Governance of quality of care: a qualitative study of health service boards in Victoria, Australia.

OBJECTIVES: To describe the engagement of health service boards with quality-of-care issues and to identify factors that influence boards' activities in this area.
METHODS: We conducted semistructured interviews with 35 board members and executives from 13 public health services in Victoria, Australia. Interviews focused on the role currently played by boards in overseeing quality of care. We also elicited interviewees' perceptions of factors that have influenced their current approach to governance in this area. Thematic analysis was used to identify key themes from interview transcripts.
RESULTS: Virtually all interviewees believed boards had substantial opportunities to influence the quality of care delivered within the service, chiefly through setting priorities, monitoring progress, holding staff to account and shaping culture. Perceived barriers to leveraging this influence included insufficient resources, gaps in skills and experience among board members, inadequate information on performance and regulatory requirements that miss the mark. Interviewees converged on four enablers of more effective quality governance: stronger regional collaborations; more tailored board training on quality issues; smarter use of reporting and accreditation requirements; and better access to data that was reliable, longitudinal and allowed for benchmarking against peer organisations.
CONCLUSIONS: Although health service boards are eager to establish quality of care as a governance priority, several obstacles are blocking progress. The result is a gap between the rhetoric of quality governance and the reality of month-to-month activities at the board level. The imperative for effective board-level engagement in this area cannot be met until these barriers are addressed. Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://group.bmj.com/group/rights-licensing/permissions.

Introduction

Effective governance is increasingly recognised as pivotal to improvements in healthcare quality, including patient experiences and the safety and effectiveness of care.1–5 While much has been written about the theory of clinical governance,6 much less is understood about the real-world factors that stimulate and retard board activities in this area.

Early evidence7–11 suggests that hospitals overseen by boards that are actively engaged in the institution's quality agenda are more likely to have quality improvement programmes in place and to perform better on a variety of indicators, including risk-adjusted mortality rates. Conversely, poor board oversight has been identified as a contributing factor12 13 to many high-profile failures in care. In the UK, the Francis Inquiry into serious problems at Mid Staffordshire NHS Foundation Trust14 found that the failure of the hospital trust board to ‘get a grip’ on its governance responsibilities left the board blind to many concerns. In Australia, the Forster Review of Queensland's health services reported concerns regarding “the inability of boards to properly understand or influence the growing complexities of health service delivery requirements”.15

Most qualitative studies of quality improvement in healthcare have focused on the views and experience of management staff and clinicians,16–18 not boards. One US study that interviewed 26 hospital board members found that they felt undereducated about evidence-based medicine and error-reduction strategies, ill-equipped to make quality-of-care decisions and hesitant to challenge the hospital culture in the clinical domain.3

To assist boards to understand what they should be doing in this area, agencies in several countries have promulgated guidance documents.19–22 In the UK, for example, the National Quality Board developed the publication ‘Quality Governance in the NHS—A guide for provider boards’.17 In Australia, the Commission on Safety and Quality in Health Care has produced an ‘improvement guide’ to assist boards with the implementation of newly introduced national standards for safety and quality.21 However, movement towards ‘best practice’ is uneven. Recent research from Australia, the USA, the UK and Canada has shown substantial inter-board variation—both in the intensity of board engagement and in the attitudes of board members to quality issues.9 23–27

We conducted interviews with board members and senior executives from 13 public health services in Australia. Our aim was to identify the tools that boards use to oversee and improve quality of care and to better understand the key influences of board activity in this area.

Methods

Setting

Our study was based in Victoria, Australia's second most populous state. Victoria has a well-established system of local health service governance with 85 separate public health services, ranging in size from large metropolitan services with more than 500 acute-care beds to small rural services with fewer than five beds.28 Each health service is governed by a local board appointed by and accountable to the Minister of Health (see table 1 for further details).

