Uncovering, creating or constructing problems? Enacting a new role to support staff who raise concerns about quality and safety in the English National Health Service

Abstract

Employee voice is an important source of organizational intelligence about possible problems in quality and patient safety, but effective systems for encouraging and supporting those who seek to speak up have remained elusive. In the English National Health Service, a novel role known as the ‘Freedom to Speak Up Guardian’ has been introduced to address this problem. We critically examine the role and its realization in practice, drawing on semi-structured interviews with 51 key individuals, including Guardians, clinicians, managers, policymakers, regulators and others. Operationalizing the new role in organizations was not straightforward, since it had to sit in a complex set of existing systems and processes. One response was to seek to bound the scope of Guardians, casting them in a signposting or coordinating role in relation to quality and safety concerns. However, the role proved hard to delimit, not least because the concerns most frequently voiced in practice differed in character from those anticipated in the role’s development. Guardians were tasked with making sense of and dealing with issues that could not always straightforwardly be classified, deflected to the right system or escalated to the appropriate authority. Our analysis suggests that the role’s potential contribution might be understood less as supporting whistleblowers who bear witness to clear-cut wrongdoing, and more as helping those with lower-level worries to construct their concerns and what to do with them. These findings have implications for how voice is understood, imagined and addressed in healthcare organizations.

Introduction

Healthcare systems worldwide increasingly recognize the importance of employee voice in improving the quality and safety of the care they provide. Long acknowledged as a source of organizational intelligence across a range of industries (Morrison, 2014), employee voice goes by a number of names (including, for example, voice behaviour, raising concerns and speaking up). There is some consensus that voice may offer especially valuable insights about problems that are not readily detected through formal, measurable indicators (Martin et al., 2015), and that it may be especially useful in high-risk settings in identifying risks to safety (Macrae, 2014). But, despite evidence that failures of voice have been implicated in crises in healthcare worldwide (Cleary and Duke, 2019; Newdick and Danbury, 2015; Weick and Sutcliffe, 2003), challenges in encouraging and responding to voice are widely reported within healthcare (e.g. Jones and Kelly, 2014a) and beyond (e.g. Kish-Gephart et al., 2009). The difficulties individuals face in raising concerns, and how they might be addressed, are the focus of a growing body of research and policy activity.

In the United Kingdom, efforts to improve voice have been driven by high-profile incidents where failure to speak up (or to listen to concerned individuals) has been implicated. Particularly notable was one the most significant disasters in English National Health Service (NHS) history – the failings at Mid Staffordshire NHS Foundation Trust, where problems of voice were endemic in the 2000s. The scandal had a profound impact on healthcare policy in England (Martin and Dixon-Woods, 2014), giving rise to a public inquiry (Mid Staffordshire NHS Foundation Trust Public Inquiry, 2013) and corresponding responses by the government, including commissioning a review of culture and practice around raising concerns among healthcare professionals (Francis, 2015). The review found a widespread reluctance to speak up among staff, linked to doubts about whether authorities would listen and to concerns about retribution. It recommended several measures to foster a culture of speaking up, where ‘injustice to whistleblowers should become very rare indeed, [and] is redressed when it does occur’ (Francis, 2015: 196). A flagship proposal was the introduction of a new role, the ‘Freedom to Speak Up Guardian’, in every healthcare provider in England across the acute, mental health, community health and ambulance sectors. The Guardian role was intended ‘to act as an independent and impartial source of advice to staff at any stage of raising a concern, with access to anyone in the organisation, including the chief executive, or if necessary, outside the organisation’ (NHS Improvement, 2016: 5).

The Freedom to Speak Up Guardian is a novel intervention, with no obvious parallels or precedents for the role internationally, in healthcare or elsewhere (Jones et al., under review). Much has been made of the role and its potential to provide guidance and support for individuals with concerns and contribute to wider cultural change about the importance of speaking up (Department of Health, 2015). Given the ubiquity of problems of voice in healthcare systems worldwide, the Guardian role is of international interest. However, though it has received significant organizational and governmental investment, it has remained unevaluated. In this article, we examine the Guardian role and its potential through empirical study of the realization of the role in practice, bringing empirical evidence and theoretical insights on voice in organizations to bear on our analysis. We derive wider insights into prevailing understandings of speaking up and speaking out, their implications for how voice is understood and imagined in healthcare organizations and the consequences for how concerns are identified, valued, addressed and learned from.

We begin by exploring some current major themes in the literature on employee voice, noting emerging insights into influences on voice and recent analytical advances that distinguish a spectrum of heterogeneous voice behaviours. We then examine the origins of the Freedom to Speak Up Guardian role, and the form it was proposed to take in English healthcare organizations.

