Article Text

Download PDFPDF

Improving patient safety: the comparative views of patient-safety specialists, workforce staff and managers
  1. Jeffrey Braithwaite1,2,
  2. Mary T Westbrook1,2,
  3. Maureen Robinson3,
  4. Sarah Michael4,
  5. Christy Pirone5,
  6. Philip Robinson6
  1. 1Centre for Clinical Governance Research, Faculty of Medicine, University of New South Wales, Sydney, Australia
  2. 2Australian Institute of Health Innovation, Faculty of Medicine, University of New South Wales, Sydney, Australia
  3. 3Quorus Consulting, Sydney, Australia
  4. 4Communio Pty Ltd, Sydney, Australia
  5. 5Safety and Quality Clinical Systems, Department of Health, Adelaide, South Australia
  6. 6Children, Youth and Women's Health Service, Department of Health, Adelaide, South Australia
  1. Correspondence to Professor Jeffrey Braithwaite, Australian Institute of Health Innovation, Faculty of Medicine, University of New South Wales, NSW 2052, Australia; j.braithwaite{at}


Introduction Methods for improving patient safety are predicated on cooperation between healthcare groups, but are the views of health professionals involved in promoting safety shared by other healthcare workforce staff and managers?

Aim To compare patient-safety suggestions from health workforce managerial and staff groups with those of patient-safety specialists.

Method Samples of managers (424) and staff (1214) in the South Australian state health system and 131 Australian patient-safety specialists were asked to write suggestions for improving patient safety. Group responses were content analysed and compared.

Results Patient-safety specialists (83.2%) were more likely to make suggestions than were workforce managers (57.8%) or staff (44.1%). Workforce members from clinical professions were more likely than non-clinicians to tender suggestions. No relationship existed between the importance specialists and managers (ρ=−0.062, p=0.880) and specialists and staff (ρ=−0.046, p=0.912) attached to nine categories of suggestions. There was a high correlation between the importance that managers and non-managers attached to safety strategies (ρ=0.817, p=0.011). Among those who made suggestions, specialists were more likely to suggest implementing reviews and guidelines, and incident reporting. Workforce groups were more likely to recommend increased and improved staffing and staffing conditions, and better equipment and infrastructure. There were no significant differences in the proportions of group members recommending: improving management and leadership; increasing staff safety education and supervision; communication and teamwork; improved patient focus; or tackling specific safety projects.

Implications Differences between safety specialists' and workforce groups' beliefs about how to improve patient safety may impede the successful implementation of patient-safety programmes.

  • Patient safety
  • workforce
  • health professionals
  • patient-safety specialists
  • attitudes
  • values
  • suggestions for improvement
  • healthcare quality improvement
  • patient safety

Statistics from


In a review of patient-safety research for the WHO World Alliance for Patient Safety, Jha et al1 reminded us that prevalence studies indicate that 3–16% of hospitalised patients suffer harm from adverse treatment. They classified causes of patient harm into those arising from structural, process and outcome factors following Donabedian's longstanding distinction. Structural causes of harm included inadequate staffing (both in terms of numbers and training), the work environment, pressure on staff, provider fatigue, communication breakdown, poor organisational safety culture and lack of regulations and accreditation. Process associated causes included misdiagnosis, inadequate follow-up of test results, use of counterfeit and substandard drugs, and unsafe injection procedures. Some outcomes of unsafe care were medication and surgical errors, device-related injuries, falls in hospital and healthcare-associated infections. The categorisation was seen as providing a practical rather than a definitive framework for addressing patient safety, as the strong links between structural and process factors meant that some causes could arguably be placed in either group.

On the basis of research evidence, Wong and Beglaryan2 recommended eight strategies for improving patient safety in hospitals: better communication within clinical teams, reporting of adverse events, increasing patient involvement, developing protocols and guidelines, managing human resources (eg, increasing staff-to-patients ratios, reducing staff overwork and fatigue), managements' commitment to a safety culture, public disclosure of adverse events and safety education for healthcare providers. A 10-year review of the technical literature on patient-safety research3 identified 21 types of remedial strategies that had been recommended to improve patient safety.

