Effective board governance of safe care: a (theoretically underpinned) cross-sectioned examination of the breadth and depth of relationships through national quantitative surveys and in-depth qualitative case studies

Case selection

As we intended to consider the enactment of internal governance arrangements and the effects of external financial and regulatory mechanisms with reference to patient safety, each case study had as its centre of investigation an acute foundation hospital trust through which links between governance and patient safety were explored. We selected FTs because this organisational form was the policy direction for acute sector organisations at the time of planning the study. The unit of analysis is thus the FT rather than the wider health economy, with a focus on the operation of the board and its effects within the specific ‘tracer condition’ area of infection control.

It is important to acknowledge that FT boards operate within a national and regional context, and for this reason we undertook some scoping of context with key commissioners and stakeholders. However, our focus on the board as the seat of governance ensured a focus on the formal enactment of governance within board structures.

In addition to normal requirements of ethical approval, cases were selected on the basis of their performance trajectory over the last 3 years on a range of safety and quality indicators selected from the Dr Foster (Dr Foster Limited, London, UK) database. Every year Dr Foster publishes a hospital guide that uses publicly available statistics to measure and assess what is happening in hospitals in England to increase the transparency around variations in performance. The data focuses on indicators in three domains of quality obtained from the available statistics, combined with information from self-reporting of safety aspects from an annual questionnaire, and is informed by the National Patient Safety Agency organisational feedback reports and CQC ‘alert’ lists. We used these data to select case study sites that were indicated to be getting worse (in the light of overall improvement of other hospitals) or getting better, as follows:

• two foundation hospital trusts with a ‘downwards’ performance trajectory, as indicated above

• two foundation hospital trusts with an ‘upwards’ performance trajectory, as indicated above.

We then made sure that the sample was drawn from across the country and included large teaching hospitals (THs) and a district general hospital. Throughout the analysis and reporting, trust sites have been anonymised by being renamed after Scottish islands. The two upwards-performing trusts are identified as Lewis and Arran, and the two downwards-performing trusts are known as Islay and Skye. Appendix 2 contains the case study invitation letter and Appendix 3 contains the participant information sheet.

Brief description of the four selected case studies

Lewis is one of the bigger NHS trusts in the UK, delivering services across several sites within a single city. It offers many specialist services, with a large consultant body seen as being at the cutting edge of clinical care. Its long-standing reputation for high-quality clinical care is reflected in it consistently being in the upper quartile of hospital ratings.

Arran is a renowned TH providing general and specialist services across multiple sites. Over the last 5 years the board has been characterised by relative stability, with low levels of turnover. The trust aims to continue to build its ‘brand’ as a provider of leading specialist services, and it is undertaking large-scale redevelopment in order to deal with excess demand within accident and emergency (A&E) services. Organisational performance against the Hospital Standardised Mortality Ratio (HSMR) is very good, but the trust has faced some performance issues with regard to expressed staff satisfaction and, more recently, infection control rates.

Islay is an integrated provider of hospital, community and primary care services. It has a well-documented profile for its achievements in relation to quality and safety, with the board in particular receiving a number of awards and plaudits for these efforts. The trust holds high scores for both patient and staff satisfaction.

Skye is a district general hospital trust providing services across two sites. It was one of the earlier NHS trusts to be awarded foundation status. Over the past 5 years, the board has been characterised by continuous change, with the arrival and departure of several chief executive officers (CEOs), medical directors (MDs) and two nursing directors (NDs) during this period. The trust has one of the highest patient demands in the country and has faced a number of challenges related to bed capacity. Organisational performance is currently a cause for concern, with the HSMR above average, high A&E demand and increasing infection rates.

Data collection in the case studies

In the context of board governance, the focus of our data collection in the case studies was to explore the following areas:

1. to examine board of director oversight of patient safety

2. to understand the relationship between the board of directors and governors with respect to overseeing patient safety issues

3. to understand how the board complies with a particular patient safety issue (conformance)

4. to understand how the board responds to a specific patient safety issue during the course of fieldwork (performance)

5. to explore the commissioning relationships related to patient safety.

Data gathering consisted of interviews and observations at management boards and the governing board.

Exploring board oversight of patient safety

Our research carried out semistructured interviews with the board of directors at each site. This included the executive and NEDs as well as other board directors deemed relevant to the research and included directors of quality and risk. These interviews built on what was understood to be the key themes and issues related to the subject area17 (see Chapter 3) in examining how board members understood their role in terms of strategic vision and leadership activities, to understand how members understood particular board mechanisms associated with effective oversight, to understand how board members understood, analysed and interpreted the information they provided, and to understand how board members understood their relationship with the wider health economy in terms of commissioning and regulation associated with patient safety.

In addition to these interviews, overt non-participant observation was undertaken by the research team at four management board meetings at each case study site. This totalled nearly 50 hours of observation (Table 2). Descriptive free-text field notes were taken by both observers at each meeting, supplemented with documentary data including the agenda, supporting papers and (retrospectively on completion) the minutes of each meeting. Data included an overview of comments made and questions raised and by whom, and a log of the time spent on each agenda item. These have been summarised to provide an overview of patient safety issues within the context of all board activity. Data from field notes were compiled across multiple board meetings within each trust to detail board operations and identify the manner in which patient safety-related issues were discussed.

We supplemented our instrumentalist analysis of conformance and performance with a dramaturgical analysis of observation field notes following the framework developed by Hajer33,34 and previously used by members of the team in a study of partnership boards. 37 Reported in Chapter 7, this analysis explores the scripting, setting, staging and performance of patient safety governance, revealing the manner in which such work was conducted and the ability (or otherwise) of board members to ensure action.

Exploring board of governors’ oversight of patient safety

Interviews were undertaken with a variety of governors across the case study sites in order to understand their experiences of working with their board and holding the executive to account in relation to patient safety matters. These interviews asked governors to reflect on their abilities to exercise this function and also explore different ways in which the role of governors could be improved.

In addition to these interviews, overt non-participant observation was undertaken at nine board of governor meetings, totalling over 21 hours of observation (Table 3). This total number was less than the board of directors observation owing to practical difficulties, as board of governor meetings tended to take place only once every 3 months or so. Such a length of time between meetings meant it was difficult to capture four meetings at each site within the study time frame.

As with the board observations, descriptive free-text field notes were taken by both observers at each meeting, supplemented with documentary data including the agenda, supporting papers and (retrospectively on completion) the minutes of each meeting.

Following ‘tracer’ issues in the trusts

In each trust we followed two ‘tracer’ issues on patient safety from board to ward. One of these (called a ‘board conformance issue’) was the same in each trust and reflected a patient safety issue requirement for all trusts imposed by external regulators. The second (called a ‘board performance issue’) was different in each trust and reflected the board’s response to a local issue requiring performance improvement.

This approach was informed by the work of Garratt. 26 As discussed in Chapter 6, Garratt suggests that an effective board is able to balance tasks simultaneously, with a ‘learning board’ having two sides: board conformance (sometimes termed compliance) and board performance. Conformance itself involves two aspects: ‘ensuring accountability’ through appropriate regulation and ‘supervision of management’ through conformance to key performance indicators. Similarly, performance also involves two aspects: ‘policy formulation and foresight’ and ‘strategic thinking’ to drive the enterprise forward by allowing it to survive and grow by maintaining, learning and developing its position in its energy niches.

The tracer condition analysis looked to analyse the different ways in which these four board tasks (policy formulation, strategic thinking, supervision of management and ensuring accountability) interacted with each other.

Commissioning for patient safety

Where possible, interviews with commissioners were carried out to further understand the external environment in relation to the governance of patient safety. These interviews reflected on the experience of commissioners in their interactions with the FT boards regarding patient safety issues and concerns, as well as documenting any important issues regarding the governance processes associated with commissioning for patient safety.

Investigation of all of these issues entailed numerous interviews in each trust, as detailed in Table 4.

Manchester Patient Safety Framework

We had originally intended to administer the Manchester Patient Safety Framework (MaPSaF) to provide further supporting information on the maturity of board safety culture within each case study site. Designed as a diagnostic and development tool, the MaPSaF identifies dimensions related to working practices, such as risk management training, the investigation of patient safety incidents and staff education. However, during the course of fieldwork it became evident that most of the case study sites were undertaking assessments of patient safety culture commissioned by external third parties, and the decision was taken to avoid the risk of ‘culture assessment overload’ occasioned by further summary assessment.

Approach to qualitative data analysis

Our approach to qualitative data analysis proceeded through the three flows of activity, as outlined by Miles et al. 38 Here, a ‘data condensation’ process was used to select, focus, simplify and abstract data from field notes of observations and interviews transcripts. A thematic approach to qualitative data analysis was taken, whereby the data were coded, labelled and clustered into themes that both emerged from data as well as directly related to our research questions and theoretical frameworks influencing the study. 39

Assessing boards with the Board Self-Assessment Questionnaire

As interest has grown in understanding the effectiveness of boards, both inside and outside health care, a range of board assessment tools have been developed and applied. Most prominent among these, and a tool that has seen some use in health care, is the 65-item BSAQ. The BSAQ is derived from research highlighting the characteristics of effective non-profit governing boards in the USA. 42,43

The initial research on the BSAQ examined the practices of boards identified by a panel of experts on board development as either reputedly very effective or reputedly very ineffective. On the basis of this dichotomy of board development, the researchers isolated observable behaviours that were distinctive to the more effective boards and, using the critical incident technique as part of a qualitative study, identified six dimensions or competencies of effective board performance. 42 Following the qualitative phase, structured interviews with boards of trustees were used to aid the development of a self-administered 65-item questionnaire, in which each item is answered using a 4-point Likert-type scale. The BSAQ has subsequently been subject to extensive testing for validity, reliability and sensitivity, and this process confirmed that the six theoretically derived dimensions also had some empirical distinctiveness. 44,45

These six dimensions are labelled: contextual, educational, interpersonal, analytical, political and strategic (Box 1 provides more details). Four of these dimensions relate directly to Garratt’s instrumental board tasks (located in Table 1) and the remaining two (educational and interpersonal) are more behavioural, reflecting recognition of the need for boards to develop group cohesion, reflection and development.

Evaluating the Board Self-Assessment Questionnaire in the English NHS context

Because the BSAQ has seen only limited use in the UK context, we revalidated the data structure of the instrument using the data from English NHS boards. Factor analysis was performed at the respondent level to explore underlying factors that characterise boards. This analytic strategy is more fully explored in Chapter 8, in which the eigenvalues and item loadings are also presented.

Relating board-level variables to organisation-level quality and safety

After constructing the factor scores representing the six BSAQ dimensions and the total BSAQ score, we explored whether or not these were correlated with patient safety measures (although we did not explore a priori hypotheses). In doing so, we estimated various multivariate models regressing patient safety measures (from Dr Foster) and measures of hospital ability to handle errors, near-misses and incidents, taken from the National NHS Staff Survey (NSS) for 2012 on the total BSAQ score, controlling for a number of hospital-level characteristics. The full range of dependent variables are described and presented in Chapter 8 and more fully in Appendix 7.

A growing field of enquiry

The US Institute of Medicine’s reports, To Err Is Human: Building a Safer Health System49 and Crossing the Quality Chasm: A New Health System for the 21st Century,50 and the UK Chief Medical Officer’s report into learning from adverse events, An Organisation with a Memory,51 were hugely influential in calling for changes to health-care systems and organisations that would improve quality and safety. It is, therefore, unsurprising that our results show large and rapid growth since 2000 in the number of published articles that relate to hospital board oversight of quality and patient safety (Figure 1), a trend that reinforces the increased policy salience of board oversight of safety.

FIGURE 1.

Results of the systematic search of board oversight and patient safety.

The study of board oversight in relation to quality and patient safety can be situated within the broader literature that addresses the role of leadership in improving quality in US hospitals. 52–56 With variable successes achieved by hospitals in implementing quality improvement programmes and initiatives,57,58 the publication of empirical studies of US hospitals emerges from 2004 onwards, revealing variation in the adoption of board practices that are thought to be associated with higher performance and better patient outcomes (Table 6). 53,66 These cross-sectional surveys of predominantly non-profit hospitals indicate the importance of examining the leadership actions that are thought to influence the effectiveness of quality improvement activities in hospital settings. 71

From 2008, larger-scale studies covering wider geographical areas begin to feature. Building on the cross-sectional survey of the prevalence and impact of board activities in US hospitals,52 Jiang et al. 64 developed the scope of their evaluation of board oversight to include impact on clinical outcome measures such as mortality, morbidity and complications, as well as differences in the processes of care. Jha and Epstein57 carried out the first national survey of board chairpersons in the USA to analyse board engagement with clinical quality and identify differences between board activities in high- and low-performing organisations.

From 2010, studies have emerged that seek to differentiate and explore board activities related to patient safety in specific socioeconomic, organisational and geographical contexts. Jha and Epstein62 pursued an explicitly socioeconomic focus to compare boards of directors’ priorities and practices in serving the interests of minority-group patients. Prybil et al. 68 examined specific board structures, practices and cultures that relate to good governance in US non-profit community health systems. Baker et al. 59 carried out the first significant study of board governance and quality and safety in Canadian health-care organisations, and studies from the UK have emerged that analyse the formal governance arrangements for HCAIs and medication errors. 70 More recently, Jha and Epstein63 conducted the first national survey of English hospital board activities, providing an international comparison with their survey of non-profit acute care hospital boards in the USA. In addition, studies have also been published examining health board oversight of quality and safety in the Netherlands,72 Scotland73 and Australia. 74

The theoretical and conceptual dimensions of board oversight have also received more explicit attention. Jiang et al. 65 employ the agency theory perspective as a lens through which to explore the role and practices of hospital governing boards, while Ford-Eickoff et al. 75 explore how concepts such as complexity absorption and requisite variety can support hospital board governance and oversight, hypothesising that boards with greater variety and breadth of expertise among its membership can better respond to complex environments and have greater potential for sense-making and learning.

Emerging storylines

Our review of board oversight of patient safety identifies a variety of empirical evidence and expert advice that suggests that specific board activities are associated with improvement in the quality and safety of hospital care. However, the results also suggest that the adoption of such activities remains variable and that our understanding of boards’ impact on patient safety is currently limited. We present the findings thematically, as four storylines derived from the narrative review: (1) leading for safer care, (2) measuring safe care, (3) internal board oversight and (4) external regulation and accountability.

Leading for safer care

Board oversight of patient safety tends to reflect a key message from the quality improvement literature as a whole; that is, that strong and committed leadership from the CEO and the board is vital to the success of quality improvement and safety programmes. 56,76–79 A review by Clarke et al. 80 suggests that leadership on patient safety should learn from the characteristics and behaviours of high-reliability organisations, such as those found in the nuclear and aviation industries. In health care, leadership is associated with perceiving lapses in patient safety to be a problem of the system rather than individual employees, and with words and actions that promote a culture that encourages the identification of mistakes, emphasises system improvements that reduce variability and makes safety a given.

Empirical evidence from cross-sectional surveys in the USA suggests that boards that demonstrate such leadership have a positive impact on the safety performance of its organisations. Boards that place a high priority on quality and safety are associated with higher performance,64 as are boards that set strategic goals for quality improvement and demand reports on the progress of action in response to adverse events. 52,72,81 A US national survey of 722 chairpersons in 2007 and 2008 by Jha and Epstein found that respondents from high-performing non-profit hospitals were more likely than respondents from low-performing non-profit hospitals to establish and publicly disseminate goals and to perceive themselves as influential throughout the organisation. 57

Although such practices have been found to be associated with effective leadership on patient safety, empirical evidence suggests significant variation in the implementation. Drawing on the 2006 TGI (The Governance Institute, San Diego, CA, USA) survey of 562 chairpersons and CEOs from hospitals across the USA, Jiang et al. 52 found that fewer than half the CEOs regarded their organisations’ governing boards as very effective in overseeing quality. Similarly, Baker et al. 59 found in case studies of Canadian and US health-care organisations that, although most boards had established strategic goals for quality improvement, many did not have specific objectives with clearly defined targets, which suggests that words are not necessarily backed up by actions. Observational research by Machell et al. 15 of how nurse executives and boards work in acute care hospitals and mental health trusts across England found that many chief nursing officers (CNOs) perceived board members to be only moderately engaged in quality improvement initiatives. This was attributed to members’ lack of knowledge about quality and patient safety issues, the limited time available for participating in quality initiatives and to a lack of quality champions at board level. Such a view is echoed in qualitative research of non-profit hospitals in the USA. Interviews with CEOs and trustees by Grey and Weiss60 in 1998 and 1999 found that the two most important issues for local non-profit hospital boards in the New York area were mergers/acquisitions and financial management, with hospital quality and safety receiving far less attention. These findings are consistent with US national survey data from Jha and Epstein,57 which revealed that approximately half of the non-profit hospital boards did not rate quality of care as a top priority for board oversight or for CEO performance evaluation. Most boards were primarily focused on financial issues, assuming that its quality of care was adequate.

Measuring safe care

Board oversight of quality and patient safety rests on directors’ ability to obtain, process and interpret information, assess current performance and set strategic direction using a range of metrics that are tailored to local circumstances. A variety of expert advice has been produced to encourage boards to understand the big picture on quality and safety performance through the use of checklists and dashboards. 82–87 Reinertsen88 summarises the steps that trustees can take to manage quality in their organisations by focusing on a limited number of internal quality measures, or ‘dots’, and argues that such an approach can generate significant improvements, particularly in terms of building organisational buy-in, maintaining constancy of purpose and nurturing collaborative effort.

Empirical evidence from US hospital surveys suggests that boards that engage in reviewing and tracking its organisation’s performance through the collection and analysis of internally generated data (quality dashboards or scorecards) and national benchmarks are likely to have better outcomes in relation to quality. 52,64,65 This finding is supported by US research by Kroch et al. ,53 whose analysis of hospital board dashboard composition found that higher-performing organisations were more likely than their lower-performing counterparts to have dashboards that were shorter and more frequently reviewed and that focused on areas that are critical for quality. In England, single case study research conducted in 2008 by Ramsay et al. 70 found that scorecard data in a hospital offered a view of the organisation as a whole and facilitated division- and ward-specific analysis and feedback on the governance of HCAIs. Empirical studies and expert advice also emphasise, alongside the use of formal performance metrics and quantitative data sets, the role of soft intelligence in capturing the qualitative experience of patients and staff, which often defies simple coding and quantification. 59,89 In the USA, interviews with board chairpersons and CEOs by Joshi and Hines66 identified measurement of patient centredness as a key issue for board oversight of quality of care and recommended tapping into informal and soft intelligence channels as a key means for safeguarding care, including through executive walkabouts, patients telling their stories at board meetings and board members shadowing clinicians so that they might better understand frontline challenges in delivering safe care.

Patient safety metrics are clearly important in any strategy that is designed to improve care, but the empirical evidence suggests a need for considerable effort to increase board members’ proficiency in the use and interpretation of metrics and to improve the credibility, validity and reliability of data. Baker et al. 59 found that many Canadian health-care organisations had struggled to develop useful measurements for board oversight of quality and patient safety. Their survey of board chairpersons from hospital, as well as regional and community, organisations reported that, despite boards receiving and discussing a range of quantitative indicators, only half rated the information as good or excellent in enabling them accurately to assess its organisation’s performance. In England, research found that the information that was accessible to hospital boards generally fell short of the then NHS regulator’s recommended range of quantitative and qualitative material90 and that more needed to be done to cascade information about safety performance to frontline staff. 78 The recent survey of 132 chairpersons in England by Jha and Epstein63 also reported variable use of quality metrics among English hospital boards and found that boards of hospitals with foundation status were more likely than non-profit boards in the USA to use quality dashboards, request quality reports and review specific quality data as part of its oversight activities. Variation in the use of metrics by US boards is illustrated by the survey and scorecard analysis of Goeschel et al. ,61 which show significant differences in the use of scorecards across the participating hospitals in Michigan and Tennessee. In contrast, interview and documentary research by Prybil et al. 69 found a high degree of consistency in measurement, with 11 of 14 large, non-profit health systems in the USA formally adopting system-wide quality measures and standards.

Expert advice has sought to address the need for board members to have greater awareness of and understanding about quality and safety, recommending that quality expertise be included in board member competency profiles and suggesting that boards receive training and continuing education in areas to do with quality and safety. 90 Evans91 suggests that exams for board members on the use and implementation of quality measures could improve quality and accountability of care in hospitals.

Empirical studies have assessed the quality literacy of hospital boards by considering members’ participation in formal training programmes and setting time aside to develop knowledge and capability in relation to improving patient safety. 73,74 Some of this work has revealed the limited time and resources that many boards devote to such activities. US and Canadian research found a ‘remarkably low’ degree of knowledge among hospital boards about published quality reports and best-practice guidance in relation to safe care,66 with many board members having only limited expertise in using and implementing such information. 59 Formal training for boards on clinical quality also appears to be underdeveloped. 57 Reflecting on the findings of the 2006 TGI survey in the USA, Jiang et al. 64 suggest that the lack of formal training poses specific difficulties for board members who are drawn from sectors outside health care, who are less likely to have the technical skills that would enable them adequately to address clinical quality issues. More recently, comparative research by Jha and Epstein63 has found that chairpersons in NHS hospitals make a higher assessment of their own expertise in relation to quality than do their US counterparts. Moreover, the surveys suggest that board chairpersons’ assessments of their hospitals’ quality performance (in both England and the USA) are over-rated when compared with the external assessments undertaken by the Hospital Quality Alliance (in the USA) and the CQC (in England).

Internal board oversight

The role of hospital board members in overseeing quality of care has been characterised as being to monitor the strategic plans that senior management develop or to act as advisors and work at the periphery of the decision-making process. 75,92 Baker et al. 59 note that health boards historically tended to delegate the oversight function to medical staff and did not consider quality and safety issues to be its top priority, which might reflect a recognition on the part of board members of clinical leaders’ expertise and the traditional separation between the responsibilities of administrative and medical staff.

The increasing interest in board oversight of patient safety has also focused on formal organisational structures and processes for safeguarding care, as well as the informal relationships and dynamics between boards and professional groups. One aspect of this is board agendas and the extent to which patient safety is discussed at board meetings. 73,74 Findings from US hospital settings suggest that having quality and safety as standing items on the board agenda provides a critical lever for engagement in quality and safety issues. 57,66 Jiang et al. 52 found that most board meetings had agenda items on quality, but only 41% of boards indicated that they spent > 20% of meeting time on quality. Hospitals in which boards spent ≥ 20% of meeting time on quality had better process-of-care rates than hospitals in which boards spent less time on quality. 64 This research supports earlier findings on variability between hospitals in relation to the amount of time that boards give to quality and safety. In England, qualitative research from the Healthcare Commission90 found that safety is rarely the first item on the board agenda. This is supported by the observational research of English hospital boards by Machell et al. ,15 whose main conclusion was that considerations of clinical quality were accorded a low priority in boardrooms, compared with financial matters, organisational restructuring and the need to meet central government performance targets.

The formal structure of boards has also been found to be related to the effective oversight of patient safety. 93 US national survey data show that boards with a separate quality committee are more likely than those without to be high performing. High-performing organisations are more likely to use quality dashboards or scorecards, issue written policy throughout the organisation and establish strategic goals for quality improvement. 52,63,64 However, qualitative research into a hospital trust in England by Ramsay et al. 70 found differing opinions about the effectiveness of subcommittees. Although there was concern about potential duplication of messages, which might lead to mistakes in reporting, duplication of key messages was also seen as necessary to sustain staff engagement in safety-related issues.

Effective board oversight of safety also requires attention to be paid to dynamics and tensions within and between boards, medical staff and senior leaders. 94 In particular, expert advice advocates having a physician leader on a board quality committee to enhance performance by facilitating communication and building trust and confidence. 88,92,94 The empirical analysis of the 2007 TGI survey by Jiang et al. 64 found that hospitals in which representatives with clinical expertise served on the board quality committee performed significantly better on process and outcomes of care than did hospitals that had no such expertise on its boards. Collaborative retreats with multidisciplinary staff groups95 and internal collaboratives, built around safety initiatives, have also been found to enhance safeguarding processes. 55

Although such intergroup dynamics are associated with benefits, expert advice and empirical research highlight the need for board development in the area. Nursing leadership, for example, remains conspicuously absent from many board deliberations and decision-making. 15,67,96–99 Grey and Weiss60 note that further discussion is required about the ethical conflicts and ambiguities that can arise when board members combine professional and clinical interests with corporate roles and duties. Boards’ relationships with the wider patient community also need consideration. A survey of US non-profit hospitals by Jha and Epstein,57 conducted in 2008, suggested that board chairpersons in hospitals that predominantly serve black populations reported less expertise and prioritisation of quality issues than chairpersons of boards in non-black-serving hospitals.

External regulation and accountability

Our primary focus has been on internal processes for board oversight of patient safety, but we also considered research into the external accountability of boards. Case study research by Baker et al. 59 found that Canadian organisations grappled with the challenge of reconciling the information needed for external accountability with that required to inform local improvements. One dilemma was about whether or not to make performance data publicly available. Some hospitals had made available information on incidence of C. difficile, MRSA and vancomycin-resistant enterococci (VRE) infections, while others were approaching the issue more cautiously.