The dynamics of speaking and listening in healthcare

A growing literature on employee voice across multiple industries has identified a range of psychological and organizational characteristics that encourage or inhibit people in speaking up about concerns. Employee voice covers a wide range of behaviours, from raising issues informally with colleagues, through more the formal act of ‘raising concerns’ formally with those in more senior positions, to high-profile acts of ‘whistleblowing’, often to authorities outside the organization (such as regulators or the press) (Mannion and Davies, 2015). Despite the divergent character of these different types of voice behaviour, they have often been used interchangeably in the literature. In this article, we use ‘employee voice’ and ‘voice behaviour’ to cover the range of such acts, but distinguish between ‘speaking up’ within a unit or organization and the higher-stakes, higher-consequence act of ‘whistleblowing’ externally. We exclude uses of ‘employee voice’ that relate to comparatively trivial activities such as staff engagement (e.g. in relation to branding), except insofar as they refer to raising concerns.

Two strands of literature are significant in pointing towards the need for better understandings of the influences on individuals’ decisions about whether to give voice to concerns in one way or another, the range of behaviours that ensue when individuals do decide to act, and the implications for any intervention seeking to foster voice. First, a major recent theme is the psychological and organizational influences on individuals’ voice behaviour, including the complex of factors that contribute to decisions about giving voice to concerns (e.g. Kish-Gephart et al., 2009). A common approach has been to parse these factors into attributes of the individual (candidate) speaker, and the features of the organizational context in which the speaker is positioned. For example, Okuyama et al.’s (2014) review finds a wealth of research devoted to identifying ‘the barriers and promoters of speaking up’. It distinguishes ‘general contextual factors’ affecting a decision to speak up (such as the strength of managerial support for speaking up, and professional expectations) from ‘individual factors’ (such as job satisfaction, sense of responsibility to patients, and confidence). In combination, these factors may render speaking up ‘a high-risk:low-benefit act’ (Attree, 2007: 95), such that silence is a rational response. The implication is that voice might be encouraged by creating contexts that encourage and act upon voice, and/or by endowing individuals with the inclination and ability to speak up.

This cognitive-rationalist construction of the decisions to raise concerns is, however, the subject of growing critique. Field and experimental studies demonstrate that, in any given situation, individuals draw heuristically on a range of considerations, such as their implicit understandings of whether speaking up will be deemed appropriate by a range of audiences, including managers and colleagues. For example, Detert and Edmondson (2011: 481) identify the range of ‘implicit voice theories’ that may guide behaviour, pointing out that these theories are often informed not by explicit cues, but by a sense that speaking up may be ‘wrong or out of place’. These employee-held implicit theories may have deep roots (Henriksen and Dayton, 2006). A key corollary of this line of research is that efforts to promote voice must do more than address the formal signals conveyed by organizations about the propriety of speaking up. Indeed, mere changes to formal policy that contradict years of accumulated experience and shared lore about the consequences of voice may have exactly the opposite effect to that intended (Cunha et al., 2019).

A second, related, theme in recent literature is the contention that attempts to raise concerns are not best characterized using the binary distinction between voice and silence that has traditionally predominated. One reason for this is that, in healthcare in particular, much activity that might be the subject of concern is inherently uncertain: a matter of inevitable compromise arising from ethical trade-offs, resource limitations, imperfect information and so on (Bosk, 2005). Such activity may also be emergent: perhaps only rarely will a single act or event be recognized as an unambiguous imperative to speak up. What people do in response to concerns is not simple either: they may use varied tactics, including peer-to-peer communication (Tarrant et al., 2017), informal concern-raising (Jones and Kelly, 2014b) and a spectrum of more formal options, often invoked incrementally (Martin et al., 2018).

Many of these voice-related activities cannot accurately be categorized as silence (Jones and Kelly, 2014a), but they may not be full-throated voice either. Yet, as Mannion et al. (2018) discuss, policy effort and public discourse is dominated by the loudest form of voice: ‘whistleblowing’, defined as ‘the disclosure by organization members (former or current) of illegal, immoral or illegitimate practices under the control of their employers, to persons or organizations that may be able to effect action’ (Near and Miceli, 1985: 4). For individuals facing mundane (though nevertheless important) challenges in the ambiguous field of day-to-day clinical practice described by the likes of Bosk (2005), the world of ‘illegal, immoral or illegitimate practices’ may have limited relevance. Their decisions about whether to give voice to concerns will be governed by very different considerations, implicit or explicit. Furthermore, the tropes of gravity, personal sacrifice and irreversibility associated with whistleblowing cast a long shadow over all voice behaviours (Mannion and Davies, 2015), with the possible effect of suppressing voice in general.

The literature is clear, then, that any intervention to foster voice must seek to do more than merely provide greater opportunity to speak up, proffer administrative protections, or address only the visible aspects of organizational structures and cultures that inhibit voice. It must also be cognizant of the wider socio-cultural context that informs decisions about voice, and of the range of behaviours through which employees may raise concerns, short of the ‘nuclear option’ of public whistleblowing. All of these insights are relevant to understanding the institutional context in which the Freedom to Speak Up Guardian role must operate.