To what extent are these views on how to improve patient safety held by members of healthcare systems? In acute healthcare, there are three groups with prime responsibility for the safety of patients: workforce staff, healthcare managers, and specifically trained quality and patient-safety staff. The workforce comprises doctors, nurses and allied health personnel striving to treat patients by providing high quality and safe services, as well as administrative, technical, scientific and support staff. Managers are those with delineated, formal responsibilities for running services, divisions, departments or units, involved with budgets, staffing and administering various organisational functions. Patient-safety specialists are mainly concerned with promoting enhanced quality of care and patient safety, conducting or facilitating projects, programmes, initiatives and related activities to support improvements. A priori, we can hypothesise that patient-safety strategies would work best if these groups shared similar values about improving care, and held comparable attitudes on how to go about making improvements and delivering safe services. To our knowledge, no previous study has examined whether, or the extent to which, the extant values and attitudes of healthcare groups are held in common.

Patient-safety specialists are more likely than most healthcare providers to be involved in safety policy and programme issues. Thus, they would be expected to know more about patient safety than would other groups and to recommend safety strategies supported in the patient-safety literature. Healthcare providers with backgrounds in the clinical professions would be more likely than those who come from non-clinical professions and provide indirect patient care to have direct experience of adverse events. This should lead them to have greater knowledge of ways to improve patient safety. A comparative study of attitudes towards patient safety in three health districts in New Zealand4 revealed significant differences in the safety attitudes of clinical staff and staff providing indirect patient care as measured by the Safety Attitudes Questionnaire.5

Managers have responsibility for dealing with the outcomes of adverse events and systems level change. Thus, they would be expected to have more knowledge of patient-safety issues and may advocate more and different strategies for improving safety than their staff do. A study of the workforce in a large hospital which had implemented a new organisational structure6 found that managers' attitudes towards the change were more favourable, intense and less polarised than were those of staff whose views were often more neutral or negative, and lacked consensus. When the state of New South Wales, Australia, introduced an electronic Incident Information Management System for the reporting of adverse events, healthcare providers with a managerial role were significantly more positive in their evaluation of almost all aspects of the system than were non-managerial staff.7


The research aim was to investigate and compare the suggestions for improving patient safety made by managers and staff in a large healthcare system and a group of patient-safety specialists. An additional aim was to examine which safety strategies promulgated in the research literature received most and least attention from these groups of healthcare providers. It was hypothesised that:

  1. Reflecting the varying involvement of the three groups in patient-safety issues, specialists would make more suggestions than health workforce managers who would make more suggestions than their staff. This was tested by comparing the number of members of the three groups who made suggestions and the actual number of suggestions made by those who responded from each group (whom we refer to as ‘engaged’ participants).

  2. Group members with a background in the clinical professions would be more likely to make suggestions for improving patient safety than their colleagues with non-clinical backgrounds.

  3. The groups would differ in the types of strategy they recommended. This was tested first by comparing the correlations between the frequency with which members of the three groups recommended various types of safety strategy. It was predicted that there would be positive correlations between the views of the specialist and manager groups, and between the managers and their staff, with least correlation occurring between the specialist and workforce staff groups. Second, this was tested by comparing the proportions of group members who recommended specific strategies in order to identify areas of agreement and disagreement.


Samples and questionnaires

The patient-safety specialist sample consisted of 131 senior participants who attended one of four 2-day workshops for staff in Australian state health systems whose work actively involved the promotion of patient safety. The workshops were presented by educators from the University of New South Wales, Australia and were specifically designed for such professionals. Of the 127 who indicated what they considered to be their current primary work role, 78.0% checked ‘manager,’ 13.4% ‘policy maker’ and 16.5% ‘clinician’ (some participants ticked two options). Overall, 90.9% described themselves as primarily a manager and/or policy maker. Attendees completed a questionnaire prior to the commencement of the workshops. This consisted of a modified version of the questionnaire devised by Hindle et al3 to assess a health workplace prior to the development of a safety work plan. Respondents were asked to use five-point scales to rate their workplace or organisation in terms of communication and teamwork between staff groups, effectiveness of meetings, continuous learning, sharing of responsibility, leadership, information systems and consumer involvement. They were also asked to provide demographic information and write their top three suggestions for improving patient safety.