In the English NHS, a case study of a hospital by Ramsay et al. 70 considered the board’s accountability in relation to national targets and regulatory bodies for HCAIs and medication errors. Although the researchers could not categorically conclude that stronger external governance resulted in more effective local governance, they found that, at the time of data collection in 2008, rates of HCAIs such as MRSA and C. difficile infections were amenable to meta-regulation, for example through target setting. Medication error data based on incident reports were identified as being more difficult to govern, because such events tended to be more open to interpretation and to suffer from substantial under-reporting.

Qualitative research by the Healthcare Commission90 in England found that, despite the onus on commissioners or purchasers of services to drive up quality and safety through contracting and payment systems, there were local variations in the extent to which commissioners had robust information systems in place for holding hospitals to account for the quality of care that they provided. Commissioning practice in holding provider organisations to account varied across health economies from ad hoc requests to providers for reports of serious untoward incidents to the systematic benchmarking of providers against indicators of quality and safety. The research found that the most effective accountability arrangements between commissioners and providers were those that were supported through ‘relational contracting’ and which built on strong personal relationships and collaborative, rather than competitive, partnerships between organisations.

Strategic activity

In our study sample, all the trusts were greatly exercised by strategic concerns, but case study sites were highly differentiated in the robustness of its strategic approaches to the governance of patient safety. Skye was perhaps the least developed, lacking a coherent strategy because of extensive recent periods of turnover in executive board membership and beset with financial difficulties which, understandably, focused the attention of the board during the period of observation. Although there was much scope for development at this site, there was also cause for optimism given more recent board membership reconfiguration and the potential for strategic focus that this brings. In contrast, Islay had developed a patient safety and quality improvement strategy that was closely integrated into the business of the trust. The board saw themselves as quality and safety strategy pioneers; this was reflected in highly developed leadership activities (see below) as well as more strategically in its translation of strategic goals into summary indicators and reporting systems which were reviewed within board meetings. Therefore, although the trust was able to track its compliance with performance indicators related to patient safety monitored by regulatory agencies, this was done in the context of the quality monitoring and improvement activity required to meet internal strategic goals. This emphasis was reflected in the structure of the board meeting agenda, which was designed to support a focus on continuous improvement to meet internal strategic goals.

Lewis and Arran are both widely acknowledged as elite organisations, a significant proportion of their work involving the provision of specialist clinical services. Consequently, both have highly organised medical workforces who inform strategic developments within their fields. At Lewis, the overall strategic direction of the trust is built on a very strong CEO who effectively personifies the trust. The emphasis is on securing the legacy of the organisation as the local and regional provider of choice, and board meetings during the period of observation firmly focused on issues of estates and finance to support this legacy. Performance indicators of patient safety monitored by external regulators were routinely reported to the board. However, board discussions centred on the need to challenge targets set for reductions in infection rates and the associated penalties for failure to comply, rather than the service redesign required for compliance under the regulatory regime.

Strategic planning arrangements at Arran are undergoing development, moving away from emergent and evolutionary processes of strategic decision-making towards more formalised approaches, envisaged as an annual process. These more formal arrangements have developed in response to an acknowledged lack of team cohesion with regard to strategic direction; although the global aspirations of the trust were always clear, there was perhaps less clarity or shared understanding in relation to its strategic implications. These more formal arrangements are designed to increase coherence and shared understandings between staff, with the intention of securing actions to address identified difficulties. The trust has a highly detailed performance indicator reporting framework (see below) which is used to monitor performance in relation to external regulatory requirements.

Leadership activities

Given the highly turbulent nature of board membership over recent years, it is perhaps unsurprising that leadership at Skye is experienced as ‘staccato’. Board dynamics were acknowledged as troubled; although NDs were often vocal in their challenges to executives, this was not universally the case, with some concern expressed over possible scapegoating and the existence of ‘favourites’ who were challenged less often. The adversarial nature of the board was experienced negatively by some participants and was felt to foster defensiveness and covert behaviour, with the consequence of continued uncertainty over clinical practice. This feeling of unease was exemplified in the operation of the infection control strategy; although it had helped to identify and address a long-standing quality issue, concerns continued to be raised over the engagement of medical staff because of the frequent non-attendance of executives at infection control meetings, to the extent that, following a breach of Monitor targets, they were unaware of the issues. Limited staff engagement was also evident in planning to address the ‘bed crisis’, which was hampered by a lack of involvement of medical and nursing staff, both at consultant and ward staff levels.

In contrast, leadership of patient safety activity at Islay was clearly attributed to the board and principally to the CEO, who had ensured that quality improvement infrastructure and methodologies were embedded within the organisation to the point that coaching and these methodologies were ‘mainstream’ activities within the trust for all levels of staff. The board had been developed over a period of 10 years, with a stable membership and the support of external quality improvement specialists over the long term. Board dynamics were experienced as ‘supportively challenging’; that is, depersonalised and intended to stimulate further improvement in patient services. This dynamic was supported by charging NEDs with free-ranging scrutiny roles, rather than the more usual committee roles. High staff engagement and an improvement-oriented approach were exemplified in the trust’s response to C. difficile infection, which was characterised by division-based (local) targets to drive internal competition; charging the MD with responsibility for infection control; and the direct involvement of quality improvement personnel in planning reductions. Similarly, the use of ‘staff stories’ (an extension of the quality improvement approach of ‘patient stories’) of experience of care within the trust fostered staff engagement and trust that their insights and contributions were valued.

At Lewis, board leadership of patient safety, as with strategy, was dominated by the CEO, who was highly visible within the organisation and known for ‘unannounced’ inspections, an approach also favoured by other directors. Although the benefits of this approach were acknowledged, some concern was expressed over the limited scope for clinical presence within the board, especially in light of the level of responsibility given to the medical and nursing directors with regard to patient safety. The limited nature of clinical engagement in patient safety issues was identified within the infection control strategy implementation, particularly in relation to work at the interface of primary and secondary care. Although the monitoring of external infection control targets had driven interest in and support for infection control in some quarters, many senior clinicians remained unengaged, which was hampering improvement in C. difficile infection management. Similarly, some resistance had been noted to the medical appraisal systems that had been instituted in the wake of the Francis report. The concern was that without increased engagement, the trust risked sanctioning disrespectful behaviour to patients, with inevitable negative consequences for patient experience.

Arran has devolved considerable responsibility for patient safety into its clinical boards, each of which is supported by a single medical directorate for quality and safety that brings together activity across the clinical boards and reports upwards to the board. This structure reflects clinical leadership for patient safety and charges the board with oversight. Given this arrangement, board anxieties chiefly related to the extent to which staff feel able to report incidents within the devolved structure, the concern being the extent to which staff feel that it is safe to do so (i.e. that their concerns will be heard and acted on and that no negative reputational consequences will follow from expressing such concerns). This possibility is at least tacitly acknowledged by a number of respondents at various tiers within the organisation; cultural interventions concerning values have been made to address such fear and incorporated into staff appraisal systems to provide a means of raising and challenging such behaviours. Although these interventions have been welcomed, fears over the possible negative consequences of raising issues continue to be expressed, and there is evidence from junior doctors’ responses to the General Medical Council (GMC) survey of a fear of bullying by some senior medical staff. The main issue is that this constitutes a latent patient safety problem, with potentially damaging consequences for patients (and, indeed, the trust). Similarly, in relation to infection control, although the trust has a clear, well-resourced system and a specialist team to co-ordinate activity across clinical boards, fear of negative reputational damage associated with identifying problems is reported as hampering the reporting of issues. This has led to the development of specific practices in reporting of routine data to identify potential future problems and thus speak a language of ‘response required’ that can be heard at board level before problems arise.

Allocation of board time to patient safety

Observational and documentary data identified the time spent on patient safety, the supporting information provided and the nature of these discussions. Detailed summary tables are presented for each board meeting observed at each site in Appendix 8, and narratively summarised here. Across all four observed board meetings, Skye allocated 379 minutes out of 892 (42% of the total time) to matters directly concerned with service quality, patient safety, performance measurement and risk; considerably more than that allocated to strategy and capacity (185 minutes, 21% of time), human resources (HR) (79 minutes, 9% of time) and finance (55 minutes, 6% of time). Detailed supporting information was provided to board members within their briefing materials, including internal and external benchmarking data, line graphs of monthly and quarterly trends and ‘traffic-light’ indication of breaches of targets. This information was used by board members to frame their discussions, with much time being taken up in ascertaining the accuracy of the information presented and its implications for service delivery.

In contrast, the board at Lewis spent comparatively less of its time on matters of service quality, patient safety, performance measurement and risk. Of a total of 845 minutes over four board meetings, 236 minutes (28% of time) was spent on these matters. Although this was longer than other topics, time allocation was more evenly allocated between priorities, with 198 minutes (23% of time) allocated to estates, 158 minutes (19%) to capacity and strategy, 105 minutes (12%) to finance and 13 minutes (2%) to HR. Extensive documentation was provided in relation to quality and safety, as at Skye. However, in contrast, this was often not directly referred to in discussions or used to frame discussion/debate/challenge by directors; discussion was largely framed narratively by the CEO, with little direct recourse to the documentation provided. Indeed, some materials included within directors’ briefing materials were not discussed at all including the staff survey, patient-reported outcome measures, the National Institute for Health and Care Excellence (NICE) guidelines, quarterly monitor returns, a red–amber–green (RAG)-rated end-of-life care outcomes report, a safeguarding review in learning disabilities services, minutes from the complaints panel, pressure ulcers, patient falls and, on one occasion, the previous board minutes for approval. This reflects the more narrative style of the board at Lewis, and the dominance of the CEO in board meetings. The overall impression is one of a stark contrast with Skye, with hardly any reference to the different information provided in directors’ briefing packs and limited direct discussion of performance targets except concerning C. difficile infections, which focused mostly on the financial consequences of breaching compliance, rather than the quality improvement needed to ensure it. Many of the documents included in directors’ briefing packs but not discussed cover patient experience, human dimensions and safety, which only Islay discussed in any detail.

Board meetings at Arran were considerably shorter than at the other sites, owing in part to the continuation of a closed (private) part of the meeting to which the research team was not granted access. In total, the four board meetings lasted for 421 minutes. Most of this time was spent on quality, safety and performance (252 minutes, 60% of the total time), the vast majority of which was spent in relation to assurance activity against external performance targets and frameworks. In total, 83 minutes (20%) were spent on finance, 56 minutes (13%) on HR and only 30 minutes (7%) on strategy and capacity. All of the time allocated to strategy was spent during a single board meeting; typically, strategic discussions were not held within the public board meeting. Although a wide range of performance data were made available to directors, attention tended to be concentrated closely on those showing shortfall and the board had a very ‘assurance-heavy’ feel. Furthermore, there was little discussion of quality improvement activity in relation to these indicators.

Islay board meetings lasted 830 minutes in total. Of this, 448 minutes (54% of total time) was spent on quality, safety, performance management and risk. Of the remainder, 315 minutes (38%) was spent on strategy and capacity, 31 minutes (4%) on HR, 28 minutes (3%) on finance, and 8 minutes (1%) on estates. Board meetings were split into two halves, the first of which was open to the public. Much of the strategic discussion was done in the second (closed) half of the meeting because of commercial sensitivities; consequently, the two halves had a very different balance of material. In contrast to the other sites, directors discussed the human dimensions of patient safety explicitly, each meeting starting with a ‘patient story’ narrative to focus attention. A wide range of statistical data in addition to performance management indicators were discussed, including quality improvement data from run charts. Indeed, the emphasis was largely on quality improvement (improve performance further) than assurance (conform to standards).

Facilitators

The main facilitator at Islay was understood to be a clear strategic mission (5-year plan) focused on care quality, which had been translated and codified into performance measures and incorporated in an organisational quality dashboard. Objectives were clearly identified against the plan, with risk assessment of the likelihood of the delivery and devolution of responsibility for achievement to service areas. This ensured continuous focus on organisational objectives, notwithstanding the requirements of external performance management regimes. The strategic focus was buttressed by board dynamics characterised by ‘supportive challenge’; airing of issues was not interpreted as disloyalty and openness was valued (always subject to respectful framing) in order to maintain focus and momentum towards strategic objectives. At Lewis and Arran, the stability of board membership over time was considered to be an important facilitator of patient safety governance; the detailed knowledge and connection with staff of the CEO was accorded a high value. At Arran, the extent of clinical leadership was considered an important factor in board governance, notwithstanding persistent concerns (identified above) about the bullying of juniors by senior medical staff. In contrast, it was accepted at Skye that facilitators were relatively weak; the focus here was on optimism regarding the possible impact of recent board reconfigurations and the increased emphasis on care quality over finance evident in board meetings.

Barriers

All sites faced considerable barriers to the appropriate escalation of issues, that is, how best to act following identification of a potential concern. At Skye, this was primarily caused by disputes over the validity/reliability of summary indicator data (information deficit) and communication difficulties between clinical units and the board, and between board members (executives and non-executives), leading to a sense of confusion over the meaning of indicator data received at the board. At Lewis, the difficulties were mainly caused by perceived lack of engagement of senior medical staff, resulting in little communication between corporate teams and ward level staff in many parts of the organisation. Although reporting systems for major events at Arran were judged to be robust, respondents indicated concern that staff may not come forward with issues related to patient safety for fear of negative implications for their career; this was in relation to an acknowledged problem of bullying of junior medical staff to the extent that staff were not prepared to speak out. An additional issue at Arran concerned fears of oversensitivity to financial considerations, leading to poor decisions over safety, such as failure to provide adequate cover to meet a financial target exposing patients to unnecessary risk, rather than acting as a trigger for redesign work to make efficiency savings safely. Islay also manifested concerns over appropriate escalation at point of concern, in its failure to identify a hygiene problem within its kitchens (rather than directly in clinical services). This raises the prospect that, while its planning arrangements are well developed, it risks being caught out in services areas that do not feature within the planning schedule.

Commissioning

Relationships with commissioners at Islay were acknowledged by all to be excellent, and the trust was considered an exemplar provider in terms of the depth and subtlety of its documentation and analysis. The only reported difficulty concerned managing the reactions of other local providers who did not always believe their reported outcomes. Skye reported cordial relationships with commissioners and good involvement of commissioners in clinical governance committee work, although closer working was required over proposed reconfigurations to reduce secondary care demand that would have negative repercussions for the trust. Similarly, relationships with commissioners at Lewis were coloured by the trust’s ambitions for expansion and large capital development projects, in the context of proposed expansion of integrated care. Commissioners indicated that they found the trust to be aggressive with regards to defending its interests, yet had reached accommodation of the financial consequences of breaches in infection control targets.

Incentives

Commissioning for Quality and Innovation (CQUIN) was experienced at both Islay and Lewis as too unwieldy in operation to act as effective incentives, and negatively perceived as ‘taking money off’ rather than incentivising improvement activity. At Skye, CQUIN were perceived as having a positive impact on stroke pathway redesign, but a distraction from other areas that also required redesign work. Although compliance frameworks were generally welcomed for the effect on focusing attention on aspects of patient safety, concerns were raised in relation to the negative unintended consequences, such as reduced focus on areas not covered under the framework (e.g. environmental health). Penalties also proved difficult to apply at both Skye and Lewis, as they could have potentially bankrupted the former and led to bitter dispute with commissioners at the latter.

Identification and attribution of patient safety issues

Detailed reporting of summary performance indicator data related to patient safety is common to all of our sites, informed by the requirements of external regulatory regimes and, in some instances (Islay and Arran), considerably exceeding them in scope and scale. Although such summary reporting concentrates board members’ attention and allows breaches to be clearly indicated, a number of factors were identified which caution against over-reliance on such systems. Experience at Skye shows the importance of the perceived validity and reliability of the underlying data if action is to follow, otherwise time may be spent on debating what the data reveals, feeding tensions between board members rather than acting upon the implications of summary measures. Similarly, detailed observation of board activity at Arran showed that, even where there are clear breaches of targets in RAG summary indicators accepted as appropriate, it does not necessarily follow that the data will be interpreted as showing a patient safety failure requiring action; rather, the attribution ‘patient safety problem’ is dependent on acceptance that the breach really is a breach and not a misinterpretation. Respondents at Arran and Lewis also draw attention to the important role of qualitative data in contextualising and making subtle judgements in relation to summary data. Specifically, ‘green’ indicator ratings may conceal manifesting failures in risk management that will, if left unchecked, eventually result in ‘red’ figures. Summary reporting of ‘traffic lights’ on a single page of A4 paper may offer assurance that all is well, but such assurance will be illusory.

Escalation of concerns over patient safety

In addition to accurate identification of performance shortfalls, service improvement requires the engagement of staff to identify underlying problems and implement changes. Where staff engagement is lacking, initiatives may not embed or embed only partially; this is a particular problem where there are concerns about the bullying of staff. Escalation requires the existence of a reporting structure, conditions of trust and support between staff and a quality improvement orientation to address underlying problems.

Dynamics of whistleblowing

Despite the acknowledged importance of whistleblowing for preventing unsafe clinical practice, staff reporting such events face a dilemma where they fear negative repercussions for their career for doing so, placing the individual in the invidious position of ‘damned if you do, damned if you don’t’. Such fears are more likely to flourish in an environment in which such action is considered to be an act of disloyalty to the organisation or profession; this is especially the case within elite organisations where such action threatens the self-image of high performance. Where such practices exist, the fear is that formal policies in relation to whistleblowing will prove ineffective and opportunities for identifying safety issues before they manifest will be lost.

Tensions between assurance and improvement

Despite the availability of quality improvement infrastructure, three of our sites operationalised the governance of patient safety largely in terms of assurance, in which performance indicator systems were used to alert the board about poor performance in relation to specific clinical indicators ‘after the fact’. Major problems with this approach include that it may foster non-reporting to avoid the blame and shame that goes with underperformance, that disputes may arise over the reliability and validity of the indicators on which judgements of performance are made and that it does not provide guidance on how best to address the manifesting poor performance. In short, in seeking to identify problems it may paradoxically make it harder to report them. Islay sought to counter such negative unintended consequences through strategic focus on service improvement and embedding performance indicator reporting systems within quality improvement cycles, even to the extent of structuring the board agenda around strategic quality improvement.

Strategic vision

The Islay ‘success story’ was attributed to the fact that they had a clear a strategic vision ‘to be become one of the safest hospitals in the NHS’. Inspired by the CEO and ND, the strategic vision was triggered by ‘a road to Damascus moment’ during a visit in 2004 to the Institute of Healthcare Improvement (IHI) based in the USA. This subsequently led to the adoption of the IHI quality collaborative methodology in the trust, including signing up to the Saving 1000 Lives programme and the introduction of patient stories at board meetings. The board attributed its high performance in quality/safety in part to the ability to translate its mission into hard performance measures, which were summarised in a quality dashboard. The board’s 5-year strategy was linked to specific performance objectives that were disseminated across organisational divisions. These objectives were also subject to risk assessment to determine the extent to which they could be delivered:

We launched our quality improvement strategy 4 years ago and we were, as I say, ahead of the game for what that was worth . . . with clear aims and objectives, time frames, responsible officers . . . And we’ve demonstrated that time and time again.

Board members appeared confident that they had nurtured a strong patient safety culture across the organisation. Indeed, they were so confident about this that it was reported that external patient safety improvement initiatives, such as the implementation of recommendations for the Francis Inquiry, had little impact in changing its ongoing strategy. The introduction of a Nursing Assessment and Accreditation System (NAAS) in 2008, which was used to assess hospital ward-level quality and patient experience, was cited as an example of the organisation being ‘ahead of the curve’ in its efforts to improve quality and safety. The successful adoption of the quality improvement (QI) collaborative methodology, which was responsible for the elimination of high-grade pressure ulcers, was another reported example of the way in which continuous quality improvement methods were embedded within the organisation.

Despite these much vaunted successes, there remained scepticism among some board members as to whether or not the high performance could be attributed to the approach to quality improvement taken in the organisation. The director of development, for example, suggested that the performance was not the result of the quality improvement programmes but rather of wider factors and, in particular, of the income generated from a mixture of tariff and non-tariff services:

I think we’ve been lucky again that we haven’t had financial problems like a lot of other places. Now, part of that is down to good management . . . But part of it is down to luck. We’ve got a service profile and income profile that puts us in a good position, so we’ve a good mix of general hospital-type services and specialist services.

Board members believed a number of factors could stall the momentum that had been built up around quality improvement. The difficult economic environment was cited as a key factor, particularly the tensions associated with funding the ‘7-day week’ agenda. Hospital reconfiguration in the region was also viewed as a threat to ensuring safe care. It was reported that Islay was in a unique position to lead the necessary system changes across the local health economy, yet organisations did not always want to engage with them. Islay had a reputation ‘for wanting to take everything’ and there was a feeling among some board members that they needed to build better relationships with key stakeholders across the region, otherwise they risked isolation. It was believed that being isolated within the local health economy might have serious negative consequences for the organisation. For example, it was felt that if neuroscience services were to be moved to another organisation this would precipitate serious financial difficulties for the organisation.

Leadership activities

Successful board oversight of patient safety was often linked to the leadership style of the CEO and the ND, and in particular to their efforts to invest in quality improvement programmes and embed these across the organisation. The executive directors were described as an experienced team who worked well together and who believed collectively in the strategic approach being taken.

Central to board strategy was the focus on consulting and engaging staff at all levels of the hierarchy in key decisions. This was cited as key to them, being one of the organisations with the highest engagement scores in the NHS for two consecutive years. Specific leadership styles that were frequently mentioned related to ‘asking questions’, ‘getting the best out of that individual’ and obtaining ‘rich information from that individual’. The CEO described that what made Islay different from other organisations was that the board did not have a chief operating officer (COO). The rationale was that they wanted each executive director to have a specific domain of responsibility for a division of the organisation. The role of the executive was therefore cast as one of striving to encourage, advise, motivate and coach the management staff within their particular divisions. This close contact with staff lower down the organisation meant that the board’s vision became ‘something that everyone believed in’:

I firmly believe that if you continue to chase the dollar, you will be chasing the wrong thing. You need to chase the quality and safety, and what brings quality and safety for me is deep staff engagement, and that costs you not a penny.

Although staff engagement was noted as a particular success, there was also recognition among some board members that engagement policies could be variable. Despite achieving high NSS satisfaction scores, recent qualitative research undertaken by the trust had highlighted significant variability across the divisions in terms of staff feeling that they were a member of a team, that they could speak out if they saw something untoward and that managers supported them. This breakdown in board communication was illustrated by a poorly graded environmental health inspection which, it was reported, had exposed the limitations of staff engagement in the organisation:

We’ve got a bit of a less robust connection . . . and that was brought into stark relief by the experiences in the patient kitchen . . . I would say the engagement is more variable than it might first appear.

The existence of ‘closed’ and ‘hard-to-reach’ areas of the organisation was challenging. For example, engaging with staff working in one or two specific departments and in the operating theatres was raised as an area where there was thought to be ‘room for improvement’ in terms of embedding the desired behavioural and cultural changes and, in particular, inculcating new approaches to working among senior consultant medical staff. The NEDs also noted that the board did not have the same visibility in the community services that they ran as they had in the hospital and that they needed to think of different ways of better engaging staff in care homes, intermediate care providers and the private sector. The CEO also raised the challenge of developing better engagement with the public, as he felt the board had yet to develop an effective way of gathering patient, family and carer views about the services the trust provided.

Board meetings and mechanisms

Quality and safety issues were described as being the core business of the board meetings at Islay, with the first part of meetings given over to presentations about the progress of ongoing quality improvement activities and typically ‘over one-third’ of board meetings being spent discussing and debating quality and safety issues. A feature of board meetings at Islay was the participation of all divisional directors in the first part of board meetings. The purpose of involving them was to create an opportunity for the board to cascade messages down the organisation and to hold the divisional directors to account for the quality/safety performance of their divisions.

Islay was unique among the case study sites in having a ‘patient story’ at the start of each board meeting. The aim was to ‘set the tone’ for discussion and to focus the immediate attention of board members on service quality and patient safety matters. The ND had ‘championed’ this approach, but there were conflicting views about its usefulness, with the business director questioning the overly emotional connections that could be engendered by such stories. Criticism was also made regarding the lack of time available for the board members to discuss fully the implications of the patient’s story.

The fieldwork uncovered that, out of the almost 13 hours of board discussion observed across the four meetings, a total of 160 minutes was spent discussing patient safety issues (just over 20% of the total time). Compared with the other case study sites, there was less discussion about C. difficile infection issues and attainment of external performance targets. Patient safety discussions tended to be connected to wider quality improvement activities. The only patient safety-specific issues discussed at length were those relating to issues around mortality coding in orthopaedics (just under 30 minutes) and the environmental health inspection (50 minutes). The majority of time spent at the four board meetings observed tended to be focused on broader strategic issues, such as hospital reconfiguration, developing strategic partnerships and strengthening links with the community sector.