Table 1

Public health service boards in Victoria, Australia

Characteristic	Public health service boards in Victoria
Population served	Victoria has 5.6 million residents, making it the second most populous state in Australia
Governing legislation	Health Services Act 1988
Number of boards	85 public health services are each governed by their own board: 16 in metropolitan areas 16 in regional and subregional areas 53 in rural areas
Functions	The statutory functions of the board include ensuring that effective and accountable systems are in place to monitor and improve the quality and effectiveness of health services provided any problems identified with the quality or effectiveness of the health services provided are addressed in a timely manner
Number of members	Each board has between 6 and 12 members. Both men and women must be adequately represented on the board
Independence	All board members are independent non-executive directors. No more than a quarter of members may be medical practitioners, and employees of the health service are not eligible to serve on the board
Term of appointment	Board members are appointed by the Minister of Health for a 3-year term with the possibility of re-appointment
Remuneration	Members of metropolitan and larger regional boards are remunerated. Members of subregional and rural boards serve as volunteers
Time commitment	Boards meet on average 11 times a year, with additional committee meetings
Committees	Each board is required to have a quality committee, as well as a risk committee and an audit committee. The chair of the quality committee is a member of the board.

Study sample

We aimed to interview board members and executive staff with a diversity of views on governance of quality of care. To this end, we sampled health services at which to conduct interviews using two strata: boards’ level of engagement in quality-related activities and the location of the health services.

In a recent survey of board members from all Victorian health services,24 we queried respondents about whether their boards were undertaking each of 15 specific quality-related activities (eg, board regularly reviews a quality dashboard). The activities were derived from a similar list developed by Jha and Epstein11 and a review of the international literature on clinical governance. Responses were used to classify boards into categories corresponding to high, medium and low levels of activity. These categories formed the ‘level of engagement’ strata. For the locational strata, we classified each health service as being located in a metropolitan, regional or remote area using the geographic classifications employed by the Victorian Department of Health.29 This helped to ensure that we included boards overseeing health services of varying sizes and serving different types of communities.

Combining the two strata created nine ‘cells’ into which we sorted all 82 health services that responded to our survey (table 2). Within each cell we selected one or two health services using a purposive sampling approach.30 31 Of the 14 health services selected and invited to participate, 13 agreed and participated.

Table 2

Stratified sampling of health services

Quality activity score	Metropolitan health services (n=15)	Regional and subregional health services (n=16)	Rural health services (n=51)
High (n=19)	2	2	2
Medium (n=38)	1	2	2
Low (n=25)	1	1	1

Selection and recruitment of interviewees

We aimed to interview at least two people from each health service: a senior member of the board and a senior member of the executive. We invited the chair of the board first, and if she/he was unavailable, reverted to the chair of the quality committee (who is also a board member). For executive staff, we invited the chief executive first, and if she/he was unavailable, reverted to a member of the executive staff with responsibility for quality.

For 12 of the sampled health services, an interview was completed with either the chair of the board (n=7) or the chair of the quality committee (n=5); for 11 of the sampled health services, an interview was completed with either the chief executive (n=7) and/or an executive with responsibility for quality and safety (n=7). On several occasions, when interviewees presented with colleagues, the interview proceeded in group format. When the targeted interviewees recommended follow-on discussions with other leaders in their health service, we conducted those additional interviews wherever feasible.

Content of and conduct of interviews

The interviews followed a semistructured format using an interview schedule informed by our review of the literature. (The interview schedule is available in appendix A—web only). The interviewer explained that the study goal was to understand the board's attitudes and activities in relation to the quality of care provided by their health service. The questions that followed addressed changes in clinical governance over time, the mix of skills among board members, the use of dashboards and benchmarks, opportunities for training, perceived barriers to governance of quality of care and suggestions for improvement.

One of us (MMB) conducted all of the interviews between September 2012 and January 2013. They lasted 30–60 min. One was conducted by telephone; the rest were conducted face-to-face at the health service, the board member's workplace or the researchers’ offices, according to the interviewee's preference. Interviews were digitally recorded and transcribed verbatim.

To maximise candid discussion, interviewees were assured confidentiality. The University of Melbourne's Human Research Ethics Committee approved the study.

Coding and analysis

Data were managed using NVivo V.9 software. We conducted a thematic analysis32 33 aimed at identifying a set of main themes in the views expressed. Using transcripts from the first five interviews, both investigators read and discussed the content and identified the main themes, which formed the basis of a draft coding framework.

