The challenges of safety improvement in New Zealand public hospitals

ABSTRACT

Patient Safety is central to the New Zealand (NZ) health strategy. However international experience shows that safety improvement work frequently fails to meet its objectives. This article provides a qualitative account of the challenges of safety improvement from the perspective of nurses, doctors, and managers in three departments in two NZ public hospitals. These staff described significant tensions in relation to engagement and organisational capacity. An analysis of their perspectives through the Organising for Quality model [Bate et al. 2008, Organising for quality: The improvement journeys of leading hospitals in Europe and the United States, Oxford & New York, Radcliffe Publishing] highlights the significance of failures at the level of structure, culture, politics, motivation and infrastructure.

KEYWORDS:

• Patient safety
• safety improvement
• hospitals
• New Zealand

About one in ten patients treated in advanced healthcare are harmed by an adverse event (de Vries et al. 2008). In New Zealand (NZ), where additional treatment due to preventable harm has been measured as accounting for 20% of all public hospital expenditure (Brown et al. 2002), safety has been a longstanding strategic priority for the District Health Boards (DHBs) (Minister of Health 2000, 2015).

The urgent need to improve patient safety is vexed by the complexity of the problem (Braithwaite et al. 2009). Internationally, efforts have mostly delivered ‘inconsistent and patchy results’ (Dixon-Woods et al. 2012, p. 876). Work in the NZ context may be thought about in two phases. In the first, following the creation of the DHB system in 2000, cultural and structural factors contradicted safety improvement. Relationships between managers and clinicians, who were expected to work cooperatively to prioritise safety, deteriorated (Ministerial Task Group on Clinical Leadership 2009). There was no organisation with dedicated expertise, quality was not at the centre of all policy and service work, quality focussed services were not built, there was no national infrastructure to support it, and the DHBs were not required to provide evidence of the implementation of best practice (Gauld 2009). The second and current phase followed the 2010 emergence of the Health Quality & Safety Commission, which supports a suite of patient safety programmes. Gauld and Horsburgh (2012, 2015) however maintain that while the DHBs are increasingly prioritising quality, safety and clinical leadership, senior doctors are insufficiently supported to lead safety improvement, and they tend to not prioritise it. (Gauld and Horsburgh 2012, 2015). In addition, Beaver (2017) argues that risk control work in NZ is being compromised by intense throughput pressure on nursing and medical staff at all levels.

In this article I provide an account of the challenges of safety improvement from the perspectives of staff in three departments in two NZ public hospitals. There are four parts to the analysis. First, I discuss safe practices, implementation, and socio-organisational theory about safety improvement. Second, the research methodology is described. Third, the findings of 37 qualitative interviews with staff are elaborated. Fourth, I comment on staff experiences through the lens of socio-organisational theory.

Safety improvement: practices and implementation

Practices

Safety improvement practices draw upon evidence-based medicine, and human factors and ergonomics. Evidence-based medicine, which seeks to align clinical practice with the best scientific evidence of safety and effectiveness, requires practitioners to either critically evaluate evidence, or apply clinical practice guidelines to diagnosis and treatment (Daly 2005). Human factors and ergonomics is concerned with optimising human performance in systemic contexts. Its ultimate purpose is high reliability, where errors and harm are minimised despite constant risk (Kohn et al. 2000). This outcome may be achieved through risk management, redesigning technologies and systems, and supporting a culture of safety (Bellandi et al. 2011).

In an extensive review of the evidence for safety practices the Agency for Healthcare Research and Quality has rated ten strategies ‘strongly encouraged’, and twelve ‘encouraged’. These strategies address both common specific risks, such as those involving surgery, medications, infections, and falls; and practices targeting generalised risk, such as team-training and detecting adverse events (Shekelle et al. 2013, p. ES12–13). But while ‘more than enough evidence exists to prompt decisive action’ (Wachter et al. 2013, p. 350), safety requires the routinisation of these practices.