Governor relations

A number of board members described their support for governors as an effective means for FTs to engage their local community. It was also thought that governors provided ‘a voice of the people’, along with providing important intelligence about the organisation in terms of ‘what people are feeling and thinking’. NEDs talked about the additional education and training programmes they had developed for governors to help them in their role. The board chairperson described developing a close relationship with governors where he would feed back the outcomes of executive board meetings and give his ‘own honest opinion’ of what was happening:

I meet with the council of governors personally and in private, and I go through the board papers with them . . . I wouldn’t drop that meeting because they think it is so valuable to be able to ask tough questions privately and for me to be able to give an honest and open and sometimes a personal opinion about them. They get it unvarnished.

However, some board members suggested that the role of governors in the organisation was still not clear and questioned the competencies and capabilities of governors to successfully fulfil their role. It was also noted that the time needed to volunteer meant that governors were overwhelmingly of retirement age and that there was less representation from younger social groups in the population.

Governors described being engaged with the governance of patient safety at a number of levels. These included the Quality Subgroup, where governors were to able talk and challenge NEDs about the day-to-day operational issues related to quality and safety. Governors also described generally good relationships between the board and the council of governors. Governors felt that they were able hold the board to account through a variety of channels, including involvement in the chairperson’s appraisal and being able to actively challenge NEDs at the Quality Subgroup. The main issues of concern for governors related to the ‘steep learning curve’ and the time needed to fulfil the role effectively. It was thought by some that the ‘enforced’ election process and the potential removal of governors at the end of their 3-year term resulted in a ‘loss of key talent and organisational memory’.

Fieldwork observation of governor meetings supported many of the issues raised in the interviews. Governors rarely participated in discussions, which were dominated by the CEO and the chairperson. The focus of the discussions was one of education about particular issues (in this case hospital reconfiguration) in order to seek support from governors to continue the position the board was currently taking.

Assessing organisational performance

The organisation had been a forerunner in developing a quality assurance framework that combined national-level data regarding the HSMR and standardised health mortality indicators with additional local data, such as the NAAS, and Quality Improvement projects using Statistical Process Control run charts. Board members described triangulating these data with additional and softer intelligence obtained through walkabouts, board briefings (project managers presenting reports), and the complaints panel. Walkabouts, in particular, were highlighted as a good way to make the board visible and as an opportunity to engage with frontline staff and to respond immediately to their concerns:

I think that regular attendance on the shop floor gives us a level of information and knowledge that we otherwise wouldn’t get . . . and I think that that gets pulled back into the discussions that we have at board level.

Although the current performance information and quality indicator dashboard was in large part supported, comments were made, particularly by NEDs, calling for a greater narrative interpretation and analysis of the bold statistics and graphs provided and for a greater allowance of time for the board to ‘think about the implications’ of what the information was ‘saying’. The chairperson made the point that, although disaggregation of the data was the ultimate goal, this information needed to be presented in an accessible way:

People would probably like the graphs, but I couldn’t give two hoots about the graphs because words mean more to me. So to tell me what that actually means. Is it good, bad or indifferent? Have we gone up? Have we gone down? What is the trend?

. . . looking at these graphs are leaving us cold. We’ve been doing it for 3 years, we see eight data points above or below, and then it drops down, great, what does that tell us?

Frustrations were raised by the ND about NEDs still struggling with interpreting run charts, despite the board having used them since 2007. However, rather than more training, NEDs reported that they wanted more time to develop relationships within the organisation and also between fellow NEDs.

Relationships with commissioners

Board members described good relations with commissioners that have continued during the transition from the local PCT to the local Clinical Commissioning Group (CCG). The experience of the MD supported the move to CCGs in facilitating greater clinical involvement in commissioning decisions, and it was seen as an opportunity to have ‘mature’ discussions about the current economic environment. An example of this was the recent collaboration around unscheduled care, which brought organisations across the local health economy together in an attempt to save several million pounds. Alongside these positive experiences, some board members reported that commissioners were ‘suspicious of their success’ and that this led to ‘mistrust’ in negotiations. Frustrations were also expressed about commissioners being overly focused on hospital oversight at the expense of the wider health economy, with community care settings in particular described as ‘very loose’ in relation to patient safety:

I think there is a recognition across the health economy that we’re one of the ambitious organisations who are at the top . . . but I think that also brings with it a level of unhelpful scrutiny and unhelpful, what’s the word I’m looking for, mistrust. It can’t all be good. Trying to find a problem, if you like, which is irritating, to say the least.

Interviews with commissioners suggested that they felt ‘very fortunate’ to have such a high-performing trust. Islay’s strong emphasis on quality and safety meant that they tended to ‘shine out of the information’ that was being presented. Islay was seen as ‘slick’, with their success being attributed in large part to a very stable board and a very strong chief executive. Islay had been open and transparent in sharing performance information. This was supplemented with regular meetings and conversations about issues or incidents that were reported:

I get a lot of comparative data . . . compared to other organisations across the region, and there are a range of indicators where it is performing better . . . the mortality rates are low, infection control rates are much lower; so a range of indicators where it gives us a level of assurance compared to similar organisations. It’s doing very well.

The only ongoing issue specific to Islay was where rival providers thought the organisation’s performance was ‘too good to be true’. For example, it was reported that clinical quality managers in the health economy could not believe that Islay had failed to report any grade III or IV pressure ulcers. The recent ‘absolutely excellent’ CQC inspection rating was cited as a case in point, as Islay had been inundated with requests nationally for help but there were few requests from within the local health economy. To improve relationships, the CCG was working with the MD and ND of Islay to try and encourage local general practitioners (GPs) and its own medical consultants to collaborate around a quality improvement programme. The goal was for both parties to use the same incident reporting system to encourage dialogue and the sharing of information between primary and secondary care about incidents.

Board members reported that the CQUIN mechanisms being used were not really impacting on service delivery as much as expected. Interviews with commissioners raised similar issues for example, although the CQUIN for dementia services was seen to have had an impact in terms of raising awareness, greater local ownership was needed in order to maximise impact. CQUIN targets, it was thought, were in danger of becoming ‘a bit of a tick box’ as a way of securing finance, but with little long-term service impact. As a result, the CCG was trying to take a more collaborative approach in identifying areas where they wanted to try and secure improvement. Recent examples of this were the agreement of two CQUINs around developing 7-day working in surgery and acute medicine. There appeared to be a shared interest for both parties for quality improvements in this area:

I do believe it needs to be done and in a collaborative way. It’s not good us as Commissioners saying, ‘we want you to improve by X in area Y’ if they haven’t got the appetite to do that.

Relating to regulation

The Monitor compliance framework was supported by board members in terms of ‘focusing the mind’ in relation to key performance areas. The introduction of infection control targets exemplified this, with the development of targets for MSRA and C. difficile infections seen as changing the way infection control was governed. The CQC criteria were also reported as being helpful in supporting the board to identify problems related to cleanliness and infection control:

You’ve got to take that as being a success. You know, there’s no other way of saying that . . . when we started with MRSA, we were probably having 80 [cases] a year . . . [but] we haven’t had any this year . . . So that, in part, has to be the fact that targets have been set.

The first thing I look at in the board pack is the infection control! And then I look at all the data, you know. So I think we are incredibly mindful of that.

However, the point was made that the experience of Mid Staffordshire NHS Trust highlighted the unintended consequences of focusing on performance indicators as an end in themselves. The recent case of an environmental health breach was use to illustrate where those service areas (kitchens) not covered by Monitor or CQC compliance frameworks often went unnoticed. On this basis, it was reported that the board needed to be ‘brave’ in balancing the interests of Monitor and the wider interests of the organisation and patients. The ND raised the point that central targets needed to be backed up with a supportive local culture and that the leadership of the organisation needed to be focused on quality improvement. The Islay board was therefore seen as good at combining a locally driven agenda on quality alongside the national agenda and centrally set performance standards.

CQC inspections are getting more and more focused on the minutia of the documentation, and yet you know you could have the entire documentation perfect, but the patients could be not being treated kindly.

Strategic vision

Interviews with board members at Skye suggested that the continuous change of executive board members in recent years had prevented the development of an effective strategy for patient safety. In what was described as a ‘jigsaw puzzle of piecemeal activities’, bed capacity was cited as a case in point where previous board decisions to remove beds were linked to a variety of problems facing the trust, including poor A&E performance, poor C. difficile infection management rates, high pressure ulcer rates and a demoralised workforce:

If there’d been someone in the executive with a longer term sort of organisational memory, they might have put some connections together.

Points were also raised about previous board configurations that promoted cost control over quality improvement. The dominant influence of the finance director was seen as a feature during this time, described as ‘a 100% focus on short-term financial issues’. One example of finance seen as overriding patient safety decisions related to the protracted attempt by the IPC committee to purchase a fogging machine to improve deep cleaning. Vacancies, recruitment and pathway redesign were also issues where finance would challenge progress being made.

Criticisms regarding the lack of strategy were also directed at the current board. The director of HR described it as ‘a very functional board’, with no deeper discussions about the evolution of service delivery. The style of the chairperson was also mentioned as a contributory factor, being described as ‘more hands-on than strategic’. An example of this was the proposed ‘quality day’, which was a response to pleas for more dedicated time for discussion. Although this initiative was supported, concerns were also raised about the style of the chairperson, who was described as trying to ‘cram everything in’ and focus on targets rather than developing more informal open discussions about strategy and moving forward:

I had originally thought oh good, because that means we’ll be able to stand back and look at our organisational purpose . . . Actually what it’s turned out to be now is a list of things that just looks like a long board day, so we’re going to talk about C. diff . . . it’s probably partly because the chair doesn’t like blank spaces on agendas.

Despite this lack of strategic thinking, notable improvements were connected to the latest board configuration. The appointment of a new director of finance and a new CEO were seen as improving relations within the organisation and with commissioners. The dynamic was also improving in light of the Francis Report, which was switching the board on to quality. The recent bed plan exemplified how the clinical drivers outweighed the finance issue, as decisions were taken ‘before the financial answer had been found’.

Leadership activities

The continuous turnover of board members at Skye had created a form of ‘staccato leadership’, characterised by interruptions and hiatuses. The new CEO was a case in point: it took him time to find his feet, and this ‘created gaps’, which meant the chairperson had ‘to do a bit more leading than might be expected’. Despite this turbulence, the board was aware that they needed to set the tone in promoting a culture of openness and transparency. Being visible was central to these activities, which were exemplified by the chairperson holding regular meetings with large groups of senior nurses to talk to them about the key issues of the moment. Executive and non-executive involvement in a variety of governance committees, including the mortality review group and the ‘safer care’ group within the clinical governance subcommittee, provided further examples of leadership activities.

It’s the visibility of the leadership I think is the key thing. And I mean they go on to wards to basically look around and talk to staff and actually say, ‘I’m a non-executive director, do you want to raise anything? What are the issues that you’re concerned about?’ So I think on an individual level, ours are very good.

Board leadership was also associated with challenge by creating an environment where certain behaviours were not tolerated. This was particularly the case for NEDs, who were noted for their vocal challenge in holding the executive to account. For example, the acting ND, new to the position, reflected on the difficulties of responding to the challenge from NEDs. Being in an acting position, she noted, ‘made it a hugely vulnerable and exposed place to be’ as the NEDs attacked her inexperience. NEDs very much believed that challenging the executive was key to their role; however, the point was made that they struggled to hold the MD to account because they lacked the necessary clinical knowledge. ‘Second-guessing clinicians’ was an ongoing struggle; therefore, having greater clinical knowledge in the boardroom was seen as necessary. The chairperson was particularly frustrated at holding individual clinicians to account, and described the disproportionate amount of trust given to them as particularly unhelpful as the board sought to understand and investigate any quality and safety issues. Things were changing, with the board now doing more work via the clinical directors who could then provide a peer-to-peer approach to scrutinising clinical practice.

It surprised me when I first got involved in the health sector that pretty serious things can go wrong, and accountability doesn’t seem to be as strong as it would be in other industries. A ‘never’ event should be a ‘never’ event, and people who were involved it, I would take a much stronger line than it appears the health, well the NHS certainly take at the moment.

Issues were also raised about the boards’ ability to engage staff. The board was improving the engagement mechanisms with notable updates to appraisals as well as the introduction of weekly briefings to promote communication; however, the point was raised by a NED that staff were feeling ‘underinformed’ about what was happening:

We don’t generally engage with staff too well. You can see from the survey results . . . [that] a theme running through this is the performance appraisal. There were people who’d do it in groups, so the ward manager would just do all the nurses together. I mean, absolutely dreadful, and that was if it was done [at all].

Board meetings and mechanisms

Although some board members, notably the chairperson, described how meetings over recent years had become more focused on quality and patient care, others were frustrated with the way the agenda was covered, with the chairperson seen as being driven by action plans and timescales. The chairperson, in turn, was aware of these frustrations; however, he believed that the right approach was being taken, and described how a recent action plan related to mortality was being delivered:

I could understand the frustrations on some of my colleagues . . . I’m a bit more in the ‘doing’ rather than the ‘talking’ camp here. You know, we can talk about things for forever and a day . . . but there is a time when the talking has to stop and the doing has to start.

The observations of board meetings highlighted that out of almost 15 hours of discussion across four board meetings, 4.5 hours of the discussion related directly to patient safety issues (30% of the total time). Most of this time was spent on the mortality review across patient pathways (95 minutes) followed by discussions about the breach of C. difficile infection rate performance targets and the resulting financial penalty (80 minutes). Other issues included discussions regarding performance related to patient falls, complaints and pressure ulcers, along with policy updates concerning patient safety, such as the CEO’s summary of the Francis Inquiry (which in in one board meeting lasted for around 30 minutes).

Governor relations

Board members had mixed opinions about the role of governors. On the one hand, the board understood the need to have governors holding NEDs to account and providing a voice for patients, the public and staff. The trust secretary described how the board proactively worked with governors in holding quarterly time-out sessions to brief them on any current issues, and NEDs suggested that they did feel that governors were holding them to account through being involved in their performance appraisal, as well as through governor observations of their performance in chairing different subcommittees.

However, other board members pointed to limitations of the governor role in the (perceived as poor) quality of their contribution to patient safety walk-rounds, productive wards and various other initiatives. One senior executive suggested that governors did not properly hold the board to account, and the quarterly time-out sessions were seen as a ‘bit of a hassle’. Despite well-meaning questions, the capability of governors was seen as mixed and as often diluted by personal agendas, pet topics and minor issues:

We can give a 5-minute spiel about something and talk with intent, hopefully. And people think, ‘Oh, that’s okay,’ and then you get another quarter.

There are some who are not confident to speak up and we try and encourage them but they’re not confident at all. There’s some who are very confident but actually kind of miss the point or misunderstand their role.

Interviews with governors described how they supported the latest board efforts, particularly in helping the chairperson and CEO to respond to patient safety issues. They were also impressed with the NEDs describing their observation of subcommittees where (for example) infection control reports would ‘get some hammer’. This was in contrast with the previous board, which was described as being ‘very entrenched in the old way’, that had ‘not got a grip’ of patient safety. The current board also seemed far more receptive to governors. This was evidenced by greater strategic involvement included drawing up the annual report, sitting on the Clinical Governance Committee, getting access to ‘never event’ explanations, setting the agenda of meetings, going on walkabouts and providing an opportunity to raise issues from a patient perspective.

Having spoken to other governors from other organisations, we’re better off here . . . We participate to a greater extent than any of the others I know I’ve met. And we’re not just a talking shop. We don’t just have governors’ meetings where we talk about things then go away and forget about them.

Despite this, there were still some issues with the current board. As a relatively inexperienced board, some felt that they needed to develop better relationships with the organisation and be more open to staff feedback about service improvements. Concerns were also raised about the quality reports that they received having inaccurate information, particularly in relation to finance. Furthermore, issues were also raised related to the governor role itself and the time needed to develop this properly. The first 3 years following election were seen as a very steep learning curve and it was often only in the second term that governors felt that they had real influence.

Observation of governor meetings found that the majority of discussions centred on governor education about the latest policy developments and the provision of summary performance information. As with the other sites, dialogue was mostly led by the CEO, who would summarise his report, providing a number of updates regarding policy. There were also PowerPoint^® (Microsoft Corporation, Redmond, WA, USA) presentations by the CEO about the Francis Inquiry, along with the COO, who presented a business intelligence report detailing the key performance indicators related to activity and infection control. The result of this was that there were limited contributions from the other board members, apart from the chairperson. The chairperson was, however, influential in encouraging governors to question the business intelligence report. The question and answer sessions in response to this report did tend to be active and, in one meeting, this led to a short but significant discussion that challenged the COO about poor C. difficile infection management performance. Points were made about performance being ‘not good enough’. At these times the MD and ND would enter the conversation, allaying concerns and suggesting that things were being done to improve the situation.

Assessing organisational performance

Board member interaction and analysis of performance information was, again, mixed. Some positive aspects were raised, particularly by the trust secretary, about the quality dashboard and how the board was ‘ahead of the game’ in some of the information it was collecting related to HR and the disaggregation of the HSMR broken down on a monthly basis. However, the stronger opinion, particularly from NEDs and the COO, described the current benchmarking as poor, with mistakes in the reports often being found. More work was needed to improve the methodology, as the accuracy of the data was currently being questioned.

I think the business intelligence report isn’t quality controlled brilliantly well. I’m fed up with spotting reds that should be green . . . that needs to be tightened.

Patient safety performance was seen as worrying, with both the HSMR and the standardised health mortality indicators being above average. One of the ways the board was looking to respond to these trends was by paying particular attention to the way that different directorates input information into the Corporate Risk Register. The CEO raised the point that there was currently some disconnect between the information systems used by the directorates and that used by the board, with directorates lacking a consistent structure, leading to people measuring things in different ways. This meant that when issues escalated up to board levels, not all staff were receiving consistent messages. The trust secretary described how directorates viewed the Corporate Risk Register as ticking the box rather than as an opportunity to raise any issues they had with the data. In looking to resolve the disconnect between central support units and board-level measurement systems, the MD and COO at the time of writing were undertaking a pathway redesign that was looking to standardise performance measures across central support units, which will then be reported to the executive team.

Alongside the review of quantitative data, board members described how they gathered soft intelligence through the review of patient complaints, the carrying out of mock CQC assessments and walkabouts. Board briefings provided opportunities for people to be invited or to offer a presentation to the board. A recent example of this was bringing people in who were involved in an investigation of a recent serious incident. Having a patient come to speak to the board was suggested, as well as having a commissioner on the board providing the ‘client voice’. A note of caution was raised, however, that there was a danger that board business could become diluted, with other people coming in and taking over the agenda: there are certain things that have got to be done. Connected to this point the ND, with the support of the CEO, was looking to introduce patient stories to board meetings as a way for the board to interact with more qualitative material at board level. Despite her enthusiasm, fellow board members decided that it was not appropriate. Reasons for opposition included:

It ends up with ‘Woe is me’ and we end up spending 30 minutes to three-quarters-of-an-hour on a patient’s story and it didn’t actually make any difference to the patient, right? And the one thing that concerned me through all this was: ‘Where is the learning? Where is the continuous improvement? Where is the will to not make that mistake again? Where is the will to plug that hole and say, that will not happen again?’.

The touchy-feely stuff . . . Yes. I’m a surgeon by trade, so I’m not good on touchy-feely.

In Skye, NEDs raised a number of personal challenges in relation to measuring performance. As it often took up to 6 months to get a handle on the job, understanding and processing the volume of information was an ongoing challenge. Walkabouts were mentioned as a good way for NEDs to triangulate the information and yet they were also problematic for NEDs, as they were not necessarily prepared with enough information about the area they were going into. Despite these concerns, other NEDs believed that they were starting to tap into the organisational realities. A NED, who was female, stressed the need for a greater female presence on the board, particularly when it came to connecting with female staff. In this sense, rather than further statistical training in this area, more attention should be dedicated to developing relationships and managing situations:

Actually seeing with your own eyes is difficult. The difficulty I think, that it becomes a box-ticking exercise, that you’re looking at a list which somebody’s prepared for you and you tick the boxes and say yeah, it’s really fine.

Relationships with the commissioners

Relationships with commissioners were described as cordial and constructive in working through the current pressures in the system. Despite this, there were calls for the board to have a closer relationship with the CCGs, particularly in relation to what commissioners were doing in the community to help reduce hospital demand. Having a commissioner as a NED on the board or attending regularly was tabled as an option to facilitate further communication and interaction. The COO also called for commissioners to spend more time in the hospital, particularly to see how A&E worked:

[we need to get] closer to these guys and we’ve got to, you know, have sometimes some difficult conversations like, ‘Hey guys, you’re throwing everybody into A&E because you’re not really making yourselves available to people, so come on, let’s play fair here’ . . . we need to have a chat about this because this is not how we need to do things.

The arrival of the new CE improved communication, with regular meetings now taking place. Furthermore, the board had recently witnessed a ‘very sensible approach’ by commissioners to support the investment of a fogging machine to encourage the prevention and control of C. difficile infections. Such a ‘joined-up approach’ was linked to the Francis Inquiry recommendations, to which commissioners were now looking for a more transparent and collegiate relationship with provider organisations.

Interviews with commissioners revealed positive relations with Skye. Since Skye had become a FT, commissioners commented on how progress had been made in terms of openness and communication about patient safety issues. The relationship was now at a point where it felt more like a partnership. This was exemplified by commissioners recently being invited to attend and contribute to their Clinical Governors Standards Committee and attend their mortality review group:

If we turn the clock back 6 years, for example, we weren’t allowed to be in any of their groups. They used to occasionally invite our medical director – our clinical governance leader was a GP – to the odd one, but they basically didn’t invite us as a matter of course.

A key driver for this openness was the events at Mid Staffordshire NHS Foundation Trust, which had driven a realisation that they needed to be open. Commissioners had witnessed the trust become more proactive at handling, reflecting on and reporting serious incidents. Commissioners had been impressed with the NED who was chairing the Clinical Governors’ Standards Committee in the way he challenged the committee and called for a drive to improve standards. Commissioners had also seen the trust become more open to collaboration in working together on patient pathways:

From our perspective, the fact that they then reported 4 months late, we have to take a good view of that. We have to say, ‘Well okay, you should have picked it up at the time but are we going to flog you for it when you’ve actually been open and said we’re reporting it now?’

Although it was reported that Skye had a positive baseline with high patient satisfaction, there were also ongoing problems with several issues, such as the management of the emergency care pathway, clinical rota covers and admissions at the weekend. This meant that the trust were in a high-risk category for quality and safety. This was likely to receive ongoing attention; however, commissioners had been working with the trust to resolve these issues. In doing so, commissioners were pleased and supportive of the way the trust were responding, particularly concerning urgent admissions:

We have to make sure we stick by the line with them, that they’ve done what we asked them to do, and if we didn’t then I think that would be a bit of a breach of trust between the two organisations because they’ve really tried their best on it.

Interviews with board members recognised that CQUIN payments did provide a lever by which commissioners could encourage certain behaviours and practices within the trust. The use of CQUIN to inform the stroke pathway was a good example of CQUIN being used as a carrot to encourage best practice. The CEO also suggested that the CQUIN target for a needs assessment for dementia patients was ‘absolutely the right thing to do’ and was forming part of the board’s dementia strategy. That said, the director of risk suggested that CQUIN was also presented as a double-edged sword: it was good to have such payments, but it could provide a distraction from other activities. Further issues cited included the point that any payments went into a trust ‘black hole’ rather than to the teams and departments who were delivering. As such, it became all sanction and no reward, which led some staff to become apathetic.

For commissioners, CQUIN had provided an incentive for the trust to claim significant additional monies, as well as being a driver for good practice. Falls was a case in point, with the commissioners incentivising reductions in serious falls. The introduction of a falls risk assessment scheme was correlated to a reduction in serious falls. CQUIN was also credited with reductions in pressure ulcers and with supporting Skye to fund service changes and encourage further discussions with community services.

Relating to regulation

Board members suggested that the previous emphasis on finance and cost control were connected to the fact that Monitor was financially driven in its focus, and that that skewed board behaviours towards the ‘financial baseline rather than worrying too much about quality’. Having targets attached to the regulatory framework and financial costs attached to these did focus the mind and being held to account through the Monitor compliance framework was a ‘reasonable way’ of doing so. However, the current compliance framework continued to create dysfunctional consequences, as attention and energy moved from one target to the next. VRE was mentioned as a case in point where the board invested energy to hit a national standard which then slipped as the ‘searchlight goes onto something else like C. difficile’.

As the organisation was in breach of its C. difficile infection rate target, the financial penalties that followed represented a key issue for the board. A number of frustrations were outlined, including the lack of consultation regarding how the target was formulated and the relatively small margins of error (a handful of cases can lead to multimillion pound fine). The lack of dialogue regarding the target negotiation was having a demotivating effect and potentially turning people off infection prevention as ‘we’ve been set up to fail’. The trust was eventually able to negotiate the penalty with the CCGs, but the maximum fine would have completely wiped out their surplus and put them into deficit.