One investigator (MMB) then reviewed transcripts from the rest of the interviews, applying the draft coding framework and making modifications to it through an inductive and iterative process. The other investigator (DMS) independently repeated this process for a subset of the transcripts. The two investigators then discussed the coding framework and coding choices in detail. Differences were resolved by consensus. All coding was reviewed in light of these inter-reviewer discussions and decisions about the final framework. Throughout this process we met regularly with our research advisory group to discuss the evolving framework being used to interpret the data.

Results

The 35 interviewees consisted of 7 board chairs, 5 chairs of quality committees, 5 other board members, 7 chief executives and 11 other executive staff (chiefly, medical directors, risk managers and quality managers). Board members had a range of professional backgrounds, including law, accounting, farming, politics, teaching, nursing and medicine.

Findings are organised below according to themes that emerged during the interviews. Quotes are used to illustrate and support points, with the position and health service location of quoted interviewees described in brackets after each quote. Where necessary, some minor details have been changed to preserve the anonymity of interviewees.

Changes over time

Interviewees described a dramatic change over the last 10 years in the extent to which boards were engaging with quality-related issues. One chair from a regional health service described previous boards as consisting of “good people around the town that were all well meaning but didn't have a great understanding of governance issues or of health services as such” (chair, regional). He followed this characterisation with a description of how board members' skills and focus had changed over recent years.

A recurring theme in interviews was that boards were becoming more professional and more engaged with quality issues. This shift was attributed to the international “rise of the clinical governance movement” (chair, rural) and an increased understanding that quality of care was core business and not just a “compliance requirement” (chief executive, rural).

Interviewees commonly spoke of governance of quality of care as an “ongoing journey” (chair, metropolitan). The extent to which interviewees felt their organisations had made progress on this journey varied considerably. One quality manager described her board's engagement in quality-related issues as follows:

It's not there yet. But it’s certainly been an awakening. Yeah, there’s an awareness and an awakening. (Quality manager, regional)

In contrast, the chair of one highly engaged board commented:

Four years ago there was no quality and safety reporting to the board. [It was] “secret doctors’ business”. Now quality is a big item on the agenda, being re-assured that we've got [the right reporting], processes and systems. (Chair, metropolitan)

Board ‘tools’ for governing quality of care

There was wide agreement among interviewees with the idea that “a board absolutely has the levers to influence the quality of care” (quality committee chair, metropolitan). In response to a question querying which board activities had the potential to make the greatest difference, four themes emerged: setting priorities, measuring progress, ensuring accountability and shaping culture. Table 3 sets these themes, or ‘governance tools’, alongside specific tasks mentioned by interviewees and illustrative quotes.

Table 3

Perceived board tools in improving quality of care

Tools	Tasks	Quotes from interviews
Set priorities	Develop and drive strategy	“Within the panoply of things you have to attend to, you're going to have a focus on a particular subset … [We] are driving a strategy; not just getting through the agenda.” (Chair, metropolitan)
	Allocate resources	“Our obstetrics area reported that perinatal deaths were on the high side. It was a high risk … so we've got more foetal monitors being purchased. We're not waiting for government to fund something.” (Chair, regional)
	Look to the future	“We are continually looking at where we can improve and where the future might be … The acuity level is increasing, so then it's about what equipment do we have … Making sure that we're ahead of the game.” (Deputy chair, regional)
Measure progress	Monitor performance	“I think outcomes, at the end of the day, are the yardstick by which you measure your governance progress. We have a good system [of quality indicators] in place to check and measure.” (Board member, rural)
	Establish targets	“We've got a strategy with clear targets for trying to push [adverse events] down. It's worked for some things. It hasn't worked for others. But at least we've got something to aim for.” (Chair, metropolitan)
	Identify and mitigate risks	“Our board reporting is tied to strategic risk, and we embed strategic risk into our Board reporting. So the board can see where things are tracking, they can see the medication errors, they can see the falls.” (Risk manager, rural)
Ensure accountability	Meet front-line employees	“On a regular basis the board has a meet and greet. We went to theatre and met with the infection control group. We went to the emergency department … we've met with cafeteria services.” (Chair, regional)
	Hold staff to account	“We have in place very clear procedures to make sure that we cover quality and safety, and distinct plans we follow of who's responsible, and timelines of when it needs to be done.” (Quality committee chair, rural)
	Engage with consumers	“[There] should be a relatively short piece of string between the decisions we're making and the effect on the patient … Community representatives have direct input in and get feedback out.” (Quality committee chair, regional)
Shape culture	Recruit good leaders	“I know the energy that we put in at the board level. Just making sure that we've got the right people into these positions.” (Quality committee chair, rural)
	Support ‘just’ culture	“To get openness of reporting and responsiveness of our clinicians we need to provide a ‘just culture’ where they don't fear there are going to be ramifications simply because there have been errors.” (Risk manager, rural)
	Foster innovation	“Our doctors and nurses know that they can have a lot of control. If something can be done better, and the junior staff say ‘Look, I've seen this done better somewhere else’, we'll look at it.” (Medical director, rural)