The need to routinise new practices makes implementation critical to safety. But implementation is difficult. The adoption of evidence-based medicine has been slow and uneven for reasons including an overwhelming volume of evidence, contradictory guidelines, and poor fit with the needs of multimorbid patients (Greenhalgh et al. 2014). Human factors practices may be similarly compromised. Many incident reporting and learning systems deliver substantial underreporting (Sari et al. 2007). Redesign tasks can exceed the capacity of local actors, and changes may deliver new unintended risks (Dixon-Woods and Pronovost 2016). Teamwork competency has not been systematically developed in healthcare (Salas and Rosen 2013). Checklists, which can aid task completion and prompt communication, may be performed as tick-the-box exercises, or over-prescribed as simple solutions to complex problems (Dixon-Woods et al. 2014).

Implementation

Implementation often draws upon methodologies such as quality assurance, audit, and quality improvement. Quality assurance requires measuring selected structures, processes, and outcomes, and maintaining them within acceptable standards (Donabedian 2005). Audit is a traditional system of reviewing and improving individual and collective practice (Donabedian 2003). At the local level it may involve a cycle of planning, measuring, implementing, and sustaining change (Burgess 2011). Quality improvement, which targets collective effort, is a set of analytic and organising methods for changing processes in real time (Colton 2000). Methods include continuous quality improvement, lean engineering, and plan-do-study-act cycles.

All of the implementation methods present challenges. While quality assurance can signal priorities and shape resourcing and action (Meyer et al. 2012), its weaknesses include antagonising clinicians with unrealistic measures (Mountford and Shojania 2012), the resource costs of excessive measurement (Meyer et al. 2012), focussing on what is measured while things not measured are ignored (effort substitution); and data manipulation to create the appearance of compliance (gaming) (Mannion and Braithwaite 2012). Challenges for audit and quality improvement include planning and design; working with organisational and institutional contexts, professions and leadership; and sustaining change (Dixon-Woods et al. 2012). In theory, the multiple socio-organisational challenges of implementing safety improvement can be explained as driven by the complex adaptive characteristics of healthcare systems. Hospital staff are active meaning-giving agents, and as such they adapt and respond to change in ways that may be unexpected, and contradictory to the intentions of managers and policy makers. Fortunately, with understanding, social action in complex adaptive systems can be influenced to deliver intended outcomes by carefully shaping the context of action. (Plsek and Greenhalgh 2001). A theoretical overview of what this involves in healthcare is provided by the organising for quality (OQ) model, developed by Bate et al. (2008) from a study of hospitals that have excelled in quality improvement (safety is a subset of quality). As a grand theory of change (Davidoff et al. 2015), the OQ model offers an outline, rather than a detailed perspective on how to do improvement. It identifies six universal improvement challenges that must be overcome:

1. Structure: planning and coordinating improvement

2. Culture: building a shared understanding of improvement

3. Politics: negotiating change and achieving buy-in and engagement

4. Motivation: energising and inspiring staff

5. Learning: embedding continuous learning

6. Infrastructure: technological systems

Fifty-six solutions (more may be possible) to these challenges have been identified. They become effective through their articulation with one another in a mutually-reinforcing and dynamic process of change. In any given hospital the characteristics of the solutions and their combinations are unique according to distinct differences of history, culture, and size that shape organisational planning and action, individual initiative, and local opportunities. The value of the OQ model for the research reported in this article is that it provides a theoretical context for interpreting the improvement challenges described by staff, and a guide for where efforts should be directed. The Discussion that concludes this article refers back to the model, details missing solutions that compromised safety in the departments investigated, and suggests how these may be worked with.

Methodology

The research data was gathered from naturalistic qualitative interviews that sought to understand participant’s views and experiences from their perspective in rich detail (Gubrium and Holstein 1997). Thirty-seven interviews were conducted during the twelve months to September 2013 with nurses, doctors, and managers in one medical and two surgical departments in two public hospitals (Table 1). Recruitment was conducted through posters, presentations, and word of mouth between staff. The interview questions (Appendix A) covered participants’ experiences of risk control (reported in Beaver 2017) and safety improvement (reported here). Because of the complexity and breadth of the topic it was not possible to cover all aspects of risk and improvement in any one interview. The mean interview length was 47 min. All interviews were transcribed and thematically analysed using QSR NVivo 10 software.

Table 1. Sample of interviews.