Financial penalties had also driven safety improvements in response to changes in insurance premiums set by the NHS Litigation Authority (LA). The money lost to litigation payments was a cause for concern, as the premium was likely to increase based on the number of claims made over the past 12 months. Further detailed analysis of litigation claims by the board was therefore needed:

We’ve been running on no claims, unprotected no claims discount, if you like, since the inception of the NHS LA and now they’re moving to relating your premium to how many accidents you’ve had. And again that will hopefully focus all boards, including ours, on understanding how that we can prevent those accidents, if you want to call them that, and obviously the costs that arise for the system for that.

Changes to the local authority funding formula from weighted deprivation age were also highlighted as problematic in reducing the amount of funding because of the high deprivation in the area. The local health economy was expected to lose a large sum of its revenue as a result of these changes. Thus, the external environment had a number of conflicting effects on the discussion of quality and safety by the board.

Strategic vision

Board members described how a culture of quality and safety was built into the vision and mission of the organisation and how this was led by the CEO. As one of the longer-serving CEOs in the NHS, his style of leadership was described as ‘one that reigned’ but also ‘one that people aligned to’. Other board members spoke critically of his style being ‘abrupt and antagonistic’:

I think absolutely culture is determined right from the top . . . He sort of determines the strategy and we run round after him doing the sums, improving the governance, you’ve got to work with somebody a long time, haven’t you, to get that kind of confidence? So, I think we are a little bit different here in how strategy’s determined in organisations where the chief exec’s been there for a long time.

Alongside the CEO’s power and influence, the chairperson strongly believed that success was attributed to ‘sticking close’ to the annual plan by holding the organisation to account against the corporate goals that had been set. Furthermore, both the chairperson and the CEO emphasised that success was linked to developing and nurturing a culture of clinical leadership, as a well as a research-centred culture:

Compared to some places, we’ve actually got very high calibre senior staff . . . We’ve got a lot of longstanding senior medical staff. We’ve also got the dimension of the academic and the medical school and the students. That actually brings a little bit of a different flavour to things . . . The academics . . . they have a fresh dimension and challenge because they’re not completely service imbued.

Despite quality and safety being expressed as ‘everywhere’ in driving strategy, some of the board members described how more work was required to better translate these messages across the organisation. The ND reported a ‘power gradient’ with the board not as in-control as they perhaps thought they were. The MD alluded to areas of the organisation that were ‘sweeping issues under the carpet’, while the trust secretary suggested communication and translation of board objectives was a particular issue in the large TH within the trust where the ‘old guard’ of senior medics still had ‘too much control’.

I think what we have become aware of, over the last year or 18 months, is that we’ve achieved a lot of change in some parts of the organisation, and in a few areas, very little change . . . it’s probably not fair to single them out, but other disciplines where the culture remains very much that of sweeping things under the carpet, of not learning.

The board had received lots of feedback about how the trust was a ‘good place to work’; however, an emerging theme was apparent that people felt inhibited to speak up or ‘whistleblow’ when there were problems in terms of quality and safety. The ‘listening events’ in response to the Francis Inquiry highlighted (through anonymous voting) that one-third of staff did not feel confident about raising concerns for fear of being disciplined. As in all NHS trusts a formal whistleblowing policy was in place, yet it was felt that further work was needed develop a culture where staff felt comfortable about raising concerns.

Leadership activities

As with strategic vision, board leadership of patient safety was dominated by the CEO. His long tenure meant that he was highly visible and had developed strong relationships across the organisation. His ‘organisational memory’ was seen as a key factor in driving organisational success. The ND and the MD had also worked at the organisation for many years.

You’ve got that stability at a very senior level . . . I think there is a bit of a benefit there . . . these very small trusts don’t have the critical mass, the financial stability, and therefore that then knocks on to quality and being able to, sort of, flex your muscles a bit and challenge the system. Whereas, when you’re big and you’ve got a bit of cash in the book, you can tend to, you know, be a little bit more arrogant maybe, but just say, ‘No, we’re fixing that. We don’t need your help, but, you know, we’ll call you in if needed’.

The CEO had a reputation for carrying out unannounced inspections, and the ND also exemplified this ‘leading by example’ style of leadership by doing a nightshift once a month. Of the CEO, the trust secretary said:

It’s his nature, it’s his personality, he knows the guys scrubbing pans in the kitchens, he knows the porters by name, he knows the cleaners who run the corridors around here by name and he takes a personal interest in them. It’s not just a bit of grandstanding, he genuinely does. And he gets out on the shop floor and I think that’s a big difference . . . pretty often, he informally pops up in peculiar places, he’s quite often got his camera with him and people then get a sheaf of embarrassing photographs slapped on their desks the next day, you know, ‘why is this corridor filled with all this equipment?’ ‘Who left all these rubbish bags there?’ So, it sets the standard very high.

Board members described how they addressed patient safety by being actively involved in the governance structures of the organisation. A NED described how chairing the clinical governance and standards committee made it clear that the board takes patient safety very seriously. The complaints process was seen as a good example. This was described as ‘very exec-led’ and ‘taken extremely seriously’, with the CEO scrutinising every single complaint. Further scrutiny was also undertaken at the complaints panel, which comprised governors, the director of quality and effectiveness, the ND and the MD.

Although some members of the board commented on how the CEO had created the right balance and skill mix, the dominance of the CEO was also mentioned as a potentially inhibiting factor. The MD felt that the chairperson could be stronger ‘in reining him at times’ in order to facilitate better discussion. One NED also suggested that too much of the patient safety brief ‘fell on the laps’ of the ND and MD and that these roles were already overstretched. The ND’s remit, in particular, was described as ‘colossal’. Indeed, the MD herself suggested that further support here would be helpful.

Board meetings and mechanisms

Board meetings were described as being predominantly focused on quality and safety issues. A NED described how the quality report containing MRSA and C. difficile infection rates and hospital mortality statistics (and similar) tended to occupy about one-quarter of the whole meeting.

That said, both the ND and MD still thought that more discussion could be given to the quality and safety papers rather than having protracted discussions about finance and business matters, which were thought to sometimes dominate the board agenda. The board agenda had been restructured, with a section now dedicated at the start to discuss quality and clinical effectiveness. This view was questioned by the chairperson, who was less in favour of the change, suggesting that it was leading to duplication with the governors’ meeting agenda and that the first part of an executive board meeting was starting to resemble a board of governors’ meeting. Others disagree:

I don’t think that there’s enough discussion at the board about the clinical issues and quality issues. I think there’s far too much concentration on politics and buildings and arguments about the office block and all this sort of stuff, which are important from the organisation’s point of view – there’s a lot of money associated with them – but I think the board agenda is dominated by those issues.

Non-executive directors were confident that they challenged and held the executive to account in board meetings. Despite NEDs’ confidence that they effectively challenged the board, in contrast, the ND’s experience of meetings was that there was not enough challenge. In her view, meetings tended to be dominated by the CEO. This was connected to views suggesting that the chairperson had a key role in encouraging contributions from other board members:

I don’t think the challenge is there. And I think that’s because it’s so dominated by the chief executive. Please don’t get me wrong – I’m not saying he’s not right or that he’s bad for the organisation. You can see very well that he isn’t bad for the organisation. I just . . . I think the board would be healthier if there was more challenge from the whole spectrum of board members rather than it being so dominated by one person. And I think some of that comes down to his personality, some of it comes down to relatively weak chairing.

The observation of board meetings found that out of just over 14 hours of discussion observed across the four meetings, a total of 200 minutes were directly related to patient safety issues, around 24%. This time was predominantly spent on discussions about C. difficile infections (about 70 minutes), and covered the target being set and the financial penalty imposed if the target was breached. A wide range of shorter discussions were directly linked to the Quality Report agenda item, containing performance indicators (falls, sharps, needles) as well as clinical risk factors, such as clinical negligence, and environmental health. The majority of these board meetings were based around monologues by the CEO, who provided a commentary on the ‘matters arising’ that the board faced, particularly in relation to estates and service reconfiguration.

Governor relations

Governors were emphasised as a key priority for the CEO and the chairperson. The CEO described how they were given ‘top-level’ information with which to scrutinise the board and to enable them to ‘drill down’ and ask critical questions. For the CEO, this had led to a number of challenging questions, particularly in relation to high levels of patient falls and difficult staffing levels. In contrast, other board members, particularly the finance director, supported the view that governors were not particularly effective. He felt that governor meetings tended to be ‘easily distracted’ from the key issues and got ‘bogged down’ in personal interests and issue campaigns.

Interviews with governors revealed a great deal of support for the board, with particular praise for the trust secretary, chairperson and chief executive. Governors interviewed were happy with the type and amount of information they received. Governors were involved in a variety of subcommittees, including the complaints panel. The lead governor described how she worked very closely with the chairperson and the trust secretary. The majority of quality and safety issues were covered in the Quality of Patient Experience working group.

Ongoing issues were raised about improving the relationship with NEDs. They described how NEDs tended to lack interest in governors, which was manifest in NEDs often not turning up to meetings and having little direct contact. The difficulties of finding members of the public with the required competencies to be a governor were also raised. It was reported that governors were often happier to do the softer things (‘potter round and do patient food testing’), with the general public having little understanding as to the complicated reality of the role.

The majority of discussion at board of governors’ meetings was built around governor education regarding the latest policy developments alongside the provision of summary information about performance. There were a variety of PowerPoint presentations covering issues about improving patient experience, workforce development and infection control. The CEO was central to these discussions, with his report occupying a significant part of the meeting agenda. Here, the CEO provided a summary of the key issues facing the trust and sought to mobilise governor support. Governors made only a limited contribution to the meetings, other than largely supporting the CEO and chairperson in their current negotiations. In relation to C. difficile infection rates, they supported the CEO in his criticism of the financial penalty being imposed for a breach of target. Some questions and challenges were raised regarding patient and staff experience. For example, a governor consistently made the point in the meetings that the board needed to improve hospital food by providing healthier options.

Assessing organisational performance

Board members expressed confidence and support in the ‘good reports and regular dashboards’ that were provided. In addition to the mandated performance measures, board members described other ways in which safety performance was measured. They had recently adopted the CAT, which was being adapted to the needs of the local context and used to provide a monthly online survey covering a variety of clinical quality assurance measures. The board also benchmarked their performance against other FTs.

Although the overall message was one of support for the board, concerns were raised about the amount of information that board members had to read in advance of and during board meetings. Both the MD and a NED suggested that more detailed ‘deep-dive’ analysis of particular issues was needed, as there was a danger with ‘high-level reports’. Other concerns were raised about the danger of looking at serious risks at the expense of more everyday ‘average’ issues.

The corporate governance committee spends quite a lot of time fretting about the Corporate Risk Register. Are we looking at the right things? . . . What about the stuff in the middle ground? . . . Something ticking away there that’s been running for months or even years, that might actually be more of a concern . . . and I worry about some of those things from a governance perspective because people tend not to keep their eye on those so much.

Walkabouts were cited as something the board regularly carried out, but again there were anxieties expressed by a NED who held the view that they were ‘never quite sure of what’s going on’ therefore a ‘certain amount of wit’ was needed to understand the high-risk areas. The MD wanted the board to build-in walkabouts as part of the board meeting in order to gain further soft intelligence. A NED also supported the view that further work was needed with the organisation in relation to communication and engagement with different members.

I, personally, feel that there should be more board-to-ward communication . . . some of them, I think, don’t actually . . . They’re a bit frightened of it . . . I think some of them feel out of place. It’s a bit like putting a barrister in an operating theatre – they’re fine in a courtroom, but you put them in an operating theatre and it’s a different ball game.

Relationships with commissioners

There was reported to be a problematic relationship between the trust and local commissioner. A NED, for example, described how relations with the former PCTs were based around ‘tick-box exercises of patient safety rather than looking at the bigger picture’. Relations had improved with the introduction of CCGs and the board had quality meetings with CCGs, which were described as being positive and constructive. Board members described having good professional relationships; however, there were some points of contention in relation to capital investments within the local health economy. Here, local ‘rivalries and egos’ created strained relationships, particularly regarding to the decision to build a new hospital.

Reflecting on their experiences, the interviews with commissioners suggested that the board was at times too ‘aggressive in its ambitions’ for business expansion and in its attempts to seek control of the local health economy. Despite these tensions, they recognised that there had been a lot of work within the trust to promote quality and safety issues. From this standpoint, Lewis was described as an excellent trust providing excellent care:

My opinion of the trust is they’re a trust that absolutely puts patients at the centre of everything it does and wants to do the right thing, and if things go wrong, hands in the air and tells us it’s gone wrong, this is what we’ll do to put it right.

In relation to the CQUIN, board members suggested that the financial incentive did make the trust ‘stop and think’ about quality issues. For example, the CQUIN for dementia had worked in driving improvements. However, there was a fear that within the new commissioning arrangements there was a danger that incentives such as the CQUIN would get less attention. In contrast, commissioners suggested that the CQUIN had not driven quality improvement. Rather than acting as a financial incentive, the CQUIN tended to be perceived as a sanction or penalty for trusts with ‘money being taken off them’. There was a feeling that the CQUIN therefore needed to be redesigned if it was to deliver its intended outcomes. Options proffered for improvement included redesigning the CQUIN so that they ‘joined up across the pathway’ and for the CQUIN to apply across ‘hospital stay experience as well as in the community’.

They see it as their money, it’s extremely difficult to negotiate stretching targets and you know we are sort of working on that and we certainly don’t give them all of the money up front.

Relating to regulation

For the CEO, Monitor was valued as being a ‘genuine independent regulator’ built on a ‘constructive relationship of guidance and opinion’. Having the same relationship manager for 5 years had helped to nurture informal dialogue on what was described as a need-to-know basis. In contrast, the MD suggested that, because Monitor was primarily set up as a financial regulator, this largely framed the discussions. Although the Monitor compliance targets had driven improvements, the point was raised that the threat of the financial penalty per case of C. difficile infection over the target trajectory had negative consequences for services delivery. Problems were also raised regarding the CQC. The CEO suggested that it was up to the CQC to better communicate emerging issues. The MD also characterised the CQC as ‘very much a box-ticking apparatus’ that did not ask the right questions or ‘talk to people on the ground’.

Strategic vision

As a large specialist TH, Arran’s strategic vision was focused on supporting clinical excellence and securing a drive for further funding for research and development. This was particularly the case in clinical areas, where the trust was seen as being a leader in the field. There was a clear aspiration to further enhance the ‘global brand’ of the organisation.

In striving for clinical excellence, the fluidity that characterised the trust meant that board members often associated more with a ‘vision’ of clinical excellence than a corporate plan. The board had not yet translated this vision into operational plans, as one NED observed:

Some people want a very detailed strategy document written and [others are] of the view that so long as we broadly know where we’re going, we don’t need it, because it changes so much; it’s hard to be definitive about where we’re going, even though we think we want to know where we want to go. So writing it down, it would be out of date as soon as you’d written it.

One board member explained that the reason for the lack of strategy by the board was in large part connected to a preference for a ‘crisis management’ style of governance. The point was also raised that safety and quality improvements tended to be patchy across the organisation or merely bolted on to local initiatives at the division level rather than being seen as an organisation-wide approach:

We are much better when we’ve got a crisis, a burning platform to deal with rather than seeing a problem and dealing with it in a structured, planned way. I’m overstating the distinction but I just think that things tend to be quite operational, crisis driven rather than improvement focused.

Some board members expressed concerns that the drive for innovations to increase efficiency could compromise service quality and patient safety. This view was further supported by the observation that the trust had a ‘strong audit culture’ and was ‘strongly focused on research’ but less focused on continuous quality improvement:

I do hear stories about people making perhaps not very well advised shortcuts because they thought that they had to hit the budget overall . . . a more able manager will find a creative way of spending what they need to deliver the service. Your less creative, innovative or less financially enabled person will hide behind the fact of, ‘Oh I didn’t think I’d got the budget to do that’ but it’s not an acceptable behaviour.

However, other board members disputed this view and suggested that the Quality and Safety Agenda was given priority, with ‘the days of having finance at number one long gone’. Discussions about patient safety were said to have had high visibility throughout the board and the Quality and Safety Committee. The strong clinical presence running throughout the organisation also ensured that issues around finance did not compromise quality and safety.

Leadership activities

Interviews with board members in Arran highlighted how the leadership associated with patient safety contrasted with other organisations. In support of its strategic vision of clinical excellence, the governance structures of the trust were led by medical professionals, with the trust having multiple medical directors each operationally responsible for the financial, workforce and patient safety issues within their own directorate. These MDs formed part of ‘clinical boards’, consisting of divisions that reported upwards to the executive board through a range of feedback and audit systems. In addition, the trust had a medical directorate, which provided forums for sharing quality and patient safety information across the various clinical boards:

It’s a clinically led organisation with nursing, doctor and AHP (allied health professional) feed-in to specialty and division . . . If you believe that the right systems and processes will stop mal-alignment and errors, we have a number of checks on that.

Although board members generally expressed confidence in the medically led assurance mechanisms with regard to patient safety, some concerns were expressed about the underlying culture around incident reporting. This was acknowledged as ‘suboptimal’ and was connected to deeper staff concerns about the reputational and career consequences of reporting in an environment perceived as being hierarchical and where people could be ‘scapegoated’ for raising issues about poor quality and unsafe care:

We still have a way to go about developing a really positive reporting culture; for everyone to feel that we’ve got a just reporting culture that’s flexible, that then acts effectively. So we still get ‘never events’, we still get serious incidents.

It’s a matter of record that we have a less-than-fantastic record on bullying and harassment which we’ve tried to address over the last 3 or 4 years, but it’s a fact.

In response to these issues and concerns, considerable effort was being made by the board to encourage a more open and responsive organisational culture. The ‘Making a Difference Together’ initiative was set to support staff in translating the values of kindness, teamwork, safety and improvement into their working values and practice. This initiative was incorporated into all appraisals and consultant appointment panels in an attempt to spread these values throughout the organisation:

Whereas before if you were a UK leading surgeon with some good R&D [research and development] behind you and the support of the university I think you could have been pretty assured you’d get a clinical place. Here, we’re starting to get to the place now where we’re saying, ‘Actually no. If you don’t appreciate the importance of kindness, if you don’t appreciate the importance of working in a team . . .’

Although this development was considered to be important and a priority, some concern was expressed over the extent to which these cultural changes were embedded within the organisation. An issue of significant ongoing tension was raised by a board member, which related to his perception of the corporate culture of the trust. As an ambitious organisation which promoted the view that ‘we want to be the best’, it was felt that this had been a distraction from the day-to-day business of providing ‘the best things for patients’. It was also felt that more needed to be done to translate the ‘high-level vision’ into something that could ‘change behaviour on the front line’:

My concern has been when we say we want to be the best at the same time as we’re taking a defensive position on a safety issue or an access issue: I’m suspicious because it’s just who appears to be the best rather than actually [who is] providing the best service.

An example of this was the introduction of the World Health Organization’s ‘Safe Surgical Checklist’. Theatre teams were taken out for a half-day at a time as it became apparent that there were cultural issues with staff feeling that they were unable to speak out openly if there were any issues about the quality of care, as well as being unable to address problems associated with clinical teams not working well together.

Board meetings and mechanisms

It was reported that over recent years there has been an increase in time spent on quality and safety at the expense of finance. This was driven in large part by the arrival of new NEDs, many of whom had extensive clinical experience. In addition to increased time devoted to patient safety, the manner in which these issues were addressed was also changing, from previously ‘noting requirements and ensuring compliance’ to attempting to understand such requirements and integrate them within organisational routines and practice:

I mean [patient safety has] certainly changed in terms of profile, as I’ve already indicated. I think it’s changed in terms of not just ticking the box that we do it, if you see what I mean, but actually people genuinely trying to assure themselves that it is done.

Openness was identified by some members of Arran as an important feature of the organisation’s culture. This was exemplified by the first part of board meetings being open to the public (before this was made mandatory post Francis Inquiry). Board members described how this had facilitated honest exchanges between executives and non-executives. This was particularly the case for the NEDs, who could ‘always speak absolutely freely and without consequence’. NEDs did not ‘pull their punches’, which meant that the executive was ‘held to account’:

What I would say is that [NEDs are] pretty well empowered here. They are seen and treated with a lot of respect by the executive. We don’t always agree with everything they say, but when they say ‘jump!’ to quite a large extent we jump.

Although many NEDs were confident that they could challenge other board members in meetings, some raised questions about their ability to challenge executives, especially in relation to the availability of detailed information on quality. A NED described his experience over the past 6–12 months of watching the increase in C. difficile infections. He had had to trust professionals that ‘the matter was in hand’, but was anxious that he had not seen data to support this view and had not seen a consistent action plan being formulated.

Overall, board relations were described as being good, with healthy challenge and debate alongside ‘friendly fun and banter’. Indeed, some board members welcomed the way that ‘challenge’ was starting to increase. This was seen as a positive shift from the ‘incredibly stage-managed’ meetings of the past, as ‘executives and non-executives were increasingly looking to resolve ongoing organisational problems’:

I think we’re only starting to get to a place where execs can question one another, but I would say that’s definitely an evolution. Previously, I’d have said that was almost unheard of, that would have been seen as being disloyal and, not career threatening, but just not done. That’s – I would say that’s definitely changed. So it’s better than it was but I’d say it’s got a way to go before you can speak freely.

Observations of board meetings revealed that, of the almost 14 hours of discussions across the four meetings, a total of just over 4 hours was directly related to issues of patient safety, representing almost 30% of board time. The largest amount of time spent on any single issue related to an environmental health risk involving contamination at the kitchens (just under 1 hour) – an issue which was not part of the trust strategy and, thus, was not included in proactive quality improvement work. Other notable issues included safety in operating theatres (about 30 minutes), orthopaedic performance (30 minutes) and patient falls (just over 10 minutes). Considerable additional time was devoted to more general discussions about quality improvement, including ‘patient story’ narratives (30 minutes) and quality improvement presentations related to specific service areas, such as community nursing (almost 1.5 hours).

Governor relations

Governors were viewed as ‘the eyes and ears’ of the trust. Governors were involved in the Quality and Safety Committee, where they could hold board members and managers to account. Board members described being impressed by their ‘penetrating questions’. Governors were also represented on the clinical quality review group and had their own high-quality patient care group, which carried out focus groups or interviews.

Governors and the board had a good relationship, with the board providing governors with information and the governors providing the board with patient journeys. Although some suggested that governors were ‘pretty good’ at holding the trust to account as a ‘critical friend’, it was also expressed that it would be very difficult for governors to actually hold the board members and managers properly to account with regard to the ‘bigger picture stuff’ as they lacked detailed information and knowledge about board issues. As with other sites, issues were also raised about individual governors having their own ‘hobby-horses’ and interests rather than putting forward a more rounded view.

Analysis of governor meetings found that the majority of time centred on governor education and the provision of summary information regarding trust performance. These meetings tended to have three central parts running throughout: a PowerPoint presentation about a particular topic (in this case a junior doctors survey and service line reporting); the chairperson’s report, providing a summary of recent activities and events (such as charity work, visits and any latest developments); and the CEO’s report, which provided a summary of the latest performance ratings. The meetings were largely driven by the contributions from the chairperson and the CEO. The NEDs did not contribute to any of the discussions. Overall, governor contributions were limited to questions for further information regarding the presentations, further clarification about particular trends related to hospital activity, providing updates about working group activities and discussion and reflections directly related to patient experience. There was no discussion of patient safety at the meetings outside of the CEO’s summary of progress regarding IPC.

Assessing organisational performance

Arran had developed information systems and an information pack that concentrated on safety as well as the other key aspects of performance, which were broken down across divisions within each clinical board. With access to detailed information from across all of the divisions, an executive described how her exposure to board information illustrated that they were functioning well in relation to quality and safety:

. . . it gives us a pointer in terms of who we should be speaking to, and there’s some commentary in there as well and I place a lot of reliance on this, as I think many other people do.

To complement the RAG ratings, board information was supplemented by ‘Care Thermometer’ clinical data, which provided more fine-grained information at ward level. Although this was welcomed, some concern was expressed that these disaggregated data were currently used in a largely reactive manner, with the view that the board was a little slower than they should be at addressing particular issues:

For the most part, we look back at what has happened and try and learn from it and then try and quickly apply, ‘Oh if that’s happened, what must we do looking forward’? If we were to aspire to the absolute best practice, which we do, there’s something about tracking non-safety trends in an intelligent way that you can now do with the level of data capture and using it to predict where the might be a problem.

There was also some concern about an over-reliance on quantitative summaries. To balance this, board members used a number of techniques for gathering soft intelligence, including unannounced visits, ongoing review of all complaints and walkabouts:

The secret of it all is to talk to a lot of people because people will often tell you in conversation what they wouldn’t tell you if they had to write a report about it or if it had to go through their line management process, wherever that line management process may apply.