The Freedom to Speak Up Guardian

One of the most troubling features of the Mid Staffordshire disaster (and, as noted above, a significant driver of English healthcare policy since) was the revelation that while there was no shortage of information indicating that all was not well in the organization, those with access to this intelligence failed to act on it (Mid Staffordshire NHS Foundation Trust Public Inquiry, 2013). Characterized by some commentators (e.g. Newdick and Danbury, 2015) as a failure of voice, and by others (e.g. Jones and Kelly, 2014a) as a failure of listening, the implication either way was clear. At Mid Staffordshire – and perhaps in other healthcare organizations – systems and processes to ensure that informal knowledge from the ‘sharp end’ of care was communicated and acted upon were seriously flawed.

A subsequent review of speaking up in the English NHS (Francis, 2015) appeared to confirm these worries. Undertaken by the lawyer who had also led the independent and public inquiries into the Mid Staffordshire disaster, it found widespread reticence across the system around speaking up about quality of care, patient safety and colleagues’ behaviour, linked to a sense of futility and fear of retaliation (Francis, 2015). Although framed as ‘a review of whistleblowing in the NHS’ (Francis, 2015: 2), the report was also clear that it covered the full spectrum of voice behaviours ‘relevant to safety or the integrity of the system’ (Francis, 2015: 33). A survey undertaken for the review found that 30 per cent of those who had raised concerns felt unsafe afterwards, while 18 per cent of those who had not expressed a lack of faith in the system.

The recommendation to introduce the ‘Freedom to Speak Up Guardian’ role sought to address these challenges. The review described the Guardian as

Someone to whom staff can go, who is recognised as independent and impartial, has the authority to speak to anyone within or outside the trust, is expert in all aspects of raising and handling concerns, has the tenacity to ensure safety issues are addressed, and has dedicated time to perform this role. (Francis, 2015: 16)

Implementing the recommendation, the government mandated that each organization providing healthcare in England should appoint one or more Guardians (Department of Health, 2015), explicitly to act as a point of contact for anyone with ‘a concern about risk, malpractice or wrongdoing [they] think is harming the service’ (NHS Improvement, 2016: 4).

All NHS provider organizations were required to nominate a Guardian by October 2016; by June 2017, over 500 had been appointed (National Guardian’s Office, 2017a). The roles must be funded from organizations’ own resources, with Guardians’ efforts coordinated by an independent national officer (later dubbed the National Guardian), with a budget of around £950,000 per annum (National Guardian’s Office, 2017a). A survey conducted by the National Guardian’s Office (2017b) found that Guardians come from diverse backgrounds, and many maintain other roles alongside a part-time commitment to the Guardian role (36% as nurses, midwives or allied health professionals; 18% in corporate services; 5% as medical doctors).

Notably, neither the review nor the government response offered detailed specification of the appointment, responsibilities and accountabilities of the organizational role of the Freedom to Speak Up Guardian. Introducing the role, the Secretary of State for Health described the Guardian as a ‘member of staff to whom other members of staff can speak if they have concerns that they are not being listened to, [. . .] someone independent in their organisations to whom they could talk and raise their concerns’ (Hansard, Vol. 592, Cm. 782–793, 11 February 2015). In setting out how organizations should approach implementation, the regulator NHS Improvement (2016) similarly emphasised that individuals should usually approach Guardians if initial efforts to resolve concerns with line managers had failed. Guardians, it indicated, were to have ‘special responsibility and training in dealing with whistleblowing concerns’, offering access to other paths to voice inside and outside the organization (NHS Improvement, 2016: 9). Such descriptions posit the Guardian as a ‘second-line’ route to voice for staff with a concern about quality, safety and colleague behaviour, to be pursued where ‘first-line’ acts of voice have failed. The Guardian thus is intended to act as an ‘independent and impartial source of advice’ (NHS Improvement, 2016: 5), sitting aloof from the compromised complex of relationships and local histories that can make speaking up difficult – though as noted above, in practice Guardians often undertake this role alongside others (National Guardian’s Office, 2017b), perhaps rendering the enactment of independence more challenging.

To date, beyond descriptive overviews of the role provided by the National Guardian’s Office (National Guardian’s Office, 2017b, 2018b), no research on the Guardian role has been published. Moreover, a recent literature review found that the role is largely unprecedented, in healthcare or other industries – even sectors, such as banking, vulnerable to ‘illegal, immoral or illegitimate practices’ and corresponding whistleblowing activity (Jones et al., under review). It thus represents an innovative, but entirely untested, approach to facilitating voice in healthcare organizations.

Methods

We draw on data collected through a study of policy interventions to foster a culture of openness in the English NHS. We conducted semi-structured interviews with key individuals (including clinicians and administrators) involved in delivering these policies in English healthcare organizations, along with policymakers, representatives of regulatory bodies and individuals from relevant medico-legal and third-sector organizations, as part of a wider mixed-methods policy research study. Ethical approval was provided by the University of Leicester, and approval to undertake the study with NHS staff was provided by the Health Research Authority (reference no. 18/HRA/0084).

Participants were identified through a mixture of random, purposive and snowball sampling techniques. With a view to securing representativeness, we contacted a randomly generated selection of acute trusts (20), community and mental healthcare trusts (10) and ambulance trusts (5) in England to identify potential participants. In parallel, we purposively sampled four organizations that had experienced problems with openness, as indicated by regulatory intervention and/or media coverage. Finally, we asked participants to suggest colleagues within or beyond their organizations who might be able to offer insights relating to our research questions. Wider stakeholders – such as policymakers and representatives of regulatory, third-sector and medico-legal organizations – were identified purposively in consultation with a stakeholder reference group, with snowball sampling again supplementing this initial list.