	Medical	Surgical	Quality	Participants n = 37
Nurses	4	11		15
Ward nurses	2	8
Charge nurses	2	3
Doctors	6	11		17
Junior doctors	1	6
Consultants	4	4
Clinical directors	1	1
Managers				5
Service managers	1	2
Quality managers			2

This research was ethically approved by the Health and Disability Ethics Committee (NTX/12/EXP/107) and the research office in both hospitals.

Findings

Participants discussed performance targets, incident reporting, improvement programmes, evidence-based medicine, audit, quality improvement, and implementation.

Performance targets: nurses

The six hour emergency department (ED) rule, which specifies that 95% of ED patients should be admitted, transferred, or discharged within six hours, was a point of contention amongst nurses. Charge nurses were divided about the rule. Ward nurses mostly opposed it because of logistical issues:

The ED are always calling up saying ‘I need to hand over a patient we’re going to breach’ and it’s like ‘we haven’t got a bed or the patient’s still in the room and we’ve got no space to put them.’ (Junior nurse)

Charge nurses offered three reasons in support of the rule. First, it valued patients, who disliked waiting. Second, research showed that patients with shorter waits had better outcomes. Third, targets motivated nurses, gave them ‘something to aspire to’, and prevented them being ‘too lax’. Against these benefits other charge nurses identified three perverse consequences. First, targets seemed to have created a ‘blame and shame’ culture. This was evident when hospital management, who appeared unsympathetic to nurses being ‘run ragged’ by short-staffing, identified underperforming wards in the hospital newsletter. These actions seemed to reflect an ‘underlying suspicion that staff sometimes play games’ to avoid taking patients. In one hospital a ‘silo mentality’ was said to have developed in the ED, with ‘strong figureheads saying we must play the game’. Second, patients could be admitted without adequate assessment, and moved into unsafe locations such as lounges. Third, targets drove the neglect of some tasks:

There’s a lot of focus on meeting the target, and that’s where a lot of energy goes. But the rest of the stuff that still has to be done sort of falls away. (Charge nurse)

The pressure to let some tasks slide was difficult to ignore when one ED was sometimes at 130% capacity, and across the hospital ‘everyone is putting their hands up saying we are at maximum capacity’.

Performance targets: doctors

Doctors perspectives on targets ranged from supportive, to ambivalent, to strongly opposed. One doctor described targets as a simplistic solution to a complex problem:

There’s enormous potential for us to improve things and do better but then you have to buy into adequate measurement and changing processes rather than shouting at people and waving the big stick and telling them to meet the targets. (Consultant)

A number of surgeons were ambivalent about surgical targets. Some factors were outside their control, and the demand was relentlessness:

The harder you work and the more you achieve the more is expected of you. So what are you going to do? (Consultant)

Our waiting lists at the moment have gone from six months, to five months, to four months, which we’re achieving in creative ways, but then one of the creative ways will get kai-bosched, but you still have to stick to your target, and that’s la-la-land. (Consultant)

Doctors who supported the ED target emphasised that systemic improvements had been made, supportive resourcing was provided, and the primary challenge was simply to get doctors in front of patients sooner. Ambivalent or critical doctors raised a number of concerns. They disputed the research supporting targets and pointed to studies showing inferior outcomes. They felt there was insufficient resourcing, and more pressure on nurses:

They get shoved into the ward, and I discharge them the next day. It’s little things like that which concatenate and stress the system. (Consultant)

Performance targets: managers

The service managers all noted that targets benefited patients by reduced waiting. One was also enthusiastic about the effect on clinicians:

We’ve got to do much more with much less and it’s very exciting. You’ve got to maximise every minute of your list but you can’t overrun because that costs money and you can’t under book because you’re wasting time. So it’s this constant tension, but it’s good because it’s a challenge and it keeps you on your toes. (Service manager)

However the service mangers all also indicated that targets had consequences. In particular there was discontent about the lack of correspondence between workload and target, which made staff ‘numb’ to measures that did not reflect effort. Staff could work hard on clinical priorities, but fail the target. Morale suffered, there was ‘whining and carrying on’, and a culture of blame was emergent.