There was some support for the use of ‘patient stories’ to provide additional qualitative information and soft intelligence on patient experience. However, this was not universally supported by board members and consequently patient stories were not routinely used:

We’ve done [a patient story] once and the board were quite staggered at the story that was told. But they were staggered and then they were sort of – ‘so what’? Well the ‘so what?’ is that we developed the whole ‘Making a Difference’ campaign as a result of it. So the board are starting to see that this is important. But I think they don’t want to be just hit with a sledgehammer every month and yet I think there’s nothing wrong with that! [Laughs.]

Relationships with commissioners

Arran relies heavily on specialist commissioning contracts. The majority of their income comes from specialist commissioners, especially the Nationalist Specialist Group (NSG). Although, on the whole, relations with commissioners were described as cordial, it was reported that the fragmentation in the commissioning landscape made it difficult to have meaningful discussions. However, it was thought that the NSG were helpful allies in promoting safety within the organisation:

We’re a much bigger client of National Specialist Group than most other hospitals. That will probably be a good thing for safety because I think for the first time . . . the people on the other side of the table really know what they’re talking about . . . The CCG are less likely to get hold of the Safety Agenda.

In relation to patient safety, the Clinical Quality Review group within the trust provided an opportunity for commissioners and the trust to meet and review matters related to quality and safety. Commissioners would scrutinise their performance, looking at serious incidents, infection control and waiting times and pressures ulcers, etc. Commissioners and the trust were currently in discussions about the financial penalties regarding the breach of C. difficile infection rate targets.

Relating to regulation

Monitor was well regarded for ‘keeping the board on their toes’. One result of external performance targets was that the board began to invest time and energy into developing their own performance indicators. However, questions were raised about the regulatory frameworks and the targets established in relation to C. difficile and MRSA infections. The board was in discussion with Monitor about the recent breach in this area: it was thought that Monitor did not take into consideration the specialist, tertiary nature of the trust and its policy of admitting patients from other hospitals that already had infections.

Islay’s response to C. difficile infections

Islay was seen as high performing in terms of C. difficile infection containment. Interviews with IPC committee members reflected pride in their achievements, and they expressed their support of board members’ forward thinking in relation to infection control. The Quality Improvement Strategy was central to progress, as it made infection control a key priority from the outset. The very first quality improvement collaborative undertaken by the trust was related to C. difficile infection control, with results showing significant reductions in the participating patient group.

The board provided financial support and encouragement to the infection control team; C. difficile infection control was well resourced, with six HPV (human papillomavirus) fogging machines to hand, a team of 24/7 domestics available to move patients at all times and side rooms also available. The IPC team were also able to draw on a relatively advanced electronic patient record system to manage and track patients with infections.

The team suggested that they had a slightly different dynamic to other trusts, as the MD was the director of IPC and considered to be a ‘crucial driving force’, providing strong leadership and accountability in relation to IPC performance. The MD supported the IPC team in situations where directorates were underperforming, providing the necessary ‘clout’ that meant audits and reports were taken seriously. A further example given was that of ‘antibiotic stewardship’, in which the MD was central to a ‘Green for Go’ campaign concerned with changing compliance with the antibiotic guidelines from ‘red’ to ‘green’. The ND was also important, keeping in regular contact with the IPC team and creating a direct reporting link to the board.

The arrival of performance targets was cited as crucial in shaping board behaviour, as the financial cost for breaching the C. difficile infection rate target raised its corporate profile. Building on these mandated performance targets, IPC committee members described how the board seized the moment very early to develop division-based targets to increase competition across the divisions, to reflect on progress and to ensure that divisions were held accountable.

Ongoing issues in the governance of C. difficile infection rate management were identified. Although targets had driven a number of improvements, there was now recognition that the C. difficile infection rate targets needed to be revised – continuing reduction to a point of zero tolerance was not realistic and difficult to communicate to staff groups. IPC committee members did, however, suggest that more could be done in relation to preventing and controlling C. difficile infections, including improvements in the translation and accessibility of antibiotic guidelines, availability of web-based guidelines and other IT resources for the workforce, such as apps.

There was still work to be done regarding joined-up working with the community. Despite the availability of a regional collaborative for all infection control leads to meet and discuss key issues, collaboration between hospital and community infection control teams could be improved. The fact that community and acute trusts had separate targets caused friction between teams. Maintaining the focus and resources for C. difficile infection control was also presented as a challenge given the existence of ‘rival’ infections. For example, a recent outbreak of CPE (carbapenemase-producing Enterobacteriaceae) infection in a neighbouring hospital highlighted the likelihood of more pressure on resources such as side rooms used to treat C. difficile-infected patients in the future.

Skye’s response to C. difficile infections

At the time of the fieldwork, C. difficile infections had gone up from being undertrajectory to a point where acceptable levels had been breached. The cause of this trend related to continuing problems with matching capacity with demand as bed occupancy levels were running at 90%. Such levels prevented the opportunities to carry out deep cleaning, as wards were full and essential isolation facilities, such as side rooms and en-suite rooms, were not available.

Historically, members of the IPC committee suggested that board approaches to infection control had been ‘hit and miss’, with frustrations expressed about the time taken to make decisions, putting costs before quality and a culture described as ‘ticking a box but missing the point’. The IPC team provided evidence to show one of the effective ways to control C. difficile infection through good environmental decontamination, but this had taken 3 years:

[D]eep cleans: we have been trying to get that in place for 3 or 4 years. It has gone to the finance group, it’s essentially been said ‘Oh well, yes we support it’ but there’s no money, so essentially it didn’t happen.

It was noted that responses had improved recently, with the latest board proactively looking to involve the IPC committee and placing them at the forefront of the strategic plan. The team were now in regular contact with the CEO, who had also recently decided that there should be an executive member taking part in every root cause analysis. Despite this increase in engagement, questions were still raised about the MD who, despite being a member of the IPC committee, never attended the meetings. This lack of engagement was linked to wider problems with medical involvement in infection control. The COO also tended not to attend IPC committee meetings. Embarrassingly, when the trust breached its target and was asked to report to Monitor, key personnel were not able to explain what the issues were.

Although targets were supported in getting infection control funded and recognised, questions were raised about unintended consequences and equity, as elsewhere. First, board focus on targets such as C. difficile infection rates had the potential to distract from other important issues. Second, as the organisation was already starting from a low baseline, continuous reduction in the C. difficile infection rate target meant they were more likely to get penalised when compared with a trust starting from a much higher threshold. The trust managed to negotiate a reasonable financial outcome with CCGs after they breached the target. Indeed, there were positive (unintended) consequences in that the penalty enforced further investment to reduce infections.

We’ve probably reached the stage where most organisations have probably reached the irreducible minimum. And it’s no coincidence that rates have probably plateaued or, if anything, are slowly creeping up.

Although policies and guidance were available, further education about antibiotic stewardship was needed, as knowledge of microbiology among nursing and medical staff was described as poor. While junior doctors were a problem in this regard, the introduction of consultant champions had improved antibiotic prescribing. The trust also needed to find new ways to translate information. Information technology was tabled as a solution. The trust had an electronic prescribing system which was good in flagging up resistant antibiotics on patients but it needed to be improved to build in more ‘red flags’, such as whether someone was C. difficile toxin negative. Developing an antibiotic website along with the use of apps were ways in which the trust was making information more accessible.

The IPC team was collaborating with the local health economy to deliver training around C. difficile infection control and antibiotic prescribing and was part of a clinical governance forum where old cases of C. difficile and MRSA infections were discussed. However, more work needed to be done at CCG level to fill the big gaps within nursing (residential) homes.

Lewis’s response to C. difficile infections

Interviews with IPC committee members revealed overwhelming support for the board in relation to its oversight of C. difficile infections. All IPC members commented on how the CEO, ND and MD were receptive to and supportive of organisational efforts related to IPC. Support at board level was demonstrated by successful business cases, which funded the organisational response to C. difficile infections. It was also demonstrated by CEO and MD attendance at serious infection review meetings (SIRM), which gave further power to the infection control team by signalling the backing of the board. The CEO was particularly important here in terms of knowing the different roles at each level and managing interactions between different parts of the organisation. The ND was also a member of the IPC committee and its operational review group, and her involvement was welcomed:

I’m involved in preparing the report for the board on a monthly basis, so I know the questions that they are asking on the basis of the information that we are providing. They do want to know what is happening and what we are doing to try and change the current situation and improve our rates as much as possible.

The introduction of infection control targets has driven interest in C. difficile infection and led to investment, enthusiasm and support for infection control. As a result, questions and concerns raised by the board tended to focus on the targets at the expense of wider factors affecting patient safety. Although targets have undoubtedly raised awareness, IPC committee members suggested that a different strategy was now required, as continuing reduction in C. difficile infections was considered to be no longer achievable. They were also concerned about securing greater recognition of the fact that a tertiary, specialist trust such as Lewis tended to receive the most complicated patients with existing problems (i.e. C. difficile carried from elsewhere).

At the moment it’s all about C. difficile so people are prepared to throw money at it . . . but things that are not quite so high profile and like E. coli maybe don’t get the same attention.

The board mind-set was changed to ‘we’ve got to meet these targets, we’ve got to reduce the numbers . . . we don’t want Monitor on our back, we don’t want the Care Quality Commission on our back and we certainly don’t want fines.

More work was needed to engage the workforce in IPC issues, particularly those related to C. difficile. Although information dissemination via e-mail or the intranet was considered good, staff engagement was acknowledged to be low for junior and senior clinical staff alike. This was particularly so in relation to antimicrobial stewardship (getting clinical staff to prescribe the right antibiotics). Things were improving as the IPC team were starting to apply more pressure to medical staff, but they required more support from senior medical staff, including the MD. Compulsory attendance at patient reviews, particularly those concerning C. difficile infection, was suggested as way to improve the situation. Further issues concerned how best to educate junior medics and change the practices of senior medics to be consistent with new guidance; initiatives included an induction package for junior medics, additional support from pharmacy teams and ‘antibiotic champions’ in each directorate.

Concerns were also expressed about the epidemiology of C. difficile, with focus moving to community settings and patient profiles. It was felt that more community involvement with C. difficile was required. The trust formed part of a region-wide HCAI partnership that concentrated on C. difficile and MRSA, and had a subgroup sharing best practice, experiences and lessons learned. Despite good relations, further joined-up working between primary and secondary care was called for, as the approach to managing C. difficile infections was to look at C. difficile carriage in the patient population. As it stood, the key issue was a lack of clarity over who was responsible for infection control in primary and community settings such as GP surgeries and care homes. A ‘great deal of work’ was needed in terms of education and awareness at the interface between community and secondary care.

C. difficile is much more widespread in the greater community than we had thought and that carriage of C. difficile by patients or carers or visitors is probably a much greater risk for transmission than using an antibiotic.

Arran’s response to C. difficile infections

The IPC team at Arran outlined an extensive infection control infrastructure supported by ‘infection control link practitioners’ (mainly nurses) who were trained to act as ‘the eyes and ears on wards and within departments’ for audit purposes and charged with embedding infection control training into practice.

One IPC nurse suggested that, although some of the education was relatively simple for some audiences (e.g. educating those nurses in charge of cleaning wards), the challenges came when trying to engage the wider clinician audience to changing particular practices. Recently the approach by the IPC team has changed towards providing a more customised ‘tailor-made’ education programme for particular clinical groups. This approach has gained a number of successes, including ‘transforming hand hygiene’ within A&E by developing a series of must-do rules:

[Y]ou need to go into the silos and actually see, you know, who are the champions there, who are the opinion leaders, how can we influence them to get them involved and then recruit them and try and get them to join in.

In contrast to other sites, the IPC team described having a distant relationship with the board: although the IPC committee had relatively good relationships with staff, the relationship with the board proved to be challenging, particularly when attempting to escalate problems; one IPC described being told by the CEO ‘to stop coming to him with bad news’. This opinion exemplified a reluctance to engage with underlying systems and process problems within the organisation, with a preference for ‘good news’ headlines. Such a perspective was linked to wider issues with the incident reporting systems of the trust. This included ‘a culture of bullying and intimidation’ prevalent throughout the organisation, which meant that there was a tendency for middle managers to project the image of being ‘seen to be doing okay’ by achieving targets and making sure none of the issues was escalated up to board level:

In the lower ranks we have a joke about how only good news goes through and none of the bad; so that you know you’ve got a problem on the ground but by the time it gets to the chief exec people are all saying everything’s okay . . .

. . . if you do pick up on anything you will find that people will be too nervous to tell you about all the intimidation that goes on because even when the CQC came in . . . it wasn’t said but the implication was that nobody should really go over there and tell anybody about the bullying culture. It wasn’t picked up here at all but it does exist.

Although the trust had historically made significant achievements in relation to C. difficile infection control, recent trends suggested that occurrence of C. difficile infections was increasing, with a breach in the performance target likely to happen. An IPC committee member explained that this increase was in large part caused by the distorted picture that the board received in relation to C. difficile infection management performance. The lack of systems to effectively govern C. difficile infections meant the response only happened in times of crisis. Although the organisation did respond to infection control issues, it did so reactively, once an external target had been breached.

. . . suddenly people become aware of it and say, ‘Why didn’t anybody tell us?’ I’m very cynical about it.

. . . it’s not just me, it’s organisationally, they’re not a listening organisation, not to the staff that are working inside them.

The Islay board’s development of the ‘staff stories’ initiative

Board members were looking to further develop staff engagement as part of the Quality Improvement Strategy. To do so they carried out what was described as a series of ‘Berwick sessions’ to identify staff concerns in relation to quality and safety. These Berwick sessions had two objectives: to raise staff awareness of achievements currently being made and to enable staff to raise previously unidentified concerns, thereby developing emerging themes.

Building on the idea of patient stories as a way of providing a voice to untapped intelligence about the organisation, the board developed the idea of ‘staff stories’ about their past experiences of being a patient within the organisation. The board requested volunteers to ‘have a cup of tea’ and speak to them about their patient journey and to rate particular aspects of the service. The questions asked by the board built on CQC inspection questions such as ‘Did you feel safe? Did you feel that your treatment was resourceful? Was it effective? Was the outcome what you expected it to be?’ These questions were supplemented with a self-reported visual analogue scale ranking experience in terms of safety, effectiveness, resourcefulness and caring.

Feedback from participating staff identified positive and negative experiences of the initiative. One staff member described the positive experience of being listened to:

[T]hey were quite transparent throughout the whole thing and admitted that, you know, if there were any faults in the services they wanted to rectify them . . . it wasn’t about pointing blame towards anyone, it was about learning from mistakes . . . and putting strategies in place to improve the services that they offer.

However, this respondent also stated that it could have been further enhanced by having a clinical perspective from someone involved in the patient journey under discussion to triangulate patient experience with a frontline staff perspective and initiate a dialogue with the service under review. Greater transparency of dissemination of feedback across departments would also have been welcomed, as would evaluation of the extent to which lessons were being learned and applied.

if they’d had someone there as a representative from that service that might have kind of helped them to understand things a little bit more . . . I came away in some ways feeling not bad but, you know, oh I feel like I’ve just been really negative there about that part of the service . . . Whereas if the person from the service had been there, then perhaps that might have been giving the board a bit of a better understanding of how it operates and why they do things in a certain way.

The Skye board’s development of the ‘Bed Plan’

In a context of ever-rising demand for services, the trust was unable to match capacity and demand or successfully manage patient flow. Faced with a bed crisis, the COO developed proposals to redesign bed management. The key proposal was to close emergency admissions and open an additional rehabilitation ward at one of the trust sites in an attempt to reduce the number of outliers and ward transfers, improving efficiency and quality by having patients on the right wards with the staff skills.

Even on the heady days of summer we still seemed to have patients who weren’t in a bed belonging to medicine. It wasn’t entirely correct . . . that seemed to give us a signal that there was something amiss.

. . . [the bed plan] gives us a better patient experience by . . . not moving anybody overnight and, making sure people aren’t in acute beds who don’t need to be there.

Interviews with staff involved in the formulation and implementation of the bed plan indicated that the proposals had improved the situation, reducing the number of outliers. Investment in additional bathrooms and bays had also improved infection control, as it was now possible to isolate cases and provide decamp facilities. Yet despite these successes interviewees were critical of the amount of time needed to make these changes. The Bed Plan was described as ‘the outcome of a discussion many times over’, where continuous changes at executive level led to protracted negotiations over the plan.

Managers and nurses identified limitations in the engagement processes associated with the plan: consultation processes had largely failed to engage nurse and ward managers in the key admission areas of A&E. Further, consultant physicians had each been nominated a number of beds spread across up to six different parts of the trust; one manager described their anxiety with the possible implications for patient care of making physicians ‘homeless’ as a result of the changes, and suggested a need for a dialogue with the executive team.

It was further suggested that better links were needed between the board and the clinical support unit management team, clinical directors, and senior nurses if engagement was to be improved. Suggestions to enable this included staff invitations into boardroom decision-making processes and increased visibility of board members on wards. The board needed to ‘get their faces known out there in clinical practice’.

If you went round an organisation and you asked most of the matrons, who are obviously heavily involved and impacted by the Bed Plan . . . if you asked all of them for details there’d be a very different range of answers and some people wouldn’t be able to answer. We all know the new ward is opening . . . but when you go into the details of that and what that actually means and where patients . . . it gets a little bit fuzzy, especially around how we’re going to staff those wards.

Despite these difficulties there was cause for optimism. Some interviewees noted that the board should be given credit for the huge effort involved and also their increased openness, visibility and enhanced communication via team briefings. The new CEO was perceived to be working in the right direction as the forthcoming strategic plan built on a consultation process.

The Lewis board’s development of the ‘After Francis’ initiative

In response to the Francis Inquiry, the MD, the ND and the director of quality scrutinised all 290 recommendations of the Francis Inquiry and RAG rated the organisation against those recommendations. For those areas deemed less than fully compliant, a series of working groups and listening events were set up to explore issues related to communication and staff awareness of how to raise concerns. The 3000 staff who attended the listening events were asked to anonymously vote ‘Would you recommend this as a place to look after your friends and family and if you had concerns would you know how to raise them and what mechanisms are in place to do so?’ Although over 96% of attendees said they would recommend the trust, the board wanted to know why 4% would not; to do so they invited staff to express their concerns and make suggestions for improvements via an anonymous electronic survey.

The 960 responses were largely positive but a significant proportion (20%) stated they were not confident to raise concerns: specifically, a consistent theme was the inability to raise concerns about senior medical staff. Suggestions for improvement included helping staff build their own confidence to be able to raise a concern, and reassurance that they would be listened to and taken seriously, and that reported inappropriate behaviour would be challenged. Themes around communication, attitude, hierarchy and the conduct of some groups of staff were raised.

[W]hile the vast majority of our medical staff are fantastic, absolutely superb, would willingly go to any of them for referral or care, but we have a small group who have been brought up in a particular style and they manifest that, and other people think it’s OK to behave in that way.

Although the trust had a whistleblowing policy under which concerns could be managed confidentially, areas of improvement were required in relation to reporting and ensuring feedback of action to those raising issues. Strategies in response to these findings included the incorporation of a ‘core professional and leadership behaviours’ dimension into all staff appraisals to ‘talk through with you how you feel your role fits into that bigger trust vision and strategy’.

Although the introduction of the new appraisal system received good feedback from staff, some resistance to staff appraisals has been made by doctors; some ‘difficult conversations’ led by the MD were needed with a small cohort of clinicians who, although clinically competent, displayed disrespectful behaviour which negatively impacted on staff morale and patient care.

The slight concern I have in this organisation is some of the badly behaved are very, very able and competent clinicians – probably you would describe them as exceptional – and because of that their behaviour has been tolerated and appears to be tolerated and they appear to be protected . . . That person continues to be protected and I feel very uncomfortable with that, and I’m in a senior role and I am aware of the difficult individual and feel completely unable to do anything about it. And I don’t like that; I think that’s not good.

I don’t get the sense that consultants . . . see themselves as managers and leaders in the organisation. They’re here to do a job . . . You know it’s a good job, it’s a technical job, it’s a challenging job, and it’s a very skilled job. But it’s – the responsibility – I just don’t get the sense that there’s that kind of commitment to the organisation and they’re here for life any more.

Francis Inquiry feedback and listening events also indicated poor communication between corporate teams and ward-level staff in many parts of the organisation. In response, proposals for new means of communication were developed, in order to standardise operating procedures. These included inviting all staff groups to feedback events at specific places and times, consistent publication of information bulletins and increases in the ‘presence’ of board members within staff information bulletins.

[W]e want to take that back a step, refresh the standard operating procedure around that to make it much more ‘Right, that will come out on a Friday lunch-time’.

On reflection, further engagement of senior medical staff at the outset would have improved the initiative. Because of time constraints, senior clinicians were not able to come away from their clinical commitments to attend, and therefore a more proactive approach was required. Some staff members also suggested increased involvement from the MD and NEDs was required:

I suppose I’m seeing much more from the nursing taking the lead, from quality and effectiveness taking the lead – and instinctively I feel that you almost shouldn’t be able to see any light between the medical and the nursing director on this.

The Arran board’s developments on junior doctors’ dissatisfaction

The board was made aware that the annual GMC national survey had indicated junior doctors’ concerns about the possible negative consequences of reporting a colleague’s inappropriate practice. The survey also identified that ‘undermining’ (a term for bullying) was indicated in a handful of the 40 specialties within the trust. This was acknowledged as a potentially serious problem:

Junior doctors not feeling empowered, not feeling able to challenge their seniors – that’s a problem. If that led to an occasion where a junior doctor had a concern about patient care or about safety and they felt unable to say because of perceived repercussions that’s a serious issue.

In moving forward the board was described as ‘significantly sighted’ on the junior doctors’ issue. This translated into the chairperson and the governing body implementing an ongoing review of this issue with the aim of developing long-term cultural change. The executive board was also reviewing junior doctors as a standing item at the HR subcommittee, which had developed an action list of 10 things that needed improvement. Agreement was also secured from the MDs that this was something that they would support within their board. Progress would be reported on after a year. Although it was stated that the action plan was getting stuck on administrative issues, progress was being made:

It’s been very trustee proactive at a trust-wide level but firefighting in individual departments. Hopefully in the future we’ll move to a point where it will be very proactive on both fronts but at the moment that’s pretty much how it is really.

The foundations of performativity: Austin and Goffman

Austin105 coined the neologism performativity to describe those instances in which the utterance of a phrase constitutes an action which changes reality rather than describes it; a simultaneous ‘saying’ and ‘doing’ with creative potential that, once the intention of the utterance is recognised, requires others to act in accordance with its implications. Austin’s paradigmatic case is the phrase ‘I do’ when spoken within the context of a marriage ceremony; a phrase with implications beyond indicating acceptance and requiring the people exchanging the vows to act, and be acted on by others, as a married couple from that point forward. Austin additionally stipulates that performative utterances are meaningful actions that are neither true nor false; rather, they create a social reality.

Although Austin provides a working definition of performativity, Goffman’s106 dramaturgy of social interactions applies a theatrical metaphor to indirectly explore its operation, principally through framing. For Goffman, the framing of the stage – the separation of front of stage, backstage and audience, and the reciprocal acceptance of the roles of audience and players – shapes the performativity of utterances onstage. In an extension of earlier work on the presentation of self,119 he applies this metaphor to the operation of social life, in which performativities are made possible through framing. Frames are essentially classification systems that actors use to order and make sense of diverse social phenomena, so that the performative potential of an utterance depends upon the frames available to actors within specific settings.

The performativity of governance: Hajer

Applying Goffman’s work106,119 to the arena of public governance, Martin Hajer33 and others have advocated the placing of a greater emphasis on the symbolic and performative dimensions of governance, focusing attention on the setting(s) in which deliberation takes place and the norms expressed during the process. Organisational life is conceived as a series of staged performances in which interactions of scene and acts produce a particular staging of decision-making. 33,34 Hajer33 operationalises Goffman’s approach, outlining a dramaturgical governance framework for the analysis of the performative dimension of board governance. Practice is explored through consideration of the setting(s) in which deliberation takes place and the enactment of organisational frames, operationalised through a consideration of the scripting, setting, staging and performance of board governance.

Scripting refers to the determination of actors involved in the decision-making forum and, by implication, those who are excluded. Consistent with the generative potential of performativity, it considers the way in which participatory practices construct participants as either active or passive, collaborators or protesters and competent or incompetent. In contrast, setting concerns the physical environment in which interaction takes place, additionally including artefacts (e.g. minutes of previous meetings, tabled reports) that participants bring to the physical environment and which shape the performance. Deliberate attempts to organise the interaction between participants is identified as staging. It is achieved by drawing on existing symbols and the invention and use of new ones, together with conventions governing distinctions between active players and passive audiences; what might be termed the unwritten rules of engagement. The final category, performance, concerns the way in which interaction constructs new knowledge/understandings/power relationships that shape future interactions and provide opportunities for change over time.