Data collection occurred between July 2017 and January 2018. We interviewed 51 stakeholders in total: 18 participants from acute hospitals (denoted Ac in data excerpt attributions), 17 from community and mental healthcare trusts (MH) and three from ambulance trusts (Am), as well as 13 wider stakeholders (WS). The 38 NHS-employed participants included 10 Freedom to Speak Up Guardians of varying levels of seniority, mostly on a part-time basis as part of their wider (clinical or managerial) role, reflecting the profile of role-holders found by the National Guardian’s Office (2017b). The other 28 had been involved to various extents in determining the scope and position of the role in their organization, and in recruitment and management of Guardians, while some of the 13 wider stakeholders had been involved in development and oversight of the programme. Some job titles have been altered where they are particular to host organizations, and may be identifying. Interviews were guided by a topic guide based on a literature review and discussion among the authors, collaborators, the stakeholder reference group and a patient and public involvement group. The guide was intended to elicit participants’ in-depth understanding of relevant policies, including those designed to foster voice such as the Guardian role, the clarity and unity of direction these policies provided, the process of implementing them, and incentives and disincentives to increase openness. Interviews were undertaken by telephone at a time convenient to the participant. Interviews averaged 40 minutes and were audio-recorded and transcribed verbatim.

Supported by NVivo 11, our analytic approach was based on the constant comparative method (Charmaz, 2007). Interview transcripts were read independently by G.P.M. and S.C., who coded the data for high-level themes (broad areas of interest) derived from the evaluation brief and academic literature, and themes identified inductively from close reading of the data. Having coded interviews to these themes, we then re-read interview transcripts and compared data within and across themes, giving rise to more precisely defined themes as we modified, developed and amalgamated them. Coding was accompanied by ongoing discussion among the authors. G.P.M. then reviewed all codes relating to the Guardian role and undertook a further round of iterative, finer-grained coding based on this reading, which he used to develop an integrated analysis of the realization of the role. All three authors contributed to this analysis through further interrogation and iterative development of the themes presented below.

Findings

We organize our findings around three themes. First, we discuss healthcare organizations’ operationalization of the new role. Next, we examine the early experiences of Freedom to Speak Up Guardians themselves. We note that often the concerns voiced to Guardians differed in character from those anticipated in the role’s development. Finally, we consider the implications of how the Guardians organized and focused the role in practice for its potentials and limitations in helping to make concerns about quality and safety heard. We highlight that this approach to the role suggested a rather different notion of the nature and constitution of such concerns from implied by policy and by many organizations’ approaches to managing the role.

Designing and managing the Guardian role

Although some minimal guidance was available, leaders were charged with the task of designing a role for the Guardians that would work in their own organizations. Largely consistent with the role as set out by NHS Improvement (2016), they tended to cast the Guardians as a means of coordination, signposting and connectivity, with two important functions: first, raising awareness of the range of options available to those with concerns, and second, coordinating how those concerns were managed. Several participants saw the Guardian primarily as a signposter, charged with directing individuals with concerns to the right mechanisms and offices. Guardians and managers also constructed the role as crucial in ensuring that concerns were escalated to the correct level of the system, and securing an appropriate response:

It offers an avenue that’s never existed before to raise concerns. And [. . .] a number of issues, they’re doing quite a bit of facilitation and guiding and helping people, and also doing a bit of myth-busting as well. (Ac13: Director of Workforce)

People didn’t know where to go. It was either, ‘Well, shut up or go to the CQC [regulator]’, or the MP [member of parliament] or whatever. Which are two very different things to do. Whereas I’m able to not just talk through the options and find a bit of a context, but also go look at the middle ground [. . .] so that they have more escalation mechanisms open to them should they be dissatisfied with how it’s being dealt with. (Ac05: Guardian)

In a similar vein, Guardians were positioned as an independent and confidential point of contact, advising individuals who approached them of their options, assisting them in directing their concern to the appropriate authority, and identifying and addressing shortcomings in existing organizational systems:

It’s acted as another option for people to raise concerns. [. . .] The Freedom to Speak Up role gives you that opportunity to raise [an issue] and discuss it with someone one-to-one and face-to-face, who – unless it’s of a serious nature or a criminal investigation or ya-de-da – is able to protect that person’s anonymity. (Ac06: Guardian)

She follows them through the whole process, she follows through to make us account for responding to them properly. She will tell me if she’s got something about a service, she doesn’t think it’s right, and response isn’t right or the response isn’t happening. (MH10: Director of Nursing)

These formulations of the role were consistent with the broad policy goal of ensuring that ‘individuals are supported when they speak up’ and that ‘appropriate action is taken when an issue is brought to the attention of a Freedom to Speak Up Guardian’ (National Guardian’s Office, 2018a: 2–3).