Incident reporting: nurses

Nurses felt there was an inherent promise that incident reporting would improve ward organisation and resourcing. Nonetheless, they tended to under-report:

Not all the nurses are filing those reports because you can’t be stuffed. (Junior nurse)

I should be reporting every time I feel unsafe or like I am doing something that shouldn’t be done this way, but I don’t report it, and I know that there’s a lot of people that don’t report on our ward. (Junior nurse)

A number of nurses felt that non-reporting either contributed to the normalisation of unsafe practices, or was symptomatic of those risks being already normalised. They did not complete reports for three reasons. First, they were busy and did not want to report in their time. Second, they felt that reports were read, ignored, and nothing changed:

So we feel what’s the point of doing a half hour Risk-Pro when we could be doing six a day, and we are not getting any results back? What are they doing to improve the hospital and patient care? (Junior nurse)

Short-staffing was particularly contentious. Some nurses had reported it on numerous occasions before giving up because nothing changed. The third reason nurses did not report was because of the potential for blame and tension with colleagues and managers.

Charge nurses views of reporting varied. While they acknowledged underreporting, the process helped them to understand events on their ward, and, although it was difficult because of issues with time, efforts were made to use reporting to support learning.

Incident reporting: doctors

Doctors typically knew little about the reporting system, and they mostly avoided it. They preferred more trusted channels, such as the morbidity and mortality conferences, which eschewed the blame associated with the hospital system of ‘dobbing someone in’.

There was some regret among consultants about the lack of reporting. The data it could provide was unavailable to guide improvement. There were also concerns about the misuse of reports:

A lot of stuff is pushed under the carpet. People think ‘oh it’s another Risk-Pro problem so we’ll just talk to the junior staff and educate them.’ They pay lip service to stuff and it’s not followed through. (Consultant)

The information seems to disappear into a hole. Other people review what you have written and reassess and almost always downgrade the significance because there is presumably an organisational push to reduce the number of serious events. I’ve been on reviews of significant events, and the way the organisation responds to incidents is inadequate. There is a tendency to minimise the importance and minimise changes and minimise publication. So the Health and Disability Commissioner publishes investigations and recommendations. Now do we see in this hospital any publication about the major sources of risk and what are we doing about them? If there is such a document I don’t know where it is. Almost certainly it will be unintelligible to somebody like me who isn’t in on the act. (Consultant)

Quality improvement: nurses

Nurses were comprehensively involved in improvement. Activities included training, admission and discharge planning, care plans with risk assessments, reorganising and tidying wards, checking equipment, using checklists, and new clinical practices. Although some activities generated paperwork, efforts were also made to simplify written tasks.

Charge nurses found plenty of evidence from observations and audits that improvement was working. But they virtually all noted that improvement generated challenges of resourcing, culture, and motivation, which created five barriers to improvement. The first barrier was that because of resource limitations projects that proceeded in modules were only partly implemented. The second barrier was the difficulty of making any aspect of a programme work with insufficient or no dedicated resources. Management were said to support improvement ‘to a degree’ only, and while projects with allocated time made ‘huge progress’, implementation was impossible on a full patient load:

It was expected to be resourced completely internally so get your nurses to do it, we won’t give you any dedicated time. Initially there was, but then money became tight and resources dried up. It became an expectation that you would do it. (Charge nurse)

The insufficient resourcing of improvement set it into conflict with other priorities. Some improvement activities were perceived as too bureaucratic:

It’s too much about documenting and auditing and keeping a paper trail and proving that you’ve done it, rather than actually allowing and freeing up time to do stuff. That can be a frustration. I think if people want auditing and monitoring and it’s going to be useful then great. But if it’s just to say ‘yeah we’ve done a good job, pat ourselves on the back,’ it’s not a good use. Our resources are too precious to waste. (Charge nurse)

In other cases activities directly targeting patient care, such as hourly rounding, were similarly rejected as unrealistic because of time constraints:

It hasn’t taken off, it’s a good idea, but the practical aspects of hourly rounding when you’ve got a really high acuity ward with staff who are really busy. The theory absolutely, you should see your patients every hour, but the practical stuff around that was really difficult and we were auditing whether it happens every hour, and it doesn’t. (Charge nurse)

The third barrier to improvement was that programmes were neglected. This fate befell the well organised ward programme, which prioritised a tidy working environment:

Sometimes you lose momentum. Everybody’s really engaged to begin with but as time’s gone by people forget. Our ward can be so busy so full on, you get people missing breaks and everything ‘cos they’re so busy, so it gets forgotten. (Senior nurse)

The stalling of this programme was also perceived as a product of culture. Many nurses prioritised time with patients over paperwork and organising activities.