Crucially, if Goffman is correct that reality is mediated through the application of frames to make sense of the mass of available information, then it follows that all accounts of reality are shaped. The implication then, in the context of patient safety governance, is that the examination of safety could be framed in many ways with radically different consequences for action. For example, in presenting data that indicates a breach of an infection control performance target, is this an instance of unreliable data requiring a defence of organisational practice to an external regulator, a worrying event requiring detailed diagnostic work to uncover systemic problems and institute quality improvement activity or an impossibly harsh target imposed by regulators that needs to be robustly challenged?

Given the above framing, we exploited the opportunities presented by our observational data to examine the dramaturgy of board performance(s).

Arran: a world-class provider

Arran has a strong research culture and a global reputation for service innovation and development. The board acts strategically and is externally oriented, seeking to secure compliance with central policy directives while safeguarding and further developing its global reach and status as a pre-eminent provider and academic research centre. In operation, board meetings are highly formal and structured to enable the CEO to ‘steer’ interpretations of events. The executive team are experienced, with low turnover and a strong medical presence. NEDs exhibit very low levels of challenge, even when performance difficulties are under discussion; indeed, the dynamic of board meetings could be summarised as maintenance of the narrative of ‘world class’ status while engaging with potential performance difficulties evident in summary indicator data.

Skye: a local service under pressure

In marked contrast, Skye presents as a trust with high levels of turnover in the executive team in recent years and overt recognition of many problems in service provision. This had led to a focus on internal problems rather than the development of organisational strategy in relation to the wider external environment. Board meetings were steered strongly by the chairperson, who dominated a newly appointed CEO. Strong challenges from NEDs were common, and included expressions of disappointment at poor service performance, as well as strong support for those executives considered to be performing well. During the period of observation, one of the executives, who was considered to be performing poorly, stepped down, to be replaced by a deputy who received similarly high levels of challenge. The dynamic of the board could be summarised as routine non-executive challenge with limited long-term strategic direction.

Lewis: an embattled regional powerhouse

Protective of its reputation as a regional ‘leading light’, board meetings were structured to ensure that issues of strategic importance to the trust received due consideration and events were considered as they unfolded over successive meetings. The image of the trust revealed through board operation is that of a guardian of the interests of the locale and wider region, requiring continued political influence to secure those interests. The trust has an experienced and longstanding executive team; indeed, the CEO personifies the organisation, dominating board meetings and acting as a political antagonist to defend local interests. During observation the majority of time was spent on a small number of financial and estate issues and related legal disputes. NEDs were typically involved in clarifications related to their specific areas of interest rather than challenges over strategy or performance. Overall, the board dynamic could be characterised as one of defending its status as regional champion.

Islay: faith in quality improvement methodologies

The overwhelming impression of this site is of one in which board structures and processes focus attention on the reconfiguration of integrated care strategies through the lens of service quality improvement defined as ‘patient experience’. Central to this approach was the consideration of performance issues related to external regulatory requirements (e.g. infection control, reduction in incidence of pressure sores), and these were directly linked to improvement work to support the strategic plan. Patient experience was routinely invoked and normalised by the presentation of a patient experience narrative at the start of each board meeting. NEDs were encouraged to question and the executive team actively sought their contributions. This board could be characterised as drawing on a wide range of soft and hard intelligence, seeking to maintain strategic focus while discharging external regulatory accountabilities.

Scripting

Sites demonstrated distinctive variations with regard to scripting. Although often commenting on information presented to the board, at Arran NEDs were limited in their challenges to executives; collectively, the tone was one of ‘low challenge’ without overt conflict. This was achieved principally through agenda shaping by the CEO, who used his ‘CEO report’ agenda item to reframe potential performance issues.

In marked contrast, the chairperson and NEDs at Skye confidently challenged executives, particularly in relation to poor performance in infection control and capacity/demand issues, challenging the CEO to improve standards and the finance director to increase investment in capacity. There was clear differentiation in the way that executives were seen and treated. The MD was widely seen as competent and dependable, with NEDs formally acknowledging his pathway development work. In contrast, the ND received tougher questioning and her hesitancy was drawn to attention. In this manner, executives were either lauded or shamed, and such public shaming had a negative reputational impact within subsequent board meetings.

At Lewis, board meetings could be summarised as ‘the CEO show’, characterised by political statements defending organisational interests. There was much deferral to his experience, political influence and judgement. Central to this performance was the crafting of a narrative of an embattled trust fighting for local interests. Board meetings gave the impression of a trust under fire, enmeshed in structural, legal and financial disputes. The CEO’s views were rarely challenged, in large part because the board supported his position, but also because they accepted that his performance was part of the ritual, crafted with humour, emotion and political point scoring. In this context, the chairperson had a supportive presence, summarising issues as appropriate.

At Islay the board maintained a clear focus on strategy and service improvement, with positive roles for each member. The CEO projected a calm manner, as a moderniser and believer in quality improvement methodology, with a similarly oriented executive team. The CEO had heavily invested in board development in relation to quality improvement and this training clearly informed board practice. Robust yet respectful challenge was the norm among NEDs, framed and legitimated as ‘improving patient experience’. Board dynamics were driven by this central narrative, which granted permission for depersonalised challenge.

Setting I: physical

Board meetings at Arran were held within the education centre at trust headquarters in a large room with high-quality presentation facilities used in formal presentations by invited speakers to address specific agenda items. The location reinforced a sense of ‘information transmission’ rather than deliberation: a passive process. The room was large, bright and airy and could have accommodated up to 30 observers with ease. The audience for the public board meeting typically consisted of five governors and/or members of the public, ordered into rows of (sparsely populated) chairs facing presentation screens and set back from board members, who were themselves seated in a ‘horseshoe’ arrangement of three tables at the front of the room; a non-adversarial framing consistent with a ‘common purpose’.

Meetings at Skye were held at different locations across the trust. With one exception, meeting rooms were poor environments in which to facilitate discussion: cramped, hot and lacking air circulation. Because of the cramped conditions, the audience (3–5 governors) tended to split into two groups at opposite sides of the room, very close to the board table. The chairperson, CEO, trust secretary and trust minute-taker always sat together, with other seats haphazardly assigned with no consistency across meetings.

Board meetings at Lewis were held at the boardroom at the main trust site, whose décor, original to the building, consisted of dark-wood tables and chairs with red leather covers: an imposing environment reflecting civic pride. The seating arrangement was stable across meetings. The chairperson, CEO and observers sat together, the business development director and NEDs together on an adjacent table to the ND and finance director, with some spaces for guest speakers and observers. The trust secretary, MD and more NEDs sat at an adjacent table. One important feature was that those invited to present to the board faced an experience akin to an inquiry or investigation; waiting in an anteroom for the appointed time on the agenda, invited in when required and positioned on a table in front of the board to face questions.

Although the venue for Islay’s board meetings rotated, they were always held in well-furnished, low-key ‘office’ environments. Rooms were spacious and able to accommodate observers as required. Seating was arranged around a large single table on an ad hoc basis. Attendees included members of the QI team and clinical leads who attended to report updates regarding specific clinical reviews.

Setting II: artefacts

All sites provided board members with ‘board papers’, a briefing pack for information consisting of a formal, standardised agenda and appended reports for each agenda item in sequence, ahead of the board meeting. The format and contents was very similar across sites. Two frames of reference were identified that underpinned these board papers: RAG ratings and ‘narrative’ executive summaries. Patient safety performance information contained within the board papers was assigned red, amber or green (RAG) ratings to document performance. Additional quantitative material would include line graphs, along with run charts and bar charts to present performance. In addition, board papers presented information in the form of an executive summary of the ‘key issues’ over the past month and the options for moving forward.

At some sites, notably Arran, special presentations were typically undertaken by medical staff in response to specific issues highlighted within previous board meetings. The format of these presentations broadly adhered to conventions associated with academic medical lectures: presenters using the electronic facilities available to display detailed presentations, summarising the main points of associated reports appended within the compendium of board papers and then responding to questions from the floor. Presentations anticipated likely questions and were generally of a high standard. While affording non-executive board members the opportunity to raise detailed questions, the presentation format provided the opportunity for prior framing material to create the desired effect.

Although similar conventions operated at Skye, the presentations were generally more informal in tone and lacked vibrancy. They typically afforded NEDs the opportunity to raise questions in relation to detail rather than broader framing.

Staging

Very formal in operation, the ordering of the agenda at Arran was significant in affording the CEO scope to craft an over-arching narrative that reinforced the perception of the organisation as a high performer. A separate ‘performance report’ delivered by the director of strategy provided a high-level summary and narrative overview of key performance indicator (KPI) data related to activity, efficiency, access, cancer, infection, quality and safety, workforce, finance and the Monitor compliance framework. Performance data were presented for the current month and year to date, disaggregated by subdivision (medicine, surgery and cancer and specialist) on a range of indicators for each of the performance areas listed above, summarised on a single A4 page, with additional supporting data for each performance area on subsequent pages. Presentation adopted a traffic-light system in which performance against each indicator is colour coded: green for ‘good’, amber for ‘attention’ and red for ‘breach’. This summary of performance acts as a strong visual cue of actual and potential performance shortfalls, drawing the eye to patterned clusters of ‘red’ and ‘amber’ on the page. Consequently, board attention was often drawn to, and time was spent on, those performance indicators with ‘red’ and/or ‘amber’ status. Executive behaviour at the board seemed strongly influenced by the need to publicly demonstrate performance against standards as indicated by the performance indicators (‘conformance’), and account for any observed shortfall.

The structure (KPIs) and form (‘traffic-light’ RAG indication of performance against targets) of reporting used to inform the board potentially facilitates the discharge of public governance in the form of assurance (‘conformance’). Although such a system could be used by non-executives in a bullish manner to vociferously challenge executives on performance shortfalls, this was not done at any of the observed board meetings. Rather, the CEO drew attention to shortfalls in his CEO report, recontextualising the figures and outlining ongoing work to diagnose and/or address shortfall. Non-executives seemed generally supportive of these reassurances.

The CEO’s report was wide-ranging, provided opportunities for clarification and discussion and, consequently, took a considerable amount of time. The CEO used this composite section, delivered very early in the board meeting, to consider multiple items related to trust activity/performance and its implications, in the light of external policy directives and/or the strategic opportunities they provided. Examples include Dr Foster performance rankings, Monitor KPI governance ratings, CQC visit outcomes and implications for the trust of the Francis Report into patient safety. This section was not delivered as a monologue – rather, the CEO outlined his perspective and invited others (typically executive members, also notably the chairperson) to comment in the light of this analysis and/or to clarify issues. Although there was opportunity for NEDs to ask questions during proceedings, contributions invited by the CEO came from executive board members, with minimal NED input.

In contrast to the close executive control exhibited above, at Skye the staging was typified by the chairperson’s dominance of the agenda. The chairperson exercised his non-executive role in challenging the executive and largely supporting the views of other non-executives. His steering of the agenda, and of contributions from other non-execs, focused principally on the breach of targets (particularly C. difficile infections) and issues related to coding and validity of information provided, with very limited external horizon scanning or evidence of facilitating strategic thinking. The chairperson was active throughout, but particularly in relation to quality and patient safety, as evinced in steering challenge to the ND about infection control and the failure to achieve the C. difficile infection target. There was a sense (picked up by an audience member in an aside to the researchers) that he dominated the meeting; by controlling the staging and encouraging the NED performance the chairperson prevented other narratives from unfolding.

Staging at Lewis was shaped by an extensive ‘matters arising’ section, placed immediately after the opening item of ‘apologies’. This was by far the longest section and invariably consisted of the following issues considered most important: a major capital investment project, ‘safe and sustainable’; paediatric cardiac surgery; any Qualified Provider procurements; an Academic Health Science Network and commissioner penalties in relation to C. difficile infections.

Board meetings at Islay were strategically focused and tightly structured around the agenda. The tone of each meeting was set by an opening ‘patient story’ narrative, introduced by the chairperson as ‘something to focus our minds’. These narratives, collected by the improvement team and presented in the patient’s own words, concerned issues such as the fears, anxieties and negative experiences following diagnosis, processes of care and treatment outcomes. They were emotive and used by the chairperson to legitimate challenge and concern with service improvement, and this language of ‘patient experience’ was available to, and used by, board members to depersonalise challenging questions. The strategic focus on service improvement was also supported by the structure of the agenda, which considered ‘service improvement’ to be a standing agenda item placed immediately after the ‘patient story’ opening, and led by members of the quality improvement team presenting data and updating progress on strategically important projects, such as intensive care, community services, falls and infections. The incorporation of indicators for infections into this section is important, as it ties external regulatory requirements to internal strategy in an agenda item specifically oriented to service improvement.

Performance(s): board governance of infection control

Board meetings at Arran were dominated by the CEO and typified by a need to present trust performance in a favourable light, consistent with a ‘world-class’ reputation. This dominance was not exercised through explicit force of personality, but principally through careful structuring of the agenda and the use of framing in the ‘CEO report’ agenda item to shape perceptions of, and thus actions in relation to, items which appeared later in the agenda.

As outlined above, the CEO’s report often drew attention to shortfalls in performance. Notable examples in relation to patient safety included failure to achieve reduction in C. difficile infection targets for three quarters in succession and a potential breach of A&E waiting time target in quarter 4 (and thus for the whole year). Although both of these topics could have elicited vociferous challenges from NEDs, no such challenges were made. Rather, in relation to the former, the CEO announced that he had successfully argued with commissioners for a waiver of penalties for 10 cases, and once this had been agreed he argued with the relevant external regulatory body for reclassification as ‘green’ (healthy) under their governance performance indicator, which had been flagged as amber, placing the trust under external scrutiny. This was presented by the CEO as evidence of oversensitivity to borderline non-compliance. His analysis drew support for an internal review of antibiotic prescribing habits (supported by comments from a NED and one of the MDs). In the case of the A&E issue, the seriousness of the potential breach was minimised by reference to the ubiquity of such problems nationally (‘everyone faces such difficulties’), addressed with reference to current plans to expand A&E provision (‘we are already dealing with this’) and attention focused on the role of limited out-of-hours cover by GPs in increasing demand on A&E services (‘the problem lies elsewhere’).

In both instances, the CEO’s report identified poor performance against indicators that could have been addressed within the latter ‘performance’ report section as ‘breaches’ and subtly interpreted the data in a misleading way. That these analyses were convincing is clear from the positive reception they received and that fact that neither of these issues were referred to in the latter presentation of the ‘performance’ report or picked up by NEDs for further exploration. Given the important work involved in framing these issues in such a way that possible poor performance is acknowledged, while simultaneously offering reassurance that ‘all is well’, the CEO report was typically one of the longer, and on one occasion the longest, agenda item(s), at 46 minutes out of a total of 110 (42% of the total time).

At Skye, board performances were typified by high levels of NED activity. In relation to breach of a C. difficile infection target, the chairperson adopted a tone of disappointment at failure to reach the target, presenting it as a serious issue that needed to be tackled and required explanation by the ND and active discussion by NEDs. Although there is some acceptance that there are widespread difficulties in meeting the target, there is no serious attempt by the chairperson or any executive to reframe the target breach. The ND (and later the acting ND) is held responsible. A ‘red’ trajectory against A&E performance led to NED involvement in discussions concerning historical failures of the board. NEDs also challenged finance report data showing an operating surplus despite underperformance of general medicine and lower than anticipated emergency admissions; a downwards trajectory from amber to red in summary performance data was also used by NEDs to challenge performance.

At Lewis, the CEO’s dominance and use of humour, emotion and politics is exemplified in discussions of a breach of infection control targets. Discussion was dominated by the financial penalties associated with breach of the target, particularly regarding how the trust would pay the financial penalty if its target were breached. As it stood, the trust faced a fine of over £1M per case in breach of the target. Disputes with commissioners ensued and legal proceedings developed. The issue of discussion centred on ‘proportionality’ and perceived unfair treatment in comparison with their regional neighbours, whose targets were considered far easier to reach. Mediation was the next step, with the trust taking the commissioners to court. The CEO suggested they had no choice and he was forced to make a stand.

Although board consideration of infection control at Islay was also framed in relation to external targets, the service improvement orientation of the board was very evident. The MD presented data showing that the trust was meeting its target – although it was close to the limit. This was warmly welcomed by a NED, who praised the effort, and then quickly considered the interventions required to ensure continuous compliance, prompting a discussion of the educational interventions being made to embed infection control in staff behaviour, including associating high-risk cases with reinforcement of requirements to junior doctors.

Reading across the cases then, we see very different emphases and approaches to the performative aspects of governance. These are summarised in Table 6, and further discussed below.

Board size and composition

Given the range and diversity of roles required of NHS trust boards, it follows that they need to be of sufficient size and diversity. The smallest board in our sample had eight members and, in general, boards were tightly clustered in size with between 10 and 15 members. Around 42% of boards had 10–12 members and around 51% had 13–15 members, with the largest board counting only 17 members. We found no significant difference in board size between trusts of different types (FT/non-FT; TH/non-TH).

In our study, around 62% of boards had three or fewer serving board members with clinical backgrounds. For about two-thirds of the trusts (63%), board members with a clinical background constituted < 30% of the board. We found no significant differences between FT/non-FT and TH/non-TH trusts in the distribution of board members with clinical backgrounds. Our findings chime with recent work that focused on NEDs, which found that only half of trusts (52%) had any NED with health-care leadership experience, with the overwhelming majority of NEDs (86%) being drawn from a commercial, financial or managerial background (non-clinical).

Board engagement with patient safety

In this study, boards appear to give considerable time to safety and quality issues. Only one-fifth of trust boards (21%) reported that ≤ 30% of their time was spent discussing safety and quality issues, but only one-quarter (26%) reported that > 60% of their board time was spent on these issues.

Board practices around patient safety

A very high proportion of the trusts in our survey reported the kinds of desirable characteristics and board-related processes that research says may be associated with higher performance (Box 2). On only two of these questions (safety measures in the CEO’s performance review, and formal training for board members on quality and safety) did affirmation fall at or below 90%. There was a small (but still non-significant) difference between FT/non-FT boards with respect to board members receiving formal training in relation to patient safety, with 90% of FT boards versus 83% of non-FT boards reportedly receiving training. Other differences between trusts of different types on the raft of measures in Box 2 were minimal. These generally high (or very high) ‘desirable responses’ by trusts may represent considerable board attention being paid to quality and safety, especially in the light of scandals such as that of Mid Staffordshire NHS Foundation Trust, and/or they may reflect a growing awareness of the importance of signalling that such issues are being taken seriously. These uniformly high affirmative response rates made correlational analysis of any relationships between board responses in these areas and organisational outcomes inappropriate.

Board use of information on patient safety

In our national survey we found that hospital boards were using a wide range of hard performance metrics and soft intelligence to monitor their organisation with regard to patient safety. Quantitative data, including a range of clinical outcomes measures, infection rates and process measures, such as medication errors and readmission rates, were reportedly used at every board meeting across most hospital trusts (> 80%). It is noteworthy, however, that fewer trusts reported routine reporting of patient safety surveys (81%) and implementation of patient safety alerts (79%). A much smaller proportion of trusts (57%) routinely report morbidity rates at every board meeting. There was no appreciable difference between FT/non-FT boards and TH/non-TH boards on most reporting of explicit performance metrics except for ‘formal morbidity reporting’, for which 52% of non-FTs compared with 60% of FTs routinely presented these at board meetings.

There were some (generally small) differences between FT and non-FT boards, and also between TH and non-TH trusts, in their use of softer intelligence, but these differences were not marked and nor were they consistent. For example, the difference between FTs and non-FTs in their use of patients’ stories was 8% (80% vs. 88%); the same difference seen in their use of board members engaging with clinicians (91% vs. 83%), but with the direction of difference reversed.

In general, then, despite a very high level of desirable characteristics being reported (see Box 2), there remained considerable variation in the use of diverse sources of hard and soft information. Explicit metrics (such as morbidity rates), in particular, were used only intermittently in a considerable number of trusts.

Testing the structure of the Board Self-Assessment Questionnaire

The BSAQ online survey was conducted over two periods: between 21 May and 20 August 2012 (phase 1) and between 30 January and 5 April 2013 (phase 2). A total of 334 responses were received from 165 executive and 169 non-executive board members representing 95 NHS trusts (of the 144 that were sent the BSAQ questionnaire). Of these 95 trusts, 51 are FTs. Board composition by foundation status is shown in Table 8.

Table 8 shows that non-FTs had a slightly higher response rate and a slightly smaller board size, but a higher proportion of clinicians on the board. However, differences were not found to be statistically significant.

The 65 questions on the BSAQ are scored by respondents on a 4-point Likert scale as one of strongly disagree, disagree, agree or strongly agree. Each of the questions corresponds to one of the six dimensions of board performance.

Factor analysis

We perform factor analysis at the respondent level to explore underlying factors that characterise board performance in each of the six dimensions. We find that there is a single underlying factor for each dimension that can be expressed as a BSAQ score. This continuous score summarises the inter-relationships among the ordinal 4-point scale questions in the dimension. A total BSAQ score is also derived and is used to explore whether or not there are associations between board performance and patient safety measures. Specifically, our analysis comprises the following steps:

1. We carry out ordinal factor analysis on each of the six dimensions of board performance individually. Based on the extracted factors, for each respondent we construct six scores, one for each dimension.

2. We apply factor analysis on the factor scores representing the six BSAQ dimensions to construct a total BSAQ score.

3. We explore correlations between the constructed BSAQ scores and a number of patient safety measures.

4. We estimate multivariate models regressing patient safety measures and measures of hospital ability to handle errors, near-misses and incidents on the total BSAQ score and a number of hospital-level characteristics.

Purpose of factor analysis

Each BSAQ dimension consists of a number of questions (items). Factor analysis assumes that the variance in a given item can be explained by one or more common underlying factors and by variance that is unique to the item. The amount of variation in the item that is caused by the common factors is known as communality. Our purpose is to establish whether or not each of the pre-determined BSAQ dimensions is unidimensional. The process, known as factor extraction, is iterative. First, the squared multiple correlation for each item (a measure of how much variance an item shares with the other items) is initially placed on the diagonal of the correlation matrix to approximate communalities (how much the items share with the factors). Then factors are extracted successively, with the first factor accounting for the maximum amount of common variance. The second factor is extracted from the residual correlation matrix after factoring out the influence of the first factor, and the process continues until some criteria are met.

Revealing the latent factors

In factor analysis, eigenvalues represent the amount of common variance among all items that is explained by a particular factor. The larger the eigenvalue, the more total common variance in the items that is explained by the factor. Table 9 presents six sets of eigenvalues obtained from factor analysis performed in each of the six BSAQ dimensions. In the first dimension, the contextual dimension, there are six negative eigenvalues, which are not acceptable in factor analysis and, therefore, only factors 1–6 are further examined for this dimension. Similarly, in the other dimensions, only the positive factors are examined. The amount of common variation explained by the first factor is 3.30, significantly larger than the variance explained by the other factors. One of the criteria for determining the end of the extraction process is to select only those factors for which the eigenvalues are > 1 (Kaiser–Guttman criterion). In all six dimensions, only the first factor meets this condition. Although the criterion is criticised as being inappropriate for factor analysis, two other commonly used criteria confirm the existence of a single factor within each dimension. Unlike principal component analysis, the eigenvalues in factor analysis do not sum up to 1; therefore, values > 1 do not necessarily imply that the factor accounts for more than its share of the total variance in the items. 124 The percentage of variance extracted by the first factor exceeds 85% in all six dimensions (much higher than commonly used thresholds in social sciences).

Scree plots125 identified in all cases show distinct breaks between the slope of the larger eigenvalue (which corresponds to the first factor) and the trailing off of the other eigenvalues, suggesting the use of one factor. Figure 2 shows the scree plot for the contextual scale as an example.

FIGURE 2.

Scree plot for the 12 items of the contextual scale.

The results of finding a single factor for each of the six BSAQ dimensions supports the validity of the BSAQ dimensions and is also found in previous studies. 29,42,44,45

Contribution of items in the common factor

Each of the dimensions in BSAQ has a number of questions that load onto that dimension; for example, there are 12 questions/items in the contextual domain, but only eight questions/items in the political domain. The factor loadings presented in Table 10 indicate the extent to which each of the items correlate with the underlying factor. The square of an item’s loading represents the item’s communality. For instance, 26% of the common variance that item 1 shares with the other 11 items within the contextual domain can be explained by the common factor. In all six dimensions, items load reasonably strongly, exceeding the commonly used threshold of 0.3 for communality,124 with only few exceptions.

Scores

We generate a score on the common factor for all respondents. We do this for all six dimensions, following two approaches.

The more sophisticated approach is to estimate factor scores. In the context of BSAQ, an estimated factor score is a linear composite of the optimally weighted individual responses on the items that comprise a particular dimension. In common factor analysis, there is no unique solution to estimate factor scores. 126 We apply the following method, which is implemented in SAS version 9.2 (SAS Institute Inc., Cary, NC, USA), to calculate factor scores.

1. All of the individual’s responses on the items are standardised to form a new k × 1 vector, where k is the number of items.

2. The PROC FACTOR procedure in SAS coupled with the SCORE option produces a k ×1 vector of factor score coefficients following a multiple regression approach. These coefficients are the weights to be used in step 3.