At the same time, however, some managerial participants suggested that the boundaries of the role needed to be carefully specified and managed, not least so that it did not interfere with or undermine established channels for raising concerns. In many cases, these channels had developed in a relatively unplanned way, giving rise to a complex ‘ecosystem’ of routes to voice that were often tangled, ill-coordinated and unique to each organization. Within organizations, in some units or departments a variety of routes to voice flourished and co-existed, while other areas were barren. Providing signposting and assisting in cases where first-line mechanisms for voice had failed was one thing, but there was also concern that the new role would duplicate existing procedures or create new dysfunctions by adding yet another process to an already-complex ecology. Accordingly, integrating the role into incumbent infrastructures for employee voice posed a challenge in itself:

There’s a huge number of ways for people to raise concerns and complaints. (MH13: Deputy Chief Executive)

You looked at the outset, you thought, ‘Well how’s this going to work? How does it fit with all sorts of other things, your whistleblowing? How do you get somebody to be used in that way, to fulfil that role?’ (MH10: Director of Nursing)

Several participants sought to tightly bound the proper territory of Freedom to Speak Up Guardians in a way that mirrored official policy: ‘the raising of a concern relevant to safety or the integrity of the system’, to use Francis’s (2015: 33) phrasing. Similarly, some worried that the role could undermine cordial existing relationships between managers and staff where openness about concerns was already routine. They identified a risk of function creep, with the role straying beyond NHS Improvement’s (2016) emphasis on ‘risk, malpractice or wrongdoing’ (p. 4) and into the domain of routine line management processes:

My only concern is that it encourages people perhaps to side-line some of the other mechanisms by which they could resolve issues. So most of our Freedom to Speak Up reports at the moment are things that really should have been – they’re not really whistleblowing, if you know what I mean: they’re issues that really could have been resolved somewhere else. (Am04: Director of Organizational Development)

If you’ve got a full-time Guardian, I don’t know what they’d be doing. Without appearing to be disrespectful to that, I wonder if they’re touting for business. And that’s not the approach that we are taking here at all. We want our line managers to be able to respond appropriately. (Ac15: Associate Director of Governance)

Participants in executive or managerial roles accordingly sought to distinguish between what they saw as legitimate work for Freedom to Speak Up Guardians – helping those who approached them decide about whether and how to voice concerns, offering advocacy on particular issues, and raising awareness – and activity that strayed beyond this focus, and risked overriding existing mechanisms, ‘touting for business’, or finding or even creating problems where they did not exist:

If they’re sensible people doing a role – as long as you haven’t got people who are zealots, who are looking for problems where there aren’t some – I’m sure it’ll be a beneficial role. (WS08: Regulatory organization representative)

I think for me, in terms of the real serious whistleblowing issues, then I absolutely support Freedom to Speak Up Guardians’ role in that, because I do think that there’s – staff need that confidence. But I do think it runs the risk, like anything, of people deliberately circumventing. (Am04: Director of Organizational Development)

For those in leadership positions, this meant identifying where the Guardian role could flourish without interfering with functional processes or upsetting the balance of the existing ecosystem, and being quite clear about whether issues fell inside or outside the Guardian’s remit.

Concerns and their discontents: the role as practised

In practice, Freedom to Speak Up Guardians themselves found that the boundaries of their roles were not always easy to divine. For one thing, they had little control over the kinds of approach they received. Several Guardians interviewed found that many approaches could not readily be categorized as relating to service quality or patient safety. They instead concerned interpersonal relationships with colleagues, broadly consistent with the National Guardian’s Office (2017b) analysis of concerns raised across England, which found that only 32 per cent of reported cases included an element of quality or patient safety, but 45 per cent included an element of bullying or harassment:

They’ve been more about signposting, rather than safety issues: more about personal employment issues. I’ve had to signpost that. So none have come my way as whistleblowing. (MH05: Guardian)

I’ve had very few contacts in the 12, 13 months since my appointment. And when I have, it tends to be somebody disgruntled because of an individual employment issue. An HR issue. (MH16: Guardian)

Guardians indicated that they tended to support individuals who approached them with concerns that seemed to fall outside their remit, and were sometimes able to signpost to more appropriate routes through which to pursue such concerns. There was a sense among some managerial participants, however, that the role was being appropriated for purposes – particularly personnel and relationship issues – for which it was not really intended:

What we have got to work on is that people should be able to raise issues in their reporting line. But sadly not everybody feels they can. But at the moment my experience is that the real safety issues that people worry about, they haven’t so far used the Freedom to Speak Up Guardian. (Ac02: Chief Nurse)

For our organization, I would say that [the Guardian] has not had the contact that she was hoping she would get. [. . .] Most of the time it’s about signposting about other things rather than about freedom to speak up. (Ac01: Director of Nursing and Quality)

Some in leadership roles thus saw personnel-related matters (and other issues not directly related to quality, safety or other matters of wider concern) as ultra vires for the Guardians.