A fourth barrier to improvement was that only economically measureable actions tended to gain resources. Good communication was critically important, but difficult to measure and hard to finance. A fifth barrier was the pressure of so many improvement activities. The cost of ‘overloading’ was lost morale, and possibly reduced quality as staff turned from clinical to bureaucratic work:

It’s a lot of clerical stuff and management. A lot of demand to present stats and audits and so forth, which is taking you off the floor, a lot of paperwork. It has grown over the last year. All the demand is taking you away from the floor. (Charge nurse)

The shift from ward to office forced charge nurses to change their style of leadership. But as they turned away from ward work it became increasingly difficult to monitor events, mentor staff, and assist them to ‘put out fires’. One charge nurse managed this with additional time at work:

The expectation is very clear that it is done within the time that I am allocated to do my job, which is 40 hours a week. If I look at the breakdown of what I am expected to achieve I am in deficit for about 100 hours each month. But it’s because I am so conscientious and passionate that I’ll put in the extra, because I’m workaholic. (Charge nurse)

Another charge nurse preferred to try to reshape implementation:

I’ve had arguments about stuff that I believe in, but I am arguing against because of the way it’s being implemented. (Charge nurse)

Quality improvement: doctors

Doctors were involved in many improvement activities, including practising evidence-based medicine (EBM), and participating in morbidity and mortality conferences and quality improvement projects.

Doctors’ perspectives on EBM shifted with experience. Guidelines were relatively straightforward for juniors, who based most decisions on training and experience. For consultants EBM was more complicated. Except in the case of a doctor mentioning critical appraisal, which was dismissed as unrealistic because of the time it demanded, these doctors referred only to guidelines.

Doctors’ discussions about guidelines identified three ways they could generate confusion and patient harm. First, there were many sources of potentially inconsistent information, including hospitals, professional associations, colleagues, journals, and conferences. Second, guidelines could be impractical for multimorbid patients and vulnerable old patients. Third, guidelines could be used thoughtlessly, leading to over-testing and over-treatment, which increased costs and treatment risks.

A consultant claimed that the confusion around guidelines meant some doctors made decisions ‘according to their own prejudices’. This was perceived as partly due to insufficient clinical education. The situation was not helped by the support systems in both hospitals:

There is an increasing emphasis on the protocols and checklists and procedure lists for everything. That sounds good on paper. In terms of implementation it’s difficult. We have a couple of hundred policies, procedures and guidelines that are all available on the computer but generally lost in the bulk and the number. So a lot of the policies and procedures we have go into our system and hardly ever get seen again. There’s too many and people get a degree of comfort having written a policy, whether it actually changes practice varies. (Clinical director)

In an organisation like this we produce policies for Africa. You have no idea how many clinical policies there are. Does anybody look at them? No, because it’s much easier to write a policy than market a policy. Nobody goes to the trouble of saying, ‘here is our new policy on the treatment of asthma.’ They are hardly communicated. I certainly don’t know what most of them are. It’s a waste of time and effort. I don’t know what evidence there is for having policies for everything. In this hospital all these policies are only roughly followed and nobody knows they’re there until something goes wrong and somebody says ‘what’s your policy?’ (Consultant)

Guidelines and policies also seemed to sometimes express a defensive orientation by the hospitals. If a patient was harmed and a policy was violated it was easy to blame the doctor.

Some efforts were being made to counter the proliferation of guidelines. A consultant had edited a handbook for managing patient issues. A policy review process in one department had reduced but not eliminated the excess. A hospital was introducing a new computer system, delayed by technical issues, to make policy access easier.

There was a variety of perspectives on the morbidity and mortality meetings. Some doctors found them useful because discussions were open, without blame, and learning was emphasised. Others criticised the meetings for lacking rigour due to poorly recorded and unreliable information, insufficient self-criticism, and an emphasis on ‘ticking the boxes’. Outcomes were also questioned. A clinical director acknowledged that it was difficult to identify the systems that needed to change. A medical and a surgical consultant both claimed that necessary systems changes were usually not completed:

‘Oh well it’s been discussed in the M&M meeting,’ full stop and very little changes. They have become an end in themselves, rather than any hard changes to practice or even formal recommendations. People have a misplaced sense of security they are doing something when actually very little happens. (Consultant)

The quality improvement projects that doctors were involved in ranged from discrete components of clinical practice, to systemic change within and across departments.