Setting priorities

Two-thirds of the interviewees remarked on the role of the board in establishing strategic priorities and “deciding what direction we should follow” (chief executive, rural). One quality committee chair spoke of the board's strategic focus on quality of care as underpinning all other governance decisions:

Everything we do in this organisation, it links back—it has to—to clinical governance because that's our core business. It's what we are all about—giving the highest quality so we get the very best outcome for each of our patients and their families. (Quality committee chair, rural)

Measuring progress

Interviewees from 8 of the 13 health services cited effective measurement and monitoring as a critical facilitator of effective governance. One quality chair emphasised the importance of being able to “step back a bit and look at what the high indicators are” (quality committee chair, regional). In order to be able to do this in a meaningful way, board members sought measures that were robust, relevant and tracked over time, with a focus on understanding progress rather than merely “filling filing cabinets” (medical director, rural).

Ensuring accountability

Interviewees emphasised the role of the board in holding staff to account for the delivery of safe, appropriate and effective care. Approaches to achieving this included ensuring that “everyone is made to own their own area” (board member, rural), establishing clear milestones and timelines, and, where necessary, addressing poor performance rather than “sitting back and saying, well, they should know what they're doing” (chair, metropolitan). One chair spoke of the importance of finding a balance between acknowledging improvements in the quality of care and aiming even higher:

The board just kept pressing and pressing and saying, “Well, we're not there yet”. You've done this [but] we've still got a long way to go. (Chair, metropolitan)

Shaping culture

Interviewees frequently spoke of the role boards played in shaping the culture of the organisation. Most often, the description related to issues of openness and transparency. For example, one interviewee spoke of her board's efforts to assure staff “that there’s not a blame culture; that reporting is a good thing and it’s about improvement” (executive manager, regional). Another manager commented on the ability of a strong board to “set the tone of the expectations” throughout an organisation (medical director, metropolitan). In six organisations, interviewees attributed significant changes in organisational culture to an individual who had championed the importance of quality of care in the boardroom.

In sum, there was a high level of agreement among interviewees—both about the governance tools that boards had at their disposal and the specific tasks that could bring about change. However, there was substantial inter-board variation in how far they had progressed in exploiting those opportunities. Some boards were highly engaged in clinical governance across the full spectrum of activities described in table 3 and were viewed as leaders in the field. Others struggled to gain traction, with one board member describing his board's clinical governance activities as “lackluster” and “needing a good shake-up” (board member, rural).

Key influences on board activities

Follow-on interview questions probed what board members and senior executives perceived as the most important influences on progress in governing the quality of care. In particular, interviewees were asked to explain the factors that had influenced their board's approach to quality, the things they would do differently in an ideal world and the barriers they had encountered in seeking to undertake quality-related initiatives. Interviewees’ responses converged around four themes: organisational resources, board members’ skills and expertise, information on performance and the external regulatory environment as key influences on board activities.