3. The standardised individual’s scores on the items are weighted by the generated factor score coefficients and summed across all items to produce the individual’s factor score. This is a standardised score with a mean of zero and a variance equal to the squared multiple correlation between the estimated factor scores and the true factor values (which are not known).

The second approach is to create factor-based scales. This less sophisticated method does not use weights but has the advantage of allowing use of only the items that load satisfactorily on a factor. Because all items in BSAQ dimensions load reasonably well, this is not relevant for our analysis.

Construction of factor-based scales is done as follows. A respondent’s score on the factor-based scale is the sum of the individual’s (unstandardised) responses to the items. Because all the items are measured on a 4-point Likert scale, the possible range for the 12 items in the contextual dimension is 12–48, for the 10 items in the analytical dimension it is 10–40, and so forth. Table 11 shows that factor-based scales do not differ by foundation or teaching status. We will not use further factor-based scales, as they do not take into account item loadings.

Total Board Self-Assessment Questionnaire score

To develop a composite BSAQ score for all 65 items we performed factor analysis on the factor scores representing the six BSAQ dimensions. One single factor emerges and Table 12 shows the factor loadings. All dimensions load very strongly on the common underlying factor. About 67% of the common variance that the contextual dimension shares with the other five BSAQ dimensions is explained by the common factor.

Ordinal factor analysis

Factor analysis applied to ordinal data is criticised for using Pearson’s correlations and therefore relying on the assumption that the variables are continuous and follow a multivariate normal distribution. A more suitable approach might be to perform factor analysis on the matrix of polychoric correlations, which can be thought of as correlations between unobserved continuous variables that represent crudely the ordinal variables. We present the results of ordinal factor analysis in Tables 13 and 14. One common factor emerges and the differences in loadings compared with conventional factor analysis are small.

Conclusion regarding the data structure of the Board Self-Assessment Questionnaire score

Overall, then, this analysis supports and replicates the original six-factor structure of the BSAQ and provides reasonable grounds for using the dimensions described with English NHS boards.

Relating Board Self-Assessment Questionnaire scores to patient safety parameters from routine data sets

After constructing the factor scores representing the six BSAQ dimensions and the total BSAQ score, we explored whether or not these are correlated with patient safety measures. We found no strong correlations between any of the six BSAQ dimensions or the total BSAQ score and the various patient safety measures.

Regressions

We estimate multivariate and hierarchical models regressing patient safety measures (see Appendix 7 for a full description of the patient safety measures) and measures of hospital ability to handle errors, near-misses and incidents (taken from the NSS for 2012) based on the total BSAQ score. We control for a number of hospital-level characteristics.

The dependent variables are listed in Table 15 (patient safety measures) and Table 16 (measures of provider ability to handle errors).

We also explored the option of examining patient safety indicators published on the NHS Choices website (www.nhs.uk/NHSEngland/thenhs/patient-safety/Pages/patient-safety-indicators.aspx). These cover measures of safe staffing, infection control and cleanliness, open and honest reporting, patients assessed for risk of blood clots, NHS Safety Thermometer data on pressure ulcers and falls with harm, NHS England patient safety reporting, CQC national safety standards and the percentage of staff who would recommend the hospital, also taken from the NSS.

We did not, however, incorporate these data into the analyses with the BSAQ, as the BSAQ data are reported at trust level and the patient safety data are reported at provider site level. There is a large number of missing data and large variation within trusts across sites in the ratings, making it difficult to aggregate up to trust level.

Measures of hospital ability to handle errors, near-misses and incidents are defined as proportions of NHS staff respondents who responded agree or strongly agree to the following nine statements and questions:

1. My organisation treats staff who are involved in an error, near-miss or incident fairly.

2. My organisation encourages us to report errors, near-misses or incidents.

3. My organisation blames or punishes people who are involved in errors, near-misses or incidents.

4. When errors, near-misses or incidents are reported, my organisation takes action to ensure that they do not happen again.

5. Would you feel safe raising your concern?

6. Would you be confident that your organisation would address your concern?

7. My organisation treats reports of errors, near-misses or incidents confidentially.

8. We are informed about errors, near-misses and incidents that happen in this organisation.

9. We are given feedback about changes made in response to reported errors, near-misses and incidents.

The set of explanatory variables in the regressions includes the total BSAQ score, FT status and three dummy variables indicating hospital size (small, medium and large). The reference category is acute TH.

Regression results are presented in Tables 17a–e for the patient safety measures and Tables 18a–c for the provider measures of effective management of errors.

Results are not significant for the total BSAQ score in any of the patient safety measure regressions. For a few of the patient safety measures size is significant, suggesting differences in patient safety outcomes between small, medium and large trusts and acute teaching trusts (the reference category).

Two of the variables show a positive and significant association between total BSAQ score (better board governance) and the proportion of staff feeling safe to raise concerns about errors, near-misses and incidents and also feeling confident that their organisation would address their concern. We ran further regressions on the factor scores representing the six BSAQ dimensions for these two variables to explore which aspects are correlated with staff’s ability to raise concerns over patient safety.

Tables 19a and 19b show results for the six BSAQ dimensions against staff feeling ‘safe’ to raise concerns about errors, near-misses, and incidents, whereas Tables 20a and 20b show the regressions for the six BSAQ dimensions against staff feeling ‘confident’ that their organisation would address their concern.

Results are positive and significant for staff feeling ‘safe’ in the contextual, analytical, political and strategic domains, though the effect is strongest in the contextual dimension, which focuses on the board taking cognisance of the organisation’s values and norms; for example, it may support a more open and transparent culture around whistleblowing, thereby enabling staff to feel safe to raise concerns.

Results are positive and significant for staff feeling ‘confident’ in the contextual and strategic dimensions, suggesting that a focus on the organisational values and institutional direction gives staff a sense of security and assurance that patient safety concerns will be dealt with.

Research objective 1: to identify the types of governance activities undertaken by hospital trust boards in the English NHS with regard to ensuring safe care in its organisation

NHS hospital boards as corporate entities have a fundamental strategic role in developing the strategy, systems and cultures for its organisations. These approaches can be ‘hard’, for example the setting and monitoring of polices, standards and procedures, or ‘soft’, for example nurturing appropriate values and beliefs. We uncovered a range of instrumental and symbolic/performative behaviours and activities undertaken by boards in relation to fulfilling its oversight duties with regard to safeguarding care.

In instrumental terms, a ‘successful’ board is one able to take decisions on corporate strategy efficiently and effectively. 19 In the national survey, a very high proportion of trust boards reported the kinds of desirable characteristics and board-related processes that research says may be associated with higher performance, with all having quality subcommittees and proactive procedures in place to address patient safety concerns and almost all having explicit objectives related to improving patient safety. However, given the current focus on patient safety in the NHS, it was surprising that only 87% of trusts reported that board members had actually received any received training in patient safety issues. This, alongside the in-depth case study work, indicates that some members of hospital boards may be struggling to fulfil their roles and responsibilities concerning safety. In this regard it is notable that, in the USA, a number of states have introduced initiatives and legislation around board training, particularly in the form of accreditation, and mandatory induction and orientation programmes for board members. 76 This may also be an area that requires greater attention in the UK.

In the case studies we found that, although each organisation had developed its own unique style of board governance and approaches to safeguarding care, all sought to provide strategic assurance by:

• establishing organisational structures and processes for reporting safety-related information throughout the organisation and to the board

• setting the tone for the whole organisation by making patient safety a strategic priority

• developing and nurturing an ‘open and fair culture’ within the trust which encouraged the reporting of patient safety-related incidents

• using high-level information to ensure that the trust is complying with safe practices and standards and meeting externally led targets.

However, the degree to which these aspirations were fully met was moot, with persistent reports (away from the board) about the risks and challenges of communicating bad news or of challenging senior management perspectives and actions.

The four case studies exhibited features associated with the range of instrumental theories and models of board governance outlined in Chapter 2. However, it was also clear that the boards adopted different models at different times and in relation to particular aspects of board governance. Governance activities, then, were seen to be contingent on a variety of internal factors, such as board dynamics, and on external factors, such as the regulatory environment and commissioning relationships. The case studies (see Chapters 4–7) show evidence of board governance approaches related to:

• agency theory, in the way that boards sought to measure its patient safety performance to ensure compliance and to hold staff accountable for their actions

• stewardship theory, in the way that boards attempted to implement a framework of shared values built on trust–based relationships

• stakeholder theory, in the way that boards have attempted to manage complex trade-offs between stakeholders, including staff, patients and the public

• resource dependency theory, in the way that boards attempted to manage internal and external relationships to leverage influence and resources.

In our four case studies, three of the sites operationalised the governance of patient safety largely in terms of assurance and performance data were used retrospectively to alert the board of poor performance. However, such an approach may encourage under-reporting of patient safety matters and does not provide guidance on how best to address past performance deficiencies. In only one case of our four study sites did the board seek to mitigate the potential for such dysfunctional organisational consequences, through a strategic focus on quality improvement (in this case, even to the extent of structuring the board agenda around strategic quality improvement).

Our analysis of the performative and symbolic aspects of board activities highlights the role and differences in local and ritualistic processes of organising in relation to the governance of patient safety. Although each case study identified different performativities at play in different situations, it was clear that the setting(s) in which board decisions and deliberation take place, and the local enactment of organisational frames (operationalised through different approaches to the scripting, setting, staging and performance of board activity), have an important bearing on how organisations approach the oversight of patient safety. These findings highlight the challenges faced by board members in terms of the artefacts that they have available and the key role of scripting and staging, especially when decisions in these areas often pass unchallenged, unremarked or even unnoticed. A better understanding of these issues may feed into revised training and induction processes for board members.

The amount of time boards devote to discussing patient safety-related issues has been identified as a key issue related to performance in this area. 57 In our study we found that most boards do allocate a considerable amount of time to discussing patient safety- and quality-related issues. The NSS obtained responses which indicate that only one-fifth of trust boards reported that ≤ 30% of their time was spent discussing safety and quality issues. In the case studies we found similar evidence, with only one study site allocating < 30% of their board time to debating quality/safety matters. However, we also identified (through the case study sites) that boards used this time rather differently. For example, for some a wide range of performance data were made available to directors and attention tended to be concentrated closely on those showing shortfalls. Such approaches tended to be quality assurance oriented, with less discussion of proactive quality improvement. In other sites, the emphasis was more on the use of quantitative data to support quality improvement rather than secure conformance against externally set standards.

A key issue in the study was the type and amount of information that board members used to help exercise scrutiny over safety issues. In the national survey we found that hospital boards were using a wide range of hard performance metrics and soft intelligence to monitor their organisation with regard to patient safety, including a range of clinical outcomes measures, infection rates and process measures, such as medication errors and readmission rates.

It is noteworthy, however, that fewer trusts reported routine reporting of patient safety surveys and the implementation of patient safety alerts, and a much smaller proportion of trusts routinely report morbidity rates at every board meeting. Softer intelligence, used organisationally and reported at all board meetings, was also more variably reported, with discussions with clinicians and executive walkabouts being most often reported, alongside use of patient stories. However, in only about two-thirds of trusts did board members shadow clinicians and report back to the board. These findings indicate some scope for broadening and enriching board activities with regard to understanding and assessing patient safety.

In the case studies we also identified a range of information and performance data reported to the board, and surfaced some of the key issues and challenges in their use by board members. For example, where RAG indicator reporting systems were used, concerns were expressed about the validity and reliability of the judgements so presented. In contrast, board members who focused on quality assurance data expressed greater confidence in the validity and reliability of these data. One way out of the apparent tension between assurance and improvement seemed to lie in attempts to combine the requirements of external assurance with ongoing quality improvement through the development of national- and local-level data sets and qualitative information from reports by clinical service areas. Despite these efforts, we heard reports of the need for further training of NEDs to understand and interpret standard quantitative data sets and for further disaggregation of data at ward level.

Understanding of these tensions is augmented by our parallel view of board activities as performative and symbolic. This analysis highlighted the important role of setting, scripting and staging in the use and interpretation of information and performance data by the board. For example, although the availability of summaries of quality indicators to board members was a key issue, so equally were the local processes of organising in each board that made it possible for non-executive board members to use such information to hold executives to account. The setting and staging could be seen as working in support of such questioning or (more often) working against any breaking out from set scripting patterns. Thus, the dramaturgical arrangements of boards require greater attention if they are truly to become fora for effective debate, challenge and the instigation of action.

Research objective 2: to explore the role of boards versus board of governors in foundation trusts with regards to the oversight of patient safety in its organisation

A strong theme from across study sites was that, although the board of governors are generally perceived as well-meaning, they were also considered to be largely ineffective in helping to promote and deliver safer care for their organisations. Boards of governors seemed to serve a useful educative and community linkage role, but with limited challenge or holding of executives to account. Meetings frequently resembled seminars for information sharing, rather than a formal board meeting where key organisational strategies were debated and discussed. Thus, while governors typically reported good relations with the executive board, there was often a lack of awareness and understanding of respective roles. Although some sites did report engagement of governors in quality assurance work in relation to, for example, NED appraisal and committee work, there were also frustrations at what was seen as only limited active engagement from NEDs. In summary, therefore, governor meetings appeared to play little active and positive role in the governance and safeguarding of care across the organisations we studied. This is, of course, not to say that boards of governors may not fulfil other important functions beyond the safety and quality remit of this study.

Research objective 3: to assess the association between particular hospital trust board oversight activities and patient safety processes and clinical outcomes

The national surveys were designed to explore any relationship(s) between board structures, board activities, board competencies (as measured by the BSAQ) and a range of patient safety process and outcome variables. The headline finding is that we did not find any statistically significant relationship between these board attributes and (self-reported) processes and any patient safety outcome measures. However, we did find a significant relationship between two dimensions of the BSAQ and two specific and related NSS organisational ‘process’ measures: staff feeling safe to raise concerns about errors, near-misses and incidents and staff feeling confident that their organisation would address their concerns if raised.

In particular, the correlational results are positive and significant for staff ‘feeling safe’ in the contextual, analytical, political and strategic domains of the BSAQ, although the effect is strongest in the contextual dimension. This context dimension explores how the board takes cognisance of the organisation’s values and norms, and positive orientations here may support a more open and transparent culture around whistleblowing, thus enabling staff to feel safe to raise quality concerns.

Results were also positive and significant for staff ‘feeling confident’ about the contextual and strategic BSAQ dimensions, suggesting that a focus on the organisational values and institutional direction give staff a sense of security and assurance that patient safety concerns will be dealt with.

These findings then suggest some association between board governance and staff ability to address issues around patient safety, but also suggest that board governance may not directly affect patient safety outcomes. To be surer of this last point, however, we would need more complex longitudinal data than were available for this cross-sectional study.

This null finding of the effect of boards on outcomes is at odds with previous empirical work in the USA, which did find an association between board activity and quality outcomes. 17 This difference may be explained in part by a number of factors, not least the focus on different measures of outcome in the US studies and the different economic and political environment in which boards operate in the USA. It may also be the case that there is a lagged effect between current board practice/competencies and patient safety outcomes, which our study would not have picked up as it was cross-sectional rather than longitudinal in design.

The finding linking board orientations to staff attitudes from our large-scale quantitative work are further substantiated when we look at the qualitative findings, especially those that were gleaned from the following of ‘tracer conditions’ from board to ward (see Chapter 6). Here there were persistent worries that calling out quality or safety problems or challenging senior management were unlikely to be welcome actions. Thus, providing an environment safe for challenge remains difficult and may yet be one key area in which boards could have an impact.

The results linking board competencies and organisational policies, associated with the creation of more open cultures and the assurance that patient safety concerns will be dealt with appropriately, are important given the current focus on whistleblowing in the NHS and, in particular, the recently announced independent review into whistleblowing to be chaired by Sir Robert Francis QC, Freedom to Speak Up. Indeed, the terms of reference for the review state that ‘it will consider, in the light of recent policy measures in response to the Public Inquiry into Mid Staffordshire NHS Foundation Trust, what further action is necessary to protect those individual NHS workers who speak out and in doing so make a significant contribution to the open culture that is needed to ensure safe care for patients.’ Later in this chapter, and building on our findings, we outline possible future research into board governance and whistleblowing to explore these issues further.

Research objective 4: to identify the facilitators and barriers to developing effective hospital trust board governance of safe care

Organisational strategies, structures and cultures can potentially serve to enable or attenuate attempts to improve board governance; however, there has previously been little research into this important area of patient safety. In the case studies we explored these dimensions and identified a range of possible organisational facilitators and enablers to effective board governance of safe care. For example, a key facilitator at one of our study sites appeared to have been the development and implementation of a clear corporate strategy and operational plan focused on enhancing the quality and safety of care in the organisation. This plan had been further codified into bespoke performance measures, which were then incorporated in an organisation-wide quality dashboard for board review. Thus, objectives were clearly identified against the plan, which ensured continuous focus, supported by board dynamics characterised as ‘supportive challenge’ (where the open discussion of problematic issues was not interpreted as disloyalty).

At other sites, the stability of board membership over time was considered to be an important facilitator of patient safety governance, and the detailed knowledge of, and connection with, clinical staff by the CEO was thought to be crucial. Strong and committed clinical leadership was also considered to be an important facilitating factor in supporting board governance, even when there were persistent concerns over the so-called bullying of junior professionals by senior clinicians (which were, unsurprisingly, seen as being important barriers to effective governance).

We identified a range of other organisational barriers to effective board governance of safe care across the case study sites. Problems and disputes over the validity and reliability of summary performance indicator data and apparent difficulties in communication between clinical units and the board, for example, led to a sense of confusion over the meaning and the correct interpretation of performance indicator data received and reviewed by the board. In addition, problems were seen to arise because of a perceived lack of engagement among senior medical staff in many parts of the organisation, resulting in impaired communication between corporate teams and ward level.

Research objective 5: to assess the impact of external commissioning arrangements and incentives on hospital trust board oversight of patient safety

Hospital trusts operate within a broader system of incentive and constraints, which necessarily help to shape and influence approaches to board oversight of safe care. In terms of the commissioning environment, there is currently a policy focus on the role of CCGs as a means of improving quality and patient safety through the design of contracting and payment systems, as well as through establishing collaborative relationships with providers. Across the case studies we found that contracting issues appeared to play only a relatively minor role in incentivising quality and safety improvement. Building strong and trustworthy relationships between commissioners and providers was often viewed as a better way of stimulating positive change than focusing on incentives and sanctions. Relationships with commissioners across all four of our study sites were seen as largely positive, notwithstanding a few local difficulties (e.g. around plans for expansion and competition with other providers).

External financial incentives also have the potential to influence board oversight strategies. The CQUIN payment framework makes a proportion of providers’ income conditional on locally agreed goals concerning quality (covering dimensions of safety, effectiveness and experience). During the period of the research the incentive was 0.5% of contract value. The aim of the CQUIN is to support a cultural shift by embedding quality improvement and innovation within commissioner-provider discussions. The CQUIN appear to have had a mixed impact across the case study sites and were generally perceived to be a low-powered incentive for quality and safety improvement, as well as possibly inducing a range of unintended and dysfunctional consequences. For example, the CQUIN was perceived by some to be too unwieldy in operation to act as effective incentives and were negatively perceived as ‘taking money off’ rather than incentivising improvement activity. Reactions were not uniform, however, with the CQUIN being perceived as having a (albeit limited) positive impact in some areas but being a distraction in others.

Although compliance frameworks were generally welcomed for their effect of focusing attention on aspects of patient safety, concerns were raised in relation to their negative unintended consequences, notably ‘tunnel vision’ (a reduced attention and focus on important areas of health-care quality and patient safety that were not covered under the CQUIN framework, e.g. environmental health). The CQUIN penalties also proved difficult to apply in practice, with formal sanctions becoming a subject of negotiation between provider and commissioner. Therefore, external financial incentives in the form of the CQUIN appeared to have had little positive impact on board governance and safeguarding care in the case study organisations, and may even have proved an unwelcome distraction.

Research design

As with all research, a rigourous study design enhances the power and credibility of the findings and future research in this area would benefit from a number of features:

• Any studies should be mixed-method and multidisciplinary in design, drawing on a range of quantitative and qualitative methods, including detailed ethnographic approaches.

• As the phenomena under investigation are essentially dynamic (governance, performance and change), longitudinal study will offer important additional insights over cross-sectional designs. Although we achieved excellent response rates, the national surveys of board structure, activity and competency (the BSAQ), which were linked to patient safety-related process and outcomes, used a cross-sectional rather than a longitudinal research design. This means that we can only infer associations rather than estimate causality between the variables. Longitudinal research using repeated cross-sectional data would help to unravel any causal relationships.

• It is plausible that existing board activity and competencies may take time to impact on patient safety outcomes; therefore, there may be a lagged effect. However, in our study the board-level variables were linked with relatively contemporaneous patient safety variables and so may have failed to pick up any lagged associations. Future research might therefore explore such lagged effects, using patient safety data in the form of a time series.

• In our study we have shown how both instrumental and symbolic theoretical approaches can yield important insights into board behaviour and provide better opportunities for theoretical transference and generalisability to other contexts. Future research will need also need sound conceptual underpinnings rather than relying on empiricism alone.

Research on board governance, whistleblowing, employee silence and voice

Our finding that board governance/competencies appear to be linked to staff feeling safe to raise concerns about patient safety issues, and also their confidence that their organisation would address their concerns, is worthy of further and more sustained exploration, particularly in the context of the current focus on improving whistleblowing policies in the NHS. Whistleblowing (the disclosure, either to a person in authority or in public, of information concerning unsafe, unethical or illegal practices) is central to current debates about addressing poor standards of care in the NHS. Since the 1998 Public Interest Disclosure Act, all NHS trusts have been required to have policy and procedures in place for dealing with whistleblowing. This is now enshrined in the NHS Constitution, which mandates ‘an expectation that NHS staff will raise concerns about safety, malpractice or wrong-doing . . . as early as possible’. 128 Yet a recent House of Commons Public Accounts Select Committee report on patient safety concluded that ‘the NHS remains largely unsupportive of whistleblowing, with many staff fearful of going outside official channels to bring unsafe care to light’. 129 Indeed, the Francis Inquiry noted that there had been a number of whistleblowing policies in place at Mid Staffordshire NHS Foundation Trust that had ‘the clear objective to empower employees to raise concerns and to ensure that those concerns, where valid, were acted upon’. 4 However, the Inquiry heard that, despite the existence of such policies, no adequate support was given to staff who attempted to raise concerns, with witnesses describing ‘an endemic culture of bullying’, with graphic examples of victimisation of those who did raise concerns. 4 Francis concluded that the culture at the trust militated against health professionals openly raising concerns for two key reasons: they considered that such concerns would not be acted on effectively, or they feared negative repercussions from colleagues and/or victimisation by management. Yet despite these serious concerns about the effectiveness of current whistleblowing polices, we still lack a firm evidence base to guide change.

To help address this gap in knowledge, the government announced an independent inquiry into whistleblowing – Freedom to Speak Up – chaired by Sir Robert Francis QC. 130 The review invited NHS staff to share their experiences of speaking up, as well as ideas for removing the obstacles and creating the right environment for an open and honest reporting culture. The review report calls for all NHS health-care organisations and regulators to adhere to its principles and take actions to develop a safety culture, with calls for the Secretary of State for Health to review the progress at least annually. Within such a context, it will be interesting to see how boards respond to these developments and seek to nurture the cultures of safety and learning advocated by Francis.

Research on the role of systematic group bias in board decision-making

Recent research has highlighted the negative impact of systematic cognitive biases on the quality of clinical decision-making when individuals act alone. Croskerry, in the field of emergency medicine, for example, has documented many such systematic biases and has identified a range of possible remedial strategies. 131 Common cognitive biases that have been shown to afflict individual decision-making in clinical contexts include confirmation bias (where individuals mainly seek out information sources that are likely to confirm, rather than challenge, their existing opinion), hindsight bias (whereby the knowledge of the outcomes of previous decisions erroneously influence the predictability associated with a particular new decision) and the ‘fundamental attribution’ bias that arises from relying on decision problems that are seen as representative of all similar problems.

Boards make many decisions in which such individual biases may be a factor, but they also collectively make decisions that may be open to different biasing processes. 132 Group decisions, too, can be imperfect and result in organisational and clinical errors, which may harm patients. 133

Group decisions typically comprise three sequential stages: orientation (problem definition), evaluation (discussion of decision alternatives) and control (deciding which alternative is selected). Furthermore, each stage may be subject to systematic bias resulting from social influences and group dynamics. However, group-decision-making bias is a hitherto unexplored aspect of board oversight of safe care and it would be useful to undertake empirical research to explore the scope, range and impact of systematic group decision-making bias in board settings on patient safety processes and outcomes. Such work might also explore the potential for ‘de-biasing strategies’ to improve decision-making in this area. Possible systematic biases in group decision-making worthy of exploring include:

• Groupthink. Highly cohesive groups with strongly connected members may inhibit the expression of (true) opinion; in such cases group harmony and unanimity may be privileged over effective decision-making. Clearly, this militates against the delivery of high-quality and safe care, where board members should feel comfortable in expressing dissenting views and be willing to speak openly about concerns they may have regarding the quality of care.