In one of the surgical departments a consultant had worked for several years on a number of operational projects to improve quality and efficiency. One was guided by an external expert who used process engineering to improve the patient journey and staff workload. The outcome was reduced waiting and time in pre-admission clinics.

Other projects within the department however were unsuccessful. There were resourcing issues and political struggles across departments. A project to improve patient experience during pre-admissions foundered on several occasions and was discontinued. Another, using sentinel event and audit data to try to reduce returns to theatre and preventable deaths, had been difficult to resolve. Although the project identified causative factors, and it was supported by the quality department, it allegedly failed because of senior management decisions and inter-departmental politics. Barriers included the unavailability of resources and problems with culture, especially in relation to gatekeepers controlling outside services. There were delays in getting patients into the operating theatre, difficulties in gaining access to the Intensive Care Unit for patients who needed a higher level of care but didn’t fit admission criteria, and the hospital didn’t support the overnight availability of senior staff. There was pessimism about other projects because throughput pressure was growing, but resources were often not provided for quality improvement:

We’re expected to do a lot of stuff without extra resource. If we are going to look at patients journeys and improving their experiences and decreasing errors and complications then I think senior management need to resource us better to do that, and we need more nurse specialists to do research and map problems. We’re busy clinicians and we’re expected to do all the productivity and improve the quality. (Consultant)

There was ‘half a dozen projects’ that could be attempted in the department if resourcing was available.

Operationally focussed change, supported by senior management, was also evident in the medical department. Roster changes had improved consultant availability. The number of overnight registrars had increased, which improved admissions safety. A High Dependency Unit was planned, and a merger of ED and ward charts was intended to reduce medication harms. But improvements were also compromised by fiscal shortages and insufficient capacity:

I suspect that if we were a commercial enterprise designing cars we would have more people on the design floor making the modifications. We could do with more stepping back from the clinical coal face, analysing the data, seeing where the holes are, making the changes. (Clinical director)

There were many possibilities. These included the system for X-Ray charts, which could go missing, resulting in diagnostic failures because the system would not flag that an X-ray was ordered but not processed. Similarly fax transmission failures could lead to patients not being booked for follow ups, and returning later in an untreatable condition.

Four reasons were proposed to explain the incapacity to perform small improvement projects. First, the quality department were not particularly visible. Second, doctors that took on quality work often suffered burnout. Third, short term economics dominated safety:

All the additional resource that’s been put in is to achieve fiscal efficiency as opposed to safety. It’s very hard to justify in dollar terms investments into safety. (Clinical director)

Fourth, there was lack of expertise. One consultant described doctors’ ideas as impractical, and showing ‘pretty isolated thinking’. Another consultant, who noted that doctors were ‘bloody busy’, emphasised insufficient understanding of methods, and unwillingness to be involved:

It’s a very difficult thing to engineer safety and most people haven’t bought the message because it’s about re-engineering practices and redesigning processes. There are lots of little things that could be done better but people don’t see them as their work. (Consultant)

Overall, doctors expressed very little engagement with quality projects. They were busy with clinical work. Consequently some improvements were done for them, and they did not have time for the committees and meetings where decisions were made. While feedback from consultants about planned changes was sought in one department, the response rate was often around 10%. This disengagement may partly explain the irritation some doctors felt towards some new processes:

Sometimes things that are in place, a check and a balance are well meaning but in the context of our day to day life are not high on the priority list, and the people who put them in place need a reality check on what we actually do. Often it’ll be just something to do with some piece of paperwork that documents or checks something. That’s fine in principle, but in the reality of what our life is actually like it’s so low on the priority list that you almost get offended by having to be asked to do it, because that manager or whoever’s come up with that process clearly has no idea what your life is like. (Consultant)

We have charts for wound care, we have charts for nutrition, we have balance charts to see how well the patients totter to the toilet. We have all these charts. Rarely are they ever filled in completely, rarely properly, and rarely appropriate to the patient. It would be much better if we scrapped a lot of the extra safety charts, and went back to basics, had notes and prescription, and that’s probably it. (Registrar)

Quality improvement: managers

Quality managers in both hospitals identified the busyness of doctors as a barrier to improvement. Time pressure severely impacted cooperation between clinicians and quality experts:

Manager: They haven’t got time to get to meetings. I’m thinking about the global trigger tool project, we haven’t had any clinicians come to meetings. One of them is a surgeon and she’s not there some days and she’s got surgery booked and she just won’t, hasn’t got time. So we never see her. You just can’t engage them.