Resources

A commonly reported barrier to stronger oversight of quality issues was “lack of resources and time” (chair, regional). Board members from eight health services mentioned financial constraints as a perceived barrier to undertaking certain quality initiatives: some said they were “just not set up” (quality committee chair, rural) with the necessary infrastructure to undertake a full suite of quality activities; for others it was difficult to even find time to discuss quality of care at board meetings. One chair noted regretfully: “the finance consumes a lot of time because we run in deficit” (chair, regional).

Collaboration with other health services was identified as a powerful way of pooling resources to address quality-related issues of common interest. Over half of the interviewees identified a need for stronger networks with other boards and greater sharing of governance policies, templates and expertise. However, there was concern to ensure that such collaborations did not come at the cost of effective representation of local community concerns at board level.

Skills and expertise

Gaps in the skills and expertise of board members emerged as a second barrier to effective governance of quality issues. Chairs at four health services spoke of the difficulty of identifying strong candidates who did not have a conflict of interest and were willing and able to serve on a board, particularly in rural and regional areas where board members are not remunerated for the role. One chair commented:

It's a huge ask for someone that's employed full-time. So there's an imbalance of [too many] retired people on the board. (Chair, regional)

Within the metropolitan health services, two chairs expressed frustration at the inability of the board to decide who should be appointed:

That skill matrix issue which is so dominant in good boards is not present in the health sector. Because the Minister decides. Did I recently get a communication expert to put in an application? Yes, I did. Was he selected? No. What did I get? A third accountant. So how can I as Chair be held to account in the same way as I would if I was actively involved in ensuring that my board had the right matrix of skills? (Chair)

Among executives, the most common skill-related concern was that the inexperience of some board members gave rise to an undue focus on operational issues. One quality manager commented bluntly that “new board members need to understand what ‘Noses in and fingers out’ means” (quality manager, regional), meaning that board members need a sound understanding of organisational risks and issues but should avoid micromanaging operational activities.34

Interviewees from all 13 health services identified gaps in the governance training currently available and expressed an appetite for new and better offerings. Interviewees commented on the need for training on quality-of-care issues to be accessible to all board members and tailored to avoid some participants feeling overwhelmed while others were “being taught to suck eggs” (chair, metropolitan). (This view resonated with the more general perception that boards were at quite different stages in their knowledge of quality-related issues.) Specifically, board members identified a need for (1) a basic introduction to quality, safety and risk for new board members; (2) topic-specific training on issues such as patient-centred care; and (3) advanced ‘master classes’ dealing with cutting-edge innovations in healthcare quality governance.

Information on performance

A recurring concern expressed by board members was receipt of information that was of the wrong type or at the wrong level. The problem manifested in two main ways. First, board members from five boards felt inundated with volumes of paperwork that did not necessarily equip them to drive improvements in the quality of care. In the words of one quality committee chair: “We were getting so much information we couldn't actually distil it” (chair, rural).

Second, certain types of information, in which board members had a clear interest, were difficult to obtain. This concern arose most strongly in relation to external quality measures against which the health services sought to benchmark performance. Most interviewees were aware of “pockets of data” (quality manager, rural) and some were members of The Health Roundtable, a non-profit collaborative organisation that collects, analyses and shares comparative data between member organisations.35 However, the absence of a consistent and reliable set of statewide quality indicators was keenly felt and repeatedly noted.

Board members and executives expressed a willingness to share certain quality-related data with other organisations, but interviewees from five health services added the caveat that, for such information sharing to be worthwhile, the coordinating body needed to ‘close the loop’ by providing meaningful reports back to the contributing health services.

The measures sought by interviewees fell into predictable categories, relating to access (eg, waiting times), efficiency (eg, duplication of services), effectiveness (eg, patient outcome measures) and safety (eg, medication errors). Table 4 outlines specific barriers interviewees identified as blocking access at the board level to data suitable for monitoring and benchmarking performance.