Several Guardians, however, discussed their experiences of hearing concerns that appeared on first sight to relate to interpersonal dynamics among team members or grievances against superiors, but which turned out to be symptomatic of deeper pathologies – with direct or indirect consequences for quality and safety. Some Guardians noted, for example, the prerequisite of an organizational culture that was open to challenge and in which staff were valued for high-quality care (cf. Sikka et al., 2015):

This isn’t really the patient safety stuff. But the reason I take things on like that myself is I want to do something about it. If we’re not following HR policies and procedures then we’re never going to shift the culture, because there’ll always be that criticism. And I just want people treated fairly. (Ac16: Guardian)

Similarly, some suggested that concerns that manifested as interpersonal or human resources-related issues might have direct consequences for quality and safety:

I have had quite a range. [. . .] People going for different positions within the trust and feeling that the job description wasn’t particularly accurate. That they hadn’t had relevant training for certain things. Issues, as you would expect, with harassment, bullying or discrimination. Issues with managers not allowing their staff to fully undertake the role, and as a result staff feeling a bit limited in what they can deliver for patients. (MH04: Guardian)

I don’t turn anyone away who’s suffering from bullying and harassment, because the trust has said they’ve got zero tolerance. [Executive director] wanted to eradicate bullying this year, which I think is a bit unlikely, unfortunately. But out of the concerns raised, it’s a balance between health and safety and bullying and harassment. (Ac09: Guardian)

More broadly, Guardians found that many of the issues that colleagues brought to them could not easily be placed into a clear, well-bounded category. What appeared to be specific, delimited concerns – whether about a practice or a policy or a person – often turned out to have deeper and knottier roots. People approached Guardians both when they were unsure where to turn, and when they had concerns that were too complex, sensitive or amorphous for existing mechanisms, with their well-specified remits and terms of reference:

A lot of the concerns that have come my way are not about what I would term simple safety issues. [. . .] They’re in the too-difficult box, or they’re in the ‘Who’s going to sort it out?’ box. (Ac17: Guardian)

I’m finding the stuff that comes my way tends to be stuff that doesn’t fit in boxes, because if it did then those people would know to go to their unions or HR or their manager. [. . .] I do pick up anomalies. But often it’s the anomalies where the big scandals come from. All this business about lessons learned et cetera et cetera, those things: the large scandals where there’s the national reviews, usually they pick on processes that weren’t fit for purpose. (Ac05: Guardian)

Consequently, Guardians found that they were tasked with making sense of and dealing with concerns that sometimes could not straightforwardly be dismissed, deflected to the right system or escalated to the appropriate authority. And this meant that acting only as a passive conduit for clearly defined concerns, with responsibilities neatly demarcated from other mechanisms for voice as the high-level policy blueprint and some of their colleagues envisaged, was not viable.

Unearthing roots: maximizing the role’s problem-sensing function

The most challenging – but also, potentially, the most valuable – component of some Guardians’ caseload thus turned out to not to be discrete issues that could be appropriately signposted or guided through to a satisfactory conclusion, but threads that led into wider concerns in which systems, structures and past events were also implicated. And this could mean that their job was as much about helping their colleagues to make sense of problems as sorting and directing them. When approached by individuals with issues whose shape and scope were not obvious, Guardians found that, whatever their formal terms of reference might suggest, further work was often necessary to make sense of the concern. This could often reveal more complicated issues than first apparent:

Because her role takes her to different wards, she was seeing the problems and thinking, ‘There’s something really bad going on here’. And one of the wards she particularly picked out, I wanted to see the matron, because I thought, ‘Well I’ll ask some questions’. I’m not supposed to investigate; I’m supposed to give things to people. But [. . .] I was fascinated what was going on. So I went to speak to the matron. [. . .] And immediately she was really defensive. [. . .] I sat with her for about probably an hour-and-a-quarter, hour-and-a-half in the end, and she said I’m the only person who’s listened to concerns that she has and staff have about the ward. (Ac6: Guardian)

Handling such concerns effectively, Guardians felt, required active work to trace the issues presented to them and situate them within their broader organizational context. It also implied a rather different approach to the Guardian’s position within the organization from that put forward by many of those responsible for implementing the role. We noted above that those in managerial roles tended to stress the need for clear demarcation from existing formal or informal routes to voice, and the policy blueprint positioned the role as an independent, second-line mechanism for concerns that had not been resolved through regular management processes. For Guardians themselves, however, a focus on independence and neutrality, and on maintaining a distance from the day-to-day commotion that faced people who might approach them, had important limitations:

There’s a lot of previous whistleblowers who are very dismissive of Freedom to Speak Up, and they think it’s a tick-box exercise. [. . . They] want the Freedom to Speak Up to come from outside the trust, completely independent, but I think you have an accessibility issue there. So if I had volunteers and staff who are on quite low [grades] raising concerns, I don’t think they would see some non-exec who sits outside the trust as being accessible to them. And the people who’ve whistleblown in the past about organizations, a lot of them are ex-consultants [attending physicians], who are very confident people. [. . .] I don’t think your average volunteer would do that, or someone who works in estates and facilities. (Ac09: Guardian)

