of the features of work relating to poor mental health are common to all NHS staff, some key features and patterns indicate unique differences that are important to note and take into account when designing, implementing and evaluating interventions to improve well-being of NHS staff. This review presents some of this data, providing a resource to support this endeavour. In relation to demographics, there are some stark differences by gender and whilst our focus is on work factors in this paper, various social and economic factors can put women at greater risk of poor mental health than men and thereby may go some way to explaining the high prevalence of poor mental health in the female dominated professions. These factors include being more likely to undertake caring roles, live in poverty and experience domestic abuse. 24 Furthermore, female dominated professions may be more open to reporting poor mental health. In relation to age profile of the workforce, medicine has a younger workforce, and nursing and midwifery have an ageing workforce. This suggests that there may be greater problems with workforce retention in medicine and/or that the peak at an earlier age in medicine is the result of greater investment in medical student numbers working their way through the system. This also indicates that there are difficult times ahead for nursing and midwifery, as many experienced professionals near retirement. In nursing this has been referred to as a demographic timebomb. 25 It is critically important to consider ways of encouraging the next generation into healthcare careers. We know that career choices for Generation Z (those born 1995-2010, so those entering the labour market now) are influenced by wanting to work for organizations that promote healthy practices and healthy working environments, 26 and research has shown the potential 'fit' between Generation Z values and caregiving careers. 27 In relation to diversity, the professions with lower ethnic diversity (nursing, midwifery and paramedic science) also have the highest vacancy rate. There are also considerable gender and BAME pay gaps across professions. In medicine the gender pay gap has been explored more comprehensively than the data we used here allows, and a greater gap than reported here was found (18.9% for hospital and community health services doctors, 15.3% for GPs, adjusted for differences in working hours). 28 The Workforce Race Equality Standard 29 highlights variations in staff experience according to ethnicity, across NHS trusts in England, and is challenging race inequality in the health and care system. Policies and strategies that aim to improve equality, diversity and inclusion within and across professions are not only a moral imperative, but are likely to improve recruitment and retention in Generation Z cohorts, improve the wellbeing of staff (e.g. reducing potential stigma and unprofessional behaviours including bullying) and also improve quality of patient care. 30 In relation to 'service architecture' the four professions have many distinct features that may be important when trying to understand the causes of poor mental health. Although there has been net growth in numbers within each profession, there has also been an exponential growth in demand, and this is within a context of chronic under-investment and staff shortages, 31 and an exacerbation of the shortages caused by Brexit. 32 Thus it is unlikely that this growth in numbers will be sufficient. Furthermore, the numerical staffing levels we have reported can mask nuances that are important to consider, for example which NHS staff (in terms of grade and experience) are leaving and joining and the employment status of staff (e.g. nurses have a high proportion of bank staff). Replacing experienced leavers with newly qualified joiners does not plug the workforce deficits alone-it is critical to also implement strategies to retain experienced staff. Consideration also needs to be given to the speed at which healthcare professionals are trained. The F2 Career Destinations Survey for doctors 33 shows a rapid decrease in recent years of the proportion of doctors who, 2 years after graduating, continue directly onto the next stage of training. These doctors are not necessarily leaving medicine but are slowing down their progression, either to support personal or professional development, 34 and/or to manage stress, regain control of their life and work. 35 Ensuring evidencebased support for staff throughout their training and practice is essential to reduce this attrition.
All four professions experience poor levels of workplace well-being, according to all of the metrics presented in this paper. A notable finding is the difference in sickness absence rates between doctors and the other professions (over a 3-fold difference), a pattern that continues for rates of presenteeism. It is unclear why this is. It may be explained in part by gender socialization theory and gender traits: that it is more acceptable for women to report being stressed than men, and therefore the female dominated professions having higher rates. 36 This does not explain why the rates are similar in paramedics though who are a more gender-balanced profession. It is more likely a complex interplay of the biopsycho-social-cultural factors that interact with gender and these professions, for example those with lower income and status being at greater risk of poor mental health. The barriers to taking time off sick may be greater for doctors, including that it may be harder for them to report poor well-being either culturally and/or practically, as they are less likely to be registered with and/or consult with their own GP. 20,37 The stigma of mental ill-health and impact on colleagues has been reported by doctors, nurses, midwives and paramedics. [38][39][40] Media reports and now published research on experiences of staff during COVID-19 tell us that NHS staff have long been experiencing a mental health crisis, but that has been made significantly worse by the COVID-19 pandemic. [41][42][43] This is not reflected in the NHS Staff Survey findings reported here perhaps as the measures were not sensitive to the impact of COVID on the mental health of staff, or because they were collected too early in the pandemic. Increasing support for NHS staff wellbeing is thus vital. Our current research study Care under Pressure 2 (nurses, midwives and paramedics) will complete summer 2022, and the next steps are to ensure a pathway to impact by embed this research into practice by testing and refining this knowledge and optimizing its implementation in the NHS. To do this we aim to create resources to augment the NHS Health and Wellbeing Framework (HWF). 44 This Framework was first launched in 2018 by NHS England and Improvement and NHS Employers and provides an interactive toolkit that makes the case for staff health and well-being, sets out clear actionable steps and includes guidance on how organizations can plan and deliver a staff health and well-being strategy. This framework takes a 'systems and multi layered' approach to health and wellbeing (from prevention to treatment, and individual and organizational strategies). Although an excellent resource, currently the NHS HWF has a generic NHS workforce focus (not specifically for doctors, nurses, midwives and paramedics), and our ongoing planned work (through new studies Care Under Pressure 3 and 4) aim to add resources to this framework and optimize their use and implementation in practice.
Through completing this critical review, we have learned that this type of comparative work is not as straightforward as it might seem, that some key data are not available, or need transforming to be comparable for example, but it can generate significant insights, and has significant potential for impact. The findings may help NHS managers, policy makers, leaders, etc. to see where improvement strategies from one profession/setting might be transferrable to another profession/setting, and can also help with targeting/prioritizing the implementation of different initiatives given finite resources (time/money). This review is limited by the data available, which in some cases is either a few years old and/or has limited comparability across professions. There are important features of work or of the workforce that we do not have reliable data about and therefore could not include: in particular the primary care workforce, which is sizeable, and the settings in which staff work. In addition, sometimes the data do not reflect the true picture on the ground, for example sometimes posts are not advertised because it is not felt they could be filled and workarounds are made to cover service needs, therefore masking the true vacancy rate.
This review presents novel inter-professional comparative work, enabling healthcare leaders, managers and other stakeholders to considerand develop strategies to mitigate-the potential impact of these distinct demographic and service architecture profiles on well-being of the workforce. Healthcare relies on interdisciplinary working, and attempts to improve workforce well-being require multilevel systems approaches, from prevention to treatment, that take into account similarities and differences across professions. The development of more harmonized inter-professional national databases, could in itself be a resource to monitor and improve healthcare staff well-being.