The workforce samples consisted of 1662 members of the health workforce in the state of South Australia (population 1.6 million). They represented a 10% random sample drawn from 16 619 respondents (52% of the state's public health workforce) who completed a questionnaire survey of attitudes towards patient safety.8 The questionnaire consisted of a slightly modified version of the Safety Attitudes Questionnaire5 which asked respondents to describe the safety and teamwork climates of their workplace, their job satisfaction, stress recognition, and perception of management and their working conditions. Respondents also provided demographic information. Of the 1638 who reported their organisational position, 424 (25.9%) had a managerial component in their role; that is, they checked their organisational role as ‘executive, senior manager, middle manager, line manager or team leader/supervisor,’ and 1214 (74.1%) did not have managerial responsibilities. This information was used to divide the workforce members into the managerial and staff workforce groups. Participants were asked to write their top three suggestions for improving patient safety.

Table 1 lists the demographic characteristics of the specialist, managerial and staff samples. χ2 analyses comparing the groups' characteristics revealed that the staff sample included significantly more female, and not surprisingly younger and less experienced persons than did the managerial and specialist groups. The specialists were the oldest group and had a significantly greater proportion of members with a background in the clinical professions. The workforce managers group resembled the staff sample in professional backgrounds and the safety specialists in gender composition. The high femininity of all groups was underscored by 10 of the 12 professional groups in the study being predominantly female. All but one profession had a greater proportion of males in its managerial than staff samples.

Table 1

Demographic characteristics of survey groups


Participants' safety suggestions were content analysed into nine categories (see list of categories in box 1 with examples). This schema was developed using procedures originally suggested by Glaser and Strauss's9 grounded theory, with the categories emerging from the suggestions. Inter-rater reliability was high; there was 98% agreement of categorisation of strategies made by two independent scorers. χ2 analyses and ANOVAs were used to compare the suggestions of different groups of participants. Bonferroni corrections for multiple comparisons were conducted to ascertain which groups differed significantly from each other. Spearman rank order correlations were calculated to examine the relationship of frequency of types of suggestions made by the three groups.

Box 1

Patient-safety suggestions: content analysis categories and examples

  1. Improve incident reporting

    • Encourage honest no-blame reporting of incidents

    • The incident-reporting system is difficult to use, and you never get any feedback

    • Taking reportable events seriously and acting on them

  2. Increase staff education and supervision

    • Mandatory regular clinical supervision

    • Continue professional development for all staff

    • Continuing education for staff regarding health and safety issues for patients and staff

  3. Implementation of guidelines and reviews

    • Staff adherence to guidelines and protocols

    • Continued audit for identified problems and routine policy/procedures

    • Random audit of files

  4. Better management and leadership

    • Total review of management structure

    • Hierarchy to work on the floor with staff at the coal face to see what actually occurs

    • More open management—more approachable

  5. Improve communication and teamwork

    • Communication across disciplines

    • More focus on teambuilding activities

    • Address issues of lack of communication between health professionals

  6. Improve staffing (numbers and quality) and staff conditions

    • Increase staff numbers

    • Improve nurses' job satisfaction so they do not leave

    • Ensure staffing levels and skills are adequate for every shift

  7. Acquire and maintain better equipment and infrastructure

    • Adequate modern diagnostic equipment

    • Better functioning equipment and more frequent maintenance of equipment

    • More room in cubicles, more accessible power points, oxygen and suction outlets

  8. Increase patient focus

    • Healthcare workers who put patients first

    • Meetings with patient's family to provide support

    • Development of patient-focused vision and culture

  9. Target specific issues

    • Ensure that the smoke-free policy is implemented

    • Improve stroke services

    • Reduce waiting lists


Characteristics associated with making suggestions

Table 2 shows the results of tests examining characteristics associated with making safety suggestions. As predicted patient-safety specialists (83.2% of group) were significantly more likely than workforce managers (57.8%) and workforce non-managers (44.1%) to answer the question asking for suggestions for improving patient safety. For the total groups, the average number of suggestions made by members was 2.14 for the specialist, 1.55 for the managerial, and 1.12 for the staff groups. The ANOVA comparing these means was significant (p=0.000). Comparison of the means indicated that the specialist group made significantly more suggestions than did the managers (t=4.35, df 551). The managerial group made significantly more suggestions compared with the staff members (t=5.62, df 1636) and the specialists made more suggestions than did the staff group (t=8.18, df 1343). All comparisons had p values of <0.001.