Interviewer: It must be hard to make the project work in that case?

Manager: Hopeless sometimes. And that varies across services. But they are really, really busy people and so time is a huge factor, a huge restraint. (Quality Manager)

These managers also felt there was culturally based scepticism amongst doctors about the need and the possibility of improvement. Doctors seemed to think of quality as the product of individual expertise rather than teamwork and systemic change. If the work was necessary they often wanted someone else to do it. Quality was an add-on, not a core component of the work, and a very high level of proof was necessary before they were comfortable putting it into practice.

Discussion

This research has investigated the challenges of safety improvement for staff in three departments in two NZ public hospitals. The commentary staff provided can be interpreted through the Organising for Quality (OQ) model (Bate et al. 2008), which identifies the need to overcome six universal improvement challenges of structure, culture, politics, motivation, learning, and infrastructure. The structural and cultural challenges are critically important because their solutions are often interdependent, and they tend to act as supports for solutions to all of the other improvement challenges. In the departments studied there appeared to be some weakness in both of these foundational aspects of improvement. Structurally, there seemed to be an absence of ‘organisational slack for quality’, which would allow staff to ‘stand back from everyday operations and think and work on service development’ (Bate et al. 2008, p. 179). Related to this there also appeared to be an absence of whole-system design, which was evident when some projects that crossed departmental boundaries seemed to become vulnerable to politically motivated conflict. These absences could compromise staff capacity to initiate and sustain improvement work.

The impact of adverse local cultures on improvement processes was evident in the way the scientific culture of data, measurement, and evidence was sometimes compromised by insufficient expertise, capacity, and analysis of inter- and intra-departmental systems. A culture of mindfulness, or vigilance to safety, was not evident in the reporting system, which was intended to promote safety vigilance, but was mostly disused and perceived as ineffectual. Staff also sometimes felt, or appeared, to be politically disconnected from decisions about improvement. They disputed practices including performance targets, the conduct of some nurses’ improvement projects, and procedural changes to medical and surgical work. With exceptions, they expressed little ownership of improvement, and those who did risked frustration and burnout. This lack of empowerment appeared to compromise motivation and contribute to a lack of collective momentum for change. The need to keep pace with change was also held back by the information technology infrastructure, which did not support the effective implementation of evidence-based medicine.

Overall, this research suggests that while some improvement processes in the departments studied were effective, difficulties were widespread and deeply embedded. Staff narratives about these troubles indicate the particular relevance of two priorities. First, the need for additional dedicated resourcing so staff can be equipped with the time, expertise, and technology necessary to work on improvement. Second, efforts to empower and motivate staff to exercise more control over improvement actions. To do this, they may need to be convinced that change is necessary, they may need methodological expertise, and they may need to see that implementation can be achieved with practical methods that support, rather than contradict, their other priorities–workflow especially. An important lesson from the OQ model about implementing improvement is that deep understanding of localised nuance is necessary to design solutions that sensitively fit the contexts they address. Qualitative research of staff experience can provide one source of this knowledge. Social theory about complex organisations, professions, and innovation offers transferrable knowledge about what works in other contexts (but may need to be translated for implementation to be effective). Synthesising these knowledge sources with the practical know-how of hospital staff can drive the development of what Bate et al. (2008, p. 9) call ‘a head full of theories’, in other words a collection of creative resources about how to do organisational change. Empowering staff to achieve this and giving them the time to do it would benefit safety improvement.

Disclosure statement

No potential conflict of interest was reported by the author.

ORCID

Peter Beaver http://orcid.org/0000-0001-8026-3406

Additional information

Funding

This research was made possible by Doctoral Scholarships from the University of Auckland and the Health Research Council of New Zealand.