• Group polarisation. Groups sometimes make more extreme (compound) decisions than the initial position of its (individual) members. Extreme decisions are those in which the group makes either a more risky or more conservative decision rather than take a more balanced appraisal or ‘middle of the road’ approach. For example, a hospital board may decide to commit significant resources, incurring high opportunity and sunk costs, to implement a new policy. Conversely, the group may decide not to commit resources to alternative courses of action because of the uncertainty involved. The term ‘risky shift’ is used to describe situations in which a group exhibits more risk-taking behaviour than the initial average risk-seeking tendencies of individual members. ‘Cautious shift’ refers to situations when group action becomes more risk averse than the initial risk tendencies of the individual members. In both cases, the average response of individual group members becomes more extreme when they are part of the group, which may have a negative impact on patient safety in some situations and contexts.

Research on board governance assessment measurement tools

There is a real need for more and better-tested bespoke instruments and tools for assessing and informing hospital board governance and competencies. In the study we used an adaptation of a US measure (the BSAQ), but there is a need for further development in the UK context. As the key characteristics of desirable board governance emerge from further sustained research, we will then be in a better position to build better instruments. Given the dynamic nature and complexity of board governance in health-care contexts, the building, testing and refining of a variety of board tools and instruments will be an ongoing task.

Research on board reach into organisational cultures and practices

As our work on the ‘tracer issues’ (see Chapter 6) demonstrates, a crucial challenge lies in understanding how different levels of organisational hierarchy articulate together. Activity within boards consists largely of discourse-creating dialogue that needs translation into actions at various levels within the hospital. Understanding how the dialogue and discourse creation at one level informs and influences the cultures and practices at another is key, as is identifying and elucidating the key conduits through which these layers are linked. Only detailed ethnographic work, particularly in well-functioning organisations, is likely to shed much light on these complexes processes.

Coding framework

Excerpt from transcript: Code of Internal Oversight and Dynamics (anonymised)

≤ Internals\\Medical Director ≥ – § 5 references coded (12.69% Coverage)

Reference 1 – 3.86% Coverage

I think things have moved on. I think the board . . . Boards are interesting because they . . . This board, the board in this organisation, has three doctors? . . . when I was made director, I had a limited clinical practice. So those people understand all of these aspects of practice. The rest of the board have no more knowledge of it than any layperson coming into hospital for a procedure. I, personally, feel that there should be more board to ward communication. Some of the board members do make time to do walkabouts and that kind of thing – many of them to it quite assiduously . . . does it pretty regularly. Others of them find it very difficult to do that because of external commitments and time constraints. And some of them, I think, don’t actually . . . They’re a bit frightened of it. The way we set it up, I or one of the senior nursing staff . . . would go with them, so they wouldn’t be sent to a ward . . . They’d go with somebody who knew the organisation and knew what questions should be asked and so on. But I think some of them feel out of place. It’s a bit like putting a barrister in an operating theatre – they‘re fine in a courtroom, but you put them in an operating theatre and it’s a different ball game. And it’s the same thing – if I had to go and give evidence in a court of law, the barrister’s very comfortable, but I’m pretty uncomfortable. So it’s the same thing when the board members go into clinical environments.

Competency 1: understands context (contextual)

6. Induction programmes for new board members, particularly non-executives, specifically include a segment about both the NHS and the organisation’s history and traditions.

12. In discussing key issues, it is not unusual for someone on the board to talk about what this organisation stands for and how that is related to the matter at hand.

13. Values are seldom discussed explicitly at our board meetings.

30. This board has made a key decision that I believe to be inconsistent with the mission of this organisation.

37. I have been present in board meetings where discussions of the history and mission of the organisation were key factors in reaching a conclusion on a problem.

39. It is apparent from the comments of some of our board members that they do not understand the mission of the organisation very well.

42. There have been occasions where the board itself has acted in ways inconsistent with the organisation’s deepest values.

45. New members are provided with a detailed explanation of this organisation’s mission when they join this board.

50. This board reviews the organisation’s mission at periodic intervals.

56. One of the reasons I joined this board was that I believe strongly in the values of this organisation.

59. Former members of this board have participated in special events designed to convey to new members the organisation’s history and values.

63. This board understands the norms of the professions working in this organisation.

Competency 2: builds learning (educational)

2. I have participated in board discussions about what we should do differently as a result of a mistake the board made.

10. At least once every 2 years, our board has a retreat or special session to examine our performance, how well we are doing as a board.

17. This board periodically sets aside time to learn more about important issues facing NHS Trusts.

18. I can recall an occasion when the board acknowledged its responsibility for an ill-advised decision.

21. Most people on this board tend to rely on observation and informal discussions to learn about their role and responsibilities.

24. When a new member joins this board, we make sure that someone serves as a mentor to help this person learn the ropes.

26. I have participated in board discussions about the effectiveness of our performance.

29. I have participated in discussions with new board members about the roles and responsibilities of a board member.

34. I have participated in board discussions about what we can learn from a mistake we have made

38. I have never received feedback on my performance as a member of this board.

46. This board does not allocate organisational funds for the purpose of board education and development.

51. This board has conducted an explicit examination of its roles and responsibilities.

Competency 3: nurtures group (interpersonal)

3. I have had conversations with other members of this board regarding common interests we share outside this organisation

15. Differences of opinion in board decisions are more often settled by vote than by more discussion.

20. This board is as attentive to how it reaches conclusions as it is to what is decided.

27. At our board meetings, there is at least as much dialogue among non-executive members as there is between non-executive members and executive members.

31. The leadership of this board typically goes out of its way to make sure that all members have the same information on important issues.

32. This board has adopted some explicit goals for itself, distinct from goals it has for the total organisation.

43. This board relies on the natural emergence of leaders, rather than trying explicitly to cultivate future leaders for the board.

52. I am able to speak my mind on key issues without fear that I will be ostracised by some members of this board.

57. This board does not recognise special events in the lives of its members.

60. This board provides biographical information that helps members get to know one another better.

64. This board meets socially as a group.

Competency 4: recognises complexity (analytical)

1. This board takes regular steps to keep itself informed about important trends in the local health economy, and in the wider national healthcare environment, that might affect the organisation.

4. I have been in board meetings where it seemed that the subtleties of the issues we dealt with escaped the awareness of a number of the members.

5. Our board explicitly examines the ‘downside’ or possible pitfalls of any important decisions it is about to make.

11. Many of the issues that this board deals with seem to be separate tasks, unrelated to one another.

22. I find it easy to identify the key issues that this board faces.

23. When faced with an important issue, the board often ‘brainstorms’ and tries to generate a whole list of creative approaches or solutions to the problem.

28. When issues come before our board, they are seldom framed in a way that enables members to see the connections between the matter at hand and the organisation’s overall strategy.

47. Recommendations from the executive directors are usually accepted with little questioning in board meetings.

53. This board tries to avoid issues that are ambiguous and complicated.

61. This board seeks information and advice from leaders of other similar organisations.

Competency 5: respects process (political)

9. This board communicates its decisions to all those who are affected by them.

14. If our board thinks that an important internal or external stakeholder or stakeholder group is likely to disagree with an action we are considering, we will make sure we learn how they feel before we actually make the decision.

19. This board has formed ad hoc committees or task forces that include staff as well as board members.

25. I have been in board meetings where explicit attention was given to the concerns of the local community.

33. The board periodically requests information on the morale of the professional staff.

41. Before reaching a decision on important issues, this board usually requests input from persons likely to be affected by the decision.

48. At times this board has appeared unaware of the impact its decisions will have within our local community.

54. The executive directors rarely report to the board on the concerns of patients.

Competency 6: shapes direction (strategic)

7. This board is more involved in trying to put out fires than in preparing for the future.

8. The board sets clear organisational priorities for the year ahead.

16. This board delays action until an issue becomes urgent or critical.

35. Our board meetings tend to focus more on current concerns than on preparing for the future.

36. At least once a year, this board asks that the chief executive articulate his/her vision for the organisation’s future and strategies to realise that vision.

40. This board has on occasion evaded responsibility for some important issue facing the organisation.

44. This board often discusses where the organisation should be headed five or more years into the future.

49. Within the past year, this board has reviewed the organisation’s strategies for attaining its long-term goals.

55. I have been in board meetings where the discussion focused on identifying or overcoming the organisation’s weakness.

58. The board discusses events and trends in the larger environment that may present specific opportunities for this organisation.

62. This board makes explicit use of the long range priorities of this organisation in dealing with current issues.

65. More than half of this board’s time is spent in discussions of issues of importance to the organisation’s long-range future.

Metric

The ratio of the observed number of in-hospital deaths with a HSMR diagnosis to the expected number of deaths, multiplied by 100, at trust level.

Denominator

Spells with a primary dominant diagnosis of any of the 56 clinical classifications software groups that comprise the HSMR basket.

Observed

Denominator superspells with method of discharge as death (DISMETH = 4,5).

Deaths are assigned to every provider within the superspell.

Expected

Expected number of in-hospitals deaths derived from logistic regression, adjusting for factors to indirectly standardise for differences in case mix.

Data sources

Secondary user service – commissioning data sets (SUS–CDS) for the financial year 2012/13 (April 2012 to March 2013).

Statistical methods

• Logistic regression is used to calculate the expected number of in-hospitals deaths. To indirectly standardise for differences in patient case mix, the model is adjusted for the following factors:

  • sex

  • age on admission (in 5-year bands up to 90+ years)

  • interactions between age on admission (in 5-year bands up to 90+ years) and Charlson Comorbidity Score

  • admission method (non-elective or elective)

  • socioeconomic deprivation quintile of the area of residence of the patient (based on the Carstairs Index)

  • diagnosis/procedure subgroup

  • comorbidities (based on Charlson score)

  • number of previous emergency admissions

  • year of discharge (financial year)

  • palliative care (if any episode in the spell has the treatment function code 315 or contains International Classification of Diseases, Tenth Edition (ICD-10) (code Z515 in any of the diagnoses fields)

  • month of admission

  • source of admission.

• Relative risk: the ratio is calculated by dividing the actual number of deaths by the expected number and multiplying the figure by 100. It is expressed as a relative risk, where a risk rating of 100 represents the national average. If the trust has a HSMR of 100, the number of patients who died was exactly as it would be expected taking into account the standardisation factors. A HSMR above 100 means more patients died than would be expected; one below 100 means that fewer than expected died.

• Control limits: control limits tell us the range of values that is consistent with random or chance variation. Data points falling within the control limits are consistent with random or chance variation and are said to display ‘common-cause variation’; for data points falling outside the control limits, chance is an unlikely explanation and hence they are said to display ‘special-cause variation’ – that is, where the trust’s rate diverges significantly from the national rate.

Data points falling above the upper 99.8% Poisson control limit are said to be significantly ‘higher than expected’, data points falling below the lower 99.8% Poisson control limit are said to be significantly ‘lower than expected’, data points falling between the lower 99.8% Poisson control limit and the upper 99.8% Poisson control limit are said to be ‘within expected range’.

Metric

The ratio of the observed number of in-hospital deaths to the expected number of deaths for all diagnosis groups, multiplied by 100, at trust level.

Denominator

Spells with any primary dominant diagnosis, linked into superspells.

Observed

Denominator superspells with method of discharge as death (DISMETH = 4,5).

Deaths are assigned to every provider within the superspell.

Expected

Expected number of in-hospitals deaths derived from logistic regression, adjusting for factors to indirectly standardise for differences in case mix.

Data sources

SUS–CDS for the financial year 2012/13 (April 2012–March 2013).

Statistical methods

• Logistic regression is used to calculate the expected number of in-hospitals deaths. To indirectly standardise for differences in patient case mix, the model is adjusted for the following factors:

  • sex

  • age on admission (in 5-year bands up to 90+ years)

  • interactions between age on admission (in 5-year bands up to 90+ years) and Charlson Comorbidity Score

  • admission method (non-elective or elective)

  • socioeconomic deprivation quintile of the area of residence of the patient (based on the Carstairs Index)

  • diagnosis/procedure subgroup

  • comorbidities (based on Charlson score)

  • number of previous emergency admissions

  • year of discharge (financial year)

  • palliative care (if any episode in the spell has the treatment function code 315 or contains ICD-10 code Z515 in any of the diagnoses fields)

  • month of admission

  • source of admission.

• Relative risk: the ratio is calculated by dividing the actual number of deaths by the expected number and multiplying the figure by 100. It is expressed as a relative risk, where a risk rating of 100 represents the national average. If the trust has a HSMR of 100, the number of patients who died was exactly as it would be expected taking into account the standardisation factors. A HSMR above 100 means more patients died than would be expected; one below 100 means that fewer than expected died.

• Control limits: control limits tell us the range of values that is consistent with random or chance variation. Data points falling within the control limits are consistent with random or chance variation and are said to display ‘common-cause variation’; for data points falling outside the control limits, chance is an unlikely explanation and hence they are said to display ‘special-cause variation’ – that is, where the trust’s rate diverges significantly from the national rate.

Data points falling above the upper 99.8% Poisson control limit are said to be significantly ‘higher than expected’, data points falling below the lower 99.8% Poisson control limit are said to be significantly ‘lower than expected’, data points falling between the lower 99.8% Poisson control limit and the upper 99.8% Poisson control limit are said to be ‘within expected range’.

Metric

Deaths per 1000 spells for conditions normally associated with a very low rate of mortality.

Denominator

Spells with a primary diagnosis associated with a low mortality diagnosis group where the mortality rate has been shown to be consistently below 0.5%.

Numerator

Denominator spells with method of discharge as death.

DISMETH:

4. Died.

Data sources

SUS–CDS (April 2012–March 2013).

Statistical methods

• To interpret the performance of a trust on a crude rate indicator, the data point is compared against the national crude rate.

• Control limits: control limits tell us the range of values that are consistent with random or chance variation. Data points falling within the control limits are consistent with random or chance variation and are said to display ‘common-cause variation’; for data points falling outside the control limits, chance is an unlikely explanation and hence they are said to display ‘special-cause variation’ – that is, where the trust’s rate diverges significantly from the national rate.

Data points falling above the upper 99.8% binomial control limit are said to be significantly ‘higher than expected’, data points falling below the lower 99.8% binomial control limit are said to be significantly ‘lower than expected’, data points falling between the lower 99.8% binomial control limit and the upper 99.8% binomial control limit are said to be ‘within expected range’.

Metric

The ratio of the observed number of deaths following surgery with complications of care to the expected number, multiplied by 100.

Denominator

Elective surgical spells, defined by a surgical HRG and an OPCS code for an operating room procedure in any procedure field, for which a diagnosis code indicating potential complications of care is present in any secondary field; and emergency surgical spells, defined by a surgical HRG and an OPCS code for an operating room procedure where the principal operating room procedure took place within 2 days of admission, for which a diagnosis code indicating potential complications of care is present in any secondary field.

Observed

Denominator spells with method of discharge as death.

DISMETH:

4. Died.

Expected

Expected number of deaths is derived from logistic regression, adjusting for factors to indirectly standardise for differences in case-mix.

Data sources

SUS – CDS (April 2012 – March 2013).

Statistical methods

• Logistic regression: expected number of deaths is derived from logistic regression, adjusting for factors to indirectly standardise for differences in case-mix. Adjustments are made for:

  • sex

  • age on admission (in 5-year bands up to 90+ years)

  • admission method (non-elective or elective)

  • socioeconomic deprivation quintile of the area of residence of the patient (based on the Carstairs Index)

  • comorbidities (categorised Charlson Comorbidity Score).

• Relative risk: the ratio is calculated by dividing the actual number of deaths by the expected number and multiplying the figure by 100. It is expressed as a relative risk, where a risk rating of 100 represents the national average. If the trust has a relative risk of 100, the number of patients who died is exactly as it would be expected taking into account the standardisation factors. A relative risk above 100 means more patients died than would be expected; one below 100 means that fewer than expected died.

• Control limits: control limits tell us the range of values that are consistent with random or chance variation. Data points falling within the control limits are consistent with random or chance variation and are said to display ‘common-cause variation’; for data points falling outside the control limits, chance is an unlikely explanation and hence they are said to display ‘special-cause variation’ – that is, where the trust’s rate diverges significantly from the national rate.

Data points falling above the upper 99.8% Poisson control limit are said to be significantly ‘higher than expected’, data points falling below the lower 99.8% Poisson control limit are said to be significantly ‘lower than expected’, data points falling between the lower 99.8% Poisson control limit and the upper 99.8% Poisson control limit are said to be ‘within expected range’.

Metric

The ratio of the observed number of in-hospital deaths following weekend emergency admissions with a HSMR diagnosis to the expected number of deaths, multiplied by 100, at trust level.

Denominator

Emergency spells with a primary dominant diagnosis of any of the 56 CCS groups that comprise the HSMR basket and an emergency admission on Saturday or Sunday, linked into superspells.

Emergency admissions are defined using the following ADMIMETH codes:

21 Emergency: via A&E services, including the casualty department of the provider.

22 Emergency: via GP.

23 Emergency: via Bed Bureau, including the Central Bureau.

24 Emergency: via consultant outpatient clinic.

28 Emergency: other means, including patients who arrive via the A&E department of another healthcare provider.

Observed

Denominator superspells with method of discharge as death (DISMETH = 4,5).

Deaths are assigned to every provider within the superspell.

Expected

The expected number of in-hospital deaths derived from logistic regression, adjusting for factors to indirectly standardise for differences in case mix.

Data sources

SUS–CDS for the financial year 2012/13.

Statistical methods

• Logistic regression: expected number of in-hospitals deaths derived from logistic regression, adjusting for factors to indirectly standardise for differences in case mix. Adjustments are made for:

  • sex

  • age on admission (in 5-year bands up to 90+ years)

  • interactions between age on admission (in 5-year bands up to 90+ years) and Charlson Comorbidity Score

  • admission method (non-elective or elective)

  • socioeconomic deprivation quintile of the area of residence of the patient (based on the Carstairs Index)

  • diagnosis/procedure subgroup

  • comorbidities (based on Charlson Comorbidity Score)

  • number of previous emergency admissions

  • year of discharge (financial year)

  • palliative care (if any episode in the spell has the treatment function code 315 or contains ICD-10 code Z515 in any of the diagnoses fields)

  • month of admission

  • source of admission.

• Relative risk: the ratio is calculated by dividing the actual number of deaths by the expected number and multiplying the figure by 100. It is expressed as a relative risk, where a risk rating of 100 represents the national average. If the trust has a HSMR of 100, the number of patients who died was exactly as would be expected taking into account the standardisation factors. A HSMR above 100 means more patients died than would be expected; one below 100 means that fewer than expected died.

• Control limits: control limits tell us the range of values that are consistent with random or chance variation. Data points falling within the control limits are consistent with random or chance variation and are said to display ‘common-cause variation’; for data points falling outside the control limits, chance is an unlikely explanation and hence they are said to display ‘special-cause variation’ – that is, where the trust’s rate diverges significantly from the national rate.

Data points falling above the upper 99.8% Poisson control limit are said to be significantly ‘higher than expected’, data points falling below the lower 99.8% Poisson control limit are said to be significantly ‘lower than expected’, data points falling between the lower 99.8% Poisson control limit and the upper 99.8% Poisson control limit are said to be ‘within expected range’.

Metric

The ratio of the observed number of in-hospital deaths following weekday emergency admissions with a HSMR diagnosis to the expected number of deaths, multiplied by 100, at trust level.

Denominator

Emergency spells with a primary dominant diagnosis of any of the 56 CCS groups that comprise the HSMR basket and an emergency admission on Monday, Tuesday, Wednesday, Thursday or Friday, linked into superspells.

Emergency admissions are defined using the following ADMIMETH codes:

21 Emergency: via A&E services, including the casualty department of the provider.

22 Emergency: via GP.

23 Emergency: via Bed Bureau, including the Central Bureau.

24 Emergency: via consultant outpatient clinic.

28 Emergency: other means, including patients who arrive via the A&E department of another healthcare provider.

Observed

Denominator superspells with method of discharge as death (DISMETH = 4,5).

Deaths are assigned to every provider within the superspell.

Expected

The expected number of in-hospital deaths derived from logistic regression, adjusting for factors to indirectly standardise for differences in case mix.

Data source

SUS–CDS for the financial year 2012/13.

Statistical methods

• Logistic regression is used to calculate the expected number of in-hospitals deaths. To indirectly standardise for differences in patient case-mix, the model is adjusted for the following factors:

  • sex

  • age on admission (in 5-year bands up to 90+ years)

  • interactions between age on admission (in 5-year bands up to 90+ years) and Charlson Comorbidity Score

  • admission method (non-elective or elective)

  • socioeconomic deprivation quintile of the area of residence of the patient (based on the Carstairs Index)

  • diagnosis/procedure subgroup

  • comorbidities (based on Charlson Comorbidity Score)

  • number of previous emergency admissions

  • year of discharge (financial year)

  • palliative care (if any episode in the spell has the treatment function code 315 or contains ICD-10 code Z515 in any of the diagnoses fields)

  • month of admission

  • source of admission.

• Relative risk: the ratio is calculated by dividing the actual number of deaths by the expected number and multiplying the figure by 100. It is expressed as a relative risk, where a risk rating of 100 represents the national average. If the trust has a HSMR of 100, the number of patients who died was exactly as would be expected taking into account the standardisation factors. A HSMR above 100 means more patients died than would be expected; one below 100 means that fewer than expected died.

• Control limits: control limits tell us the range of values which are consistent with random or chance variation. Data points falling within the control limits are consistent with random or chance variation and are said to display ‘common-cause variation’; for data points falling outside the control limits, chance is an unlikely explanation and hence they are said to display ‘special-cause variation’ – that is, where the trust’s rate diverges significantly from the national rate.

Data points falling above the upper 99.8% Poisson control limit are said to be significantly ‘higher than expected’, data points falling below the lower 99.8% Poisson control limit are said to be significantly ‘lower than expected’, data points falling between the lower 99.8% Poisson control limit and the upper 99.8% Poisson control limit are said to be ‘within expected range’.

Metric

The ratio of the observed number of non-elective admissions with an emergency readmission within 28 days of discharge to the expected number, multiplied by 100.

Denominator

Non-elective spells where the day of discharge was a weekday, linked into superspells.

Observed

Denominator superspells with a readmission to any provider within 28 days of discharge where readmission is an emergency admission:

ADMIMETH:

21. Emergency – via A&E.

22. Emergency – via GP.

23. Emergency – via Bed Bureau.

24. Emergency – via Outpatient clinic.

28. Emergency – via other means.

Readmission date minus discharge date < 28 days.

Expected

Expected number of 28-day emergency readmissions derived from logistic regression, adjusting for factors to indirectly standardise for differences in case mix.

Data sources

SUS–CDS for the financial year 2012/13.

Statistical methods

• Logistic regression: the expected number of readmissions is derived from logistic regression, adjusting for factors to indirectly standardise for differences in case mix. Adjustments are made for:

  • sex

  • age on admission (in 5-year bands up to 90+ years)

  • interactions between age on admission (in 5-year bands up to 90+ years) and Charlson Comorbidity Score

  • admission method (non-elective or elective)

  • socioeconomic deprivation quintile of the area of residence of the patient (based on the Carstairs Index)

  • diagnosis/procedure subgroup

  • comorbidities (based Charlson Comorbidity Score)

  • number of previous emergency admissions

  • year of discharge (financial year)

  • palliative care (if any episode in the spell has the treatment function code 315 or contains ICD-10 code Z515 in any of the diagnoses fields)

  • month of admission

  • source of admission.

• Relative risk: the ratio is calculated by dividing the actual number of readmissions by the expected number and multiplying the figure by 100. It is expressed as a relative risk, where a risk rating of 100 represents the national average. If the trust has a standardised readmission ratio of 100, that means that the number of patients who were readmitted is exactly as would be expected taking into account the standardisation factors. A standardised readmission ratio above 100 means more patients were readmitted than would be expected; one below 100 means that fewer than expected were readmitted.

• Control limits: control limits tell us the range of values that are consistent with random or chance variation. Data points falling within the control limits are consistent with random or chance variation and are said to display ‘common-cause variation’; for data points falling outside the control limits, chance is an unlikely explanation and hence they are said to display ‘special-cause variation’ – that is, where the trust’s rate diverges significantly from the national rate.

• Data points falling above the upper 99.8% Poisson control limit are said to be significantly ‘higher than expected’, data points falling below the lower 99.8% Poisson control limit are said to be significantly ‘lower than expected’, data points falling between the lower 99.8% Poisson control limit and the upper 99.8% Poisson control limit are said to be ‘within expected range’.

Meeting 1

Meeting 2

Meeting 3

Meeting 4

Meeting 1

Meeting 2

Meeting 3

Meeting 4

Meeting 1

Meeting 2

Meeting 3

Meeting 4

Meeting 1

Meeting 2

Meeting 3

Meeting 4