For the more discrete, overt or egregious issues, Guardians acknowledged, independence and critical distance were important. But for concerns that were less clear-cut, or more embryonic, or where the concerned individuals were simply less powerful or less confident about what they were describing (cf. Martin et al., 2018), something different was needed. In many cases, they suggested, the Guardian role had adapted to fulfil that function. To elicit and make sense of such concerns, they found, necessitated a proximity that would facilitate trust and understanding, and encourage colleagues to take the chance to confide in Guardians:

There are pluses on that, because you do have those working relationships with the staff and the board, and we do get to see things from the inside. [. . .] I know I can walk onto a ward and know most of the staff by name, because our turnover’s quite low. It takes me sometimes quite a long time to go between meetings because you stop and have conversations in the corridor. So in terms of the approachability, I think it works. (Ac07: Guardian)

Contrary to the concerns of some managers that Guardians might intrude on others’ territory or generate concerns by ‘touting’ for business, Guardians felt that fulfilling their role effectively meant making themselves accessible and showing themselves to be interested and engaged in the difficulties their colleagues faced. Accordingly, they saw relational dynamics as more important than procedural clarity or cold neutrality. Achieving strong relationships with their colleagues meant immersing themselves in the challenges and compromises of the clinical sharp end, not restricting themselves to a formally demarcated position in the organizational structure:

What I didn’t want is to just be, in the perception of the people who are actually delivering the services, part of this perceived corporate bubble. [. . .] To speak up is to stick your head above the parapet and in order to do that, the person who you are raising the concern with, if you have seen them, if you have spoken to them or heard them speak or whatever, you have at least got that beginning of a rapport. [. . . You] cease to be either a person on the end of a phone, or this kind of faceless position. [. . .] I want to develop a rapport with people. The harder-to-reach people, you know, the BAME employees, LGBT, people with mental or physical issues, junior doctors. I have the time and the resources that enables me to target people who perhaps in previous times go under the radar. (MH04: Guardian)

Discussion

Employee voice is an important source of organizational intelligence, but effective systems for encouraging staff to speak up have remained elusive. The Freedom to Speak Up Guardian role represents a potentially promising innovation in addressing this challenge, but neither the review that recommended its creation (Francis, 2015), government policy (Department of Health, 2015) nor implementation guidance (NHS Improvement, 2016) specified exactly how it should be operationalized. Our findings show that those responsible at organizational level for implementing the policy were eager that the role’s place among existing channels for voice organizations was well-bounded. One reason for this careful boundary work was to avert the risk that it might inadvertently interfere with existing systems for speaking up, undermine other processes, or risk uncovering or even creating problems where they did not really exist. For Guardians themselves, however, such a neatly demarcated role – epitomized in the notion that they should act primarily as signposters or as a second-line mechanism when initial efforts to speak up had failed – was often estranged from the realities of the role as they put it into practice at the sharp end. The nature of the concerns brought to them frequently defied easy categorization; very few reached a threshold for whistleblowing, many were not obviously quality and safety-related, and some appeared to be signals of issues that were much more complex and wide-ranging than they first appeared.

Many parts of the Guardians’ caseload turned out not to be discrete concerns of the kind that could be readily guided through to a satisfactory conclusion, but instead were fragments of larger challenges whose totality could only initially be glimpsed. Consequently, much of their activity involved active work with those who approached them, to help them construct and make sense of the hints of issues they had sensed. The Guardians’ focus on the relational aspects of their work also reflects an understanding of the importance of going beyond rigid formalities, given the long shadow cast by past experiences of speaking up, and the durability of implicit assumptions about what should and should not be said (Detert and Edmondson, 2011). If the act of speaking up is not usually the result of a conscious, rational decision-making process, then efforts to understand and influence the influences that tacitly inform people’s sense of which acts of voice are appropriate will be at least as important as providing supplementary routes to voice or offering better coordination of the opportunities available. Equally, our findings show that the issues at stake in dilemmas about whether to speak do not always present themselves clearly as vivid ‘problems’ that warrant disclosure: rather, like the patient safety episodes that were contorted into the shape of incident reports in Waring’s (2009) study, they were more diffuse, ambiguous and subject to interpretation. Such episodes can of course be readily explained away or deflected, particularly if the alternative is to construct them as ‘concerns’ that must be formally voiced (Martin et al., 2018), but to do so risks missing their potential value and eroding the potential role of the Guardian as a safe haven for concerns. From this perspective, the value of the Guardian and similar roles may lie in their ability to assimilate, distil and sometimes augment the potentially valuable intelligence that resides in ambiguous and informal voice acts.