Table 4

Perceived barriers to receipt of quality-of-care data for monitoring and benchmarking by health service boards

Perceived barriers	Quotes from interviews
Acceptability	“It's tough to find indicators that the medical staff will accept as meaningful.” (Medical director, rural)
Accuracy	“There are programs which can be easily manipulated … I think it's that old thing: rubbish in, you get rubbish out. So it's really reliant—still—back at the coal face, on reporting.” (Chief executive, rural)
Affordability	“There's quite a bit of criticism on how much money can be spent [on quality reports] and is it necessary.” (Chief executive, rural)
Comparability	“You need to ensure that apples are compared to apples because that's one of the biggest issues that we found when benchmarking projects [were] undertaken, that it's not necessarily always comparable.” (Quality committee chair, rural)
Completeness	“It comes back to those gaps in data and benchmarking … they are pretty well defined and available in the acute area, but I personally find aged care is a real vacuum. And in primary care, it's also hit and miss in terms of what data is around.” (Executive manager, rural)
Pertinence	“Major investigations in the health sector still come about through whistleblowers, not data.” (Chair, metropolitan)
Simplicity	“We actually had developed our own reporting system. Well, yeah, [the Department of Health] came in on top of that, and added what they called a minimum data set that had thousands of classifications and—you know—made our reporting system much more difficult.” (Risk manager, regional)
Sustainability	“The patient safety indicator programme looking for outliers in key areas—like complications post-surgery—was looking really good. But it's just disappeared. Gone, I'm sure.” (Executive manager, regional)
Timeliness	“We just keep hounding the Department of Health ‘til we get [benchmarking data] and it might take us six months to get the figures.” (Medical director, rural)
Validity	“Measuring outcomes is technically very difficult. It has to be unbiased, it has to be systematic, and it has to be risk-adjusted.” (Quality committee chair, metropolitan)

Regulation

Interviewees from every health service cited external regulatory requirements as a barrier to more effective clinical governance within their health service. This did not always manifest as a desire for less government intervention: interviewees cited examples of requirements that posed an unnecessary burden, but also of ‘voids’ that it would be helpful for a central agency to fill.

A recurring theme was the governance burden associated with maintaining accreditation and meeting new national standards.21 Current reporting requirements were described by one board member as “aggregated on, like mollusks on the hull of a ship over time” (chair, metropolitan). A chief executive explained:

Within this organisation we have nine accreditation systems … Now for a board to manage that from a governance perspective is nigh on impossible, yet each year with iteration from the various funding bodies, there will be attached to that a volume of additional accountabilities that go with it for no benefit to the board and no benefit to the organisation. (Chief executive, rural)

Notwithstanding clear frustration with certain reporting and accreditation requirements, most interviewees were able to envisage a productive role for external agencies—including the state Department of Health, professional liability insurers and industry bodies—in supporting improved quality governance. Interviewees pointed to three specific reforms that would go some distance towards achieving this. First, board members saw a need for more templates and tools that boards could “adapt to their own organisation and their own activities” (Chair, metropolitan).

Second, board members and chief executive officers consistently mentioned the need for rationalisation of the reporting and accreditation requirements imposed on boards, eliminating overlap and limiting mandatory reports to information with the capacity to improve performance and patient outcomes.

Third, members of boards that were already relatively sophisticated in their approach to clinical governance expressed a desire to have greater freedom to pursue their own quality improvement agendas and try innovative initiatives. They felt too tightly constrained by government-determined priorities and ring-fenced funding allocations. As one board chair put it:

When boards look prepared, when boards understand their communities, when boards can provide the proper business plans, when boards can demonstrate the return to government then there needs to be an allowance to let them go forward that way. (Quality committee chair, rural)

Discussion

Over the last 10 years two forces—increased scrutiny of institutional leadership and enhanced pressure to improve healthcare quality—have coalesced to transform the roles and responsibilities of hospital boards.2 22 Prominent government investigations, such as the Francis Inquiry in the UK, have left little doubt that ultimate accountability for ensuring a safe standard of care rests with the highest levels of governance.14 Boards do not deliver front-line services or manage operational details, but they can and should establish the leadership, accountability and organisational culture necessary for staff to deliver safe and effective services.2 10 14

An emerging body of research suggests considerable variation, within health systems, in the extent to which evolving expectations about governance of quality have penetrated board practices.11 23 36 Our study provides further evidence of such intrasystem variation. While some health service boards in Victoria had high aspirations and clear plans to improve care quality, others appeared to be ‘muddling through’. This resonated with findings from our earlier survey of board members from all 85 health service boards in Victoria,24 where wide variation was evident in the nature and extent of relevant activities being undertaken at the board level.