This in turn points towards issues with the concept of whistleblowing itself. Although charged with examining ‘the treatment of “whistleblowers” and their concerns’, Francis (2015: 4) took care to note that (public) whistleblowing usually results from the escalation of lower-key voice behaviours, consistent with research that highlights the spectrum of acts of voice between silence and whistleblowing (e.g. Jones and Kelly, 2014a; Mannion and Davies, 2015). Our findings, however, suggest that whistleblowing’s conceptual baggage goes beyond its association with grave, public and irreversible acts. It also assumes that the subjects of concern can be readily characterized as discrete activities that are evidently problematic, readily identifiable and probably deliberate, reflected, for example, in NHS Improvement’s (2016) reference to ‘concern about risk, malpractice or wrongdoing’ (p. 4). Near and Miceli’s (1985: 4) notion of the ‘disclosure [. . .] of illegal, immoral or illegitimate practices [. . .] to persons or organizations that may be able to effect action’ has, as Mannion et al. (2018) note, been largely taken for granted in the literature on employee voice within and beyond healthcare. This conceptualization of whistleblowing has evident relevance in fields such as banking, accountancy or insurance, where fraud, bribery and other deviant acts may be difficult to detect, but easy to categorize as wrongful. It is more problematic when applied to healthcare quality and safety, where – with conspicuous exceptions (e.g. Shipman and Inquiry, 2005) – catastrophic outcomes have their roots not in criminal acts, but often in decisions made in good faith in suboptimal circumstances. Where problems are more subtle, ambiguous or forgivable (Bosk, 2005), the decision to speak up may be equally momentous – but characterized by a different set of quandaries. In such situations, as the Guardians and some others we interviewed reported, independence or formal accountabilities may seem less important to potential speakers than the ability to understand and help make sense of the complex situations they faced.

This is not to suggest that acts of whistleblowing in the narrower sense of the term should be disregarded: on the contrary, whistleblowers’ actions have been crucial in exposing persistent problems of omission or commission in several systems (Bristol Royal Infirmary Inquiry, 2001; Cleary and Duke, 2019; Mid Staffordshire NHS Foundation Trust Public Inquiry, 2013). It is, however, to note that Guardians’ contribution in supporting whistleblowers (a key part of the role as intended in policy: see Department of Health, 2015; Francis, 2015) remains as yet unproven – and indeed appears from our study and from other evidence (cf. National Guardian’s Office, 2018b) to be a relatively small component of their workload.

For the future development of the Freedom to Speak Up Guardian role, and for other healthcare systems that might seek to devise a similar intervention, our findings suggest several lessons. First, work must be done to reimagine and re-present the work of raising concerns, so that it is infused less by the connotations of whistleblowing. Voice behaviours are gradated and plural, and the subjects covered by voice may go beyond issues that are easily recognizable as safety and quality concerns. Second, given both the range of forms that voice may take and the range of issues that it may cover, limiting the remit of Guardians or similar roles to more formal acts of voice about clearly defined issues may not be the most profitable way to deploy them. If problems of healthcare quality and safety tend to have their roots in misplaced optimism or tolerance of standards that slip through time (Macrae, 2014), rather than in wilful acts of commission or negligence, then attention to understanding and intervening in the localized cultures that produce these traits may be a more optimal focus. Third, this implies a need for a close and embedded – rather than distant and independent – relationship between Guardians or their equivalents and their colleagues: a role that is relational rather than procedural.

While this study has important strengths, including the wide-ranging backgrounds of the interviews, the breadth of the organizations involved and the novelty of the focus, it also has important limitations. Participants self-selected and had an interest in openness; they may not be representative of all organizations, and others may be more sceptical about the role. As an interview-based study, our analysis also rests on accounts presented by participants, which may be prone to issues such as social-desirability bias. The study also involved a relatively small number of Guardians (10); further research is required to interrogate the role, its potentials and limitations more fully.

Conclusion

The new Freedom to Speak Up Guardian role in the English NHS is rich in potential for addressing a persistent challenge for healthcare systems worldwide: inhibition of voice about quality and safety concerns. Our findings suggest, however, that to limit the role to supporting and advising those with concerns that can unambiguously be characterized as relating to quality and safety, or to cast Guardians primarily as signposters or coordinators, may not be the best way to exploit that potential. It was their relational work with colleagues with emergent, inchoate concerns that Guardians in this study found most fruitful, and it is this component of the role that may be most valuable for other healthcare systems looking to develop similar interventions to foster a culture in which employee voice may flourish.

Author biographies

Graham P Martin is Director of Research for THIS Institute, a unit funded by the Health Foundation at the University of Cambridge to develop the evidence base for and impact of work to improve healthcare quality and safety. His research focuses on social, organisational and professional issues in healthcare system change, with a particular focus on quality improvement work and policy initiatives to address quality, safety and risk in the NHS. He has published in journals such as Social Science & Medicine, Organization Studies, and the Journal of Health Services Research & Policy.

Sarah Chew is honorary research associate at the University of Leicester and works with the Sapphire research group (Social Science APPlied to Healthcare Improvement REsearch) which aims to improve the quality and safety of care in the NHS. Sarah has worked on a range of healthcare projects within the NHS. She is a qualitative researcher who specializes in ethnographic and interview data collection methods. As well as working in the field, Sarah also teaches social research methods.

Mary Dixon-Woods is Director of THIS Institute and is The Health Foundation Professor of Healthcare Improvement Studies in the Department of Public Health and Primary Care at the University of Cambridge. She leads a programme of research focused on healthcare improvement, healthcare ethics, and methodological innovation in studying healthcare. Her programme of research is concerned with generating a high quality evidence-base to support the organisation, quality and safety of care delivered to patients.