What accounts for the uneven pace of progress among boards towards substantive engagement in quality governance? One possibility is a lack of awareness or enthusiasm, with some boards not yet conceiving of such engagement as part of their brief. Our findings point away from this explanation. The vast majority of interviewees perceived boards as having both a responsibility and an array of opportunities to oversee and support improvements in the care delivered by their health service. Explanations for why some had not exploited those opportunities centred on four main barriers, articulated with remarkable consistency across different types of organisations and interviewees: insufficient resources, lack of skills and expertise among board members, inadequate information on performance and unhelpful forms of regulatory control.

With respect to resources, the pressures of operating in a publicly funded system with fixed budgets were keenly felt by most board members and executives interviewed. Although quality and safety were rarely viewed as a ‘luxury’ item for board agendas, some board members argued that the acuity of fiscal challenges left little space for other priorities. The uncertain returns on potential investments in quality improvement were also noted. Better understanding of the cost-effectiveness of different initiatives would be valuable.

With respect to skills and expertise, our findings echo messages from previous studies in the USA,11 the UK23 and Australia,24 indicating that (additional) board training on quality governance would be useful. Such training should be accessible, flexible and tailored rather than assuming that ‘one size fits all’. Further strategies for addressing skills gaps may include remunerating all board members so that they can devote a meaningful portion of their professional work to the role23 and ensuring that the process for appointing new members is merit-based and achieves a balanced matrix of skills.

The main problem with board reports appeared to be a ‘filter deficit’, whereby some board members simultaneously felt overwhelmed with data yet not satisfied that they were getting the right information. This problem was frequently located in quality metrics. Board members sought timely, accurate and pertinent measures of quality that could meaningfully be compared over time and across services. Yet, currently available quality metrics are frustratingly piecemeal and incomplete, with the benefits of some data-sharing collaborations (eg, the Health Roundtable35) restricted to member organisations.

These findings lend weight to the call for high-quality performance indicators amendable to benchmarking across peer health services.37 38 Performance indicators have well-described limitations,39–41 including the risk of inappropriate indicators, invalid inferences and perverse incentives. But without meaningful metrics, boards will remain constrained in their ability to recognise poor performance or, perhaps more importantly, to identify innovations that merit wider dissemination.42 The establishment in 2012 of a National Health Performance Authority in Australia may be a welcome development in this regard.43

Finally, our findings highlight the need for better coordination between local and national systems of governance. Boards felt both burdened and hamstrung by external reporting requirements and other forms of regulation in the quality domain. Accreditation carries a real cost for uncertain benefit,44 and duplication and inconsistency in reporting requirements divert time, energy and resources away from other initiatives. Wherever possible, the activities undertaken by health service boards should complement, and be integrated with, other activities occurring within and outside the organisation.

Our study has strengths and limitations. Board members are a difficult population to reach for research purposes, both because they tend to be busy people and because of the confidential nature of boardroom discussions. Three strengths of this study were the high rate of participation (only one health service declined to take part), the diversity of leaders involved and the apparent candour with which they shared their views. Although we could not test validity of responses, we observed substantial internal consistency in views and themes across interviewees and health services. It should be noted, however, that there are socially desirable responses to many of the questions we posed and we could not test the veracity of responses; this may have introduced some biases. Finally, the generalisability of our findings outside Victoria is unknown.

It is increasingly apparent that the quality ‘buck’ stops with boards.14 45 They have a duty to ensure that effective systems are in place to ensure the quality of care; they must also address problems quickly. This study contributes to a growing international literature documenting challenges boards face on the ground as they seek to respond to changing expectations in governance of quality. It is a shift from traditional forms of healthcare governance that will likely take many years to fully penetrate healthcare systems. The most helpful accelerants may be the development, implementation and evaluation of strategies to address common barriers to progress and the promulgation of successful board initiatives.