Table 2

Characteristics of group members who did or did not make safety suggestions

We examined whether professional background was related to making suggestions that is being ‘engaged.’ Among specialists, having a clinical or non-clinical professional background was not associated with making suggestions. In fact, 83% of the specialists had a clinical background. In the workforce managers' sample participants with a clinical background (64.9%) were significantly more likely than those from a non-clinical profession to make suggestions (45.4%). In the workforce staff group 53.2% of those with a clinical background were ‘engaged’ while only 27.8% of staff with a non-clinical background made suggestions. Thus, in the two workforce samples, groups from non-clinical professions were much less likely to demonstrate engagement in safety issues. The average number of suggestions made by ‘engaged’ members of the specialist (mean=2.58 suggestions), managerial (mean=2.69) and staff (mean=2.54) groups were compared using ANOVA. This was significant (p=0.025). Post hoc comparisons of the means indicated that the managerial group made significantly more suggestions than did the staff group (t=2.54, df 779, p<0.02) but the specialist group mean did not differ significantly from the means of either workforce group.

Types of safety suggestions made by groups

We compared the frequency of the types of safety suggestions by the three groups. The specialist sample made 281 suggestions, the managers' group made 659 and the workforce staff made 1366 suggestions. Table 3 provides the numbers and percentages of these suggestions which were classified into the nine content analysis categories shown in box 1. For example, 6.0% of specialists' suggestions were for improved staffing compared with 24.6% of managers' and 26.9% of the staff's suggestions. Spearman rank order correlations revealed a highly significant relationship between the rankings of the two workplace samples (ρ=0.817, N=9, p=0.011). However, the specialists' rankings were not related to those of the managers (ρ=−0.062, N=9, p=0.880) or the staff (ρ=−0.046, N=9, p=0.912).

Table 3

Number and ranks of suggestions for improving patient safety made by health workforce staff, managers and patient-safety specialists

We examined the proportions of ‘engaged’ members from the three samples who made each type of suggestion using χ2 analyses (table 4). Significant differences were found regarding four strategies. Patient-safety specialists (48.6%) were significantly more likely to recommend the implementation of guidelines and reviews than were workforce managers (13.9%) or staff (10.1%). Specialists (22.9%) were also more likely to recommend improved incident reporting than were managers (13.9%) or staff (9.3%). Almost half of managers (49.4%) and staff (46.8%) favoured improved staffing while only 12.8% of specialists made this recommendation. Workforce managers (26.9%) and staff (21.5%) were significantly more likely than specialists (12.8%) to suggest improvements in equipment and infrastructure. There were no significant differences in the proportions of the two groups making suggestions directed at improving staff eduction and supervision, management and leadership, communication and teamwork, patient focus, or undertaking specific targets.

Table 4

Results of χ2 analyses comparing workforce staff, managers and patient-safety specialists making various suggestions


Significant differences were found between the percentages of members in the three groups who made suggestions. Patient-safety specialists were most, and staff least, likely to make suggestions. These differences would seem to reflect the groups' differing involvement in, and knowledge of, patient-safety issues. However, the predictions that specialists who answered the question would make more suggestions than ‘engaged’ managers, who would make more suggestions than ‘engaged’ staff, were only partially supported. Managers who gave suggestions listed significantly more strategies than did staff. However, ‘engaged’ specialists did not cite significantly more suggestions than ‘engaged’ members of either workforce group. These findings suggest that although fewer managers than specialists held views on improving patient safety, the managers that were ‘engaged’ were as deeply involved as specialists. The questionnaire instructions asking for three suggestions may have resulted in some participants, particularly specialists, tendering fewer suggestions than they might have otherwise contributed.

The second hypothesis was that group members who had a background in the clinical professions, even though their roles might now be non-clinical, were more likely than their non-clinician colleagues to make suggestions about improving patient safety. The prediction was supported for the managerial and staff groups but not the specialist group. However, the vast majority (83%) of the specialist group came from clinical professions. This suggests the attraction of the role of patient-safety specialist for health professionals who are or have been direct care providers. The relative lack of engagement of non-clinical staff suggests that when specialists design safety improvement programmes for healthcare providers these may need to be tailored to the different knowledge and experiences of clinical and non-clinical staff.

The third prediction, that the groups would differ in the types of strategies they recommended, was only partially supported and not entirely as expected. The specialists' preferred strategies did not resemble those of managers more than they resembled those of staff. In fact, there was no relationship between the rankings of the specialists' preferences and those of either of the workplace groups. The very high degree of similarity between the preferred strategies of the managers and staff was greater than expected. Although the findings suggest much greater involvement in patient-safety issues on the part of managers than their staff, this involvement does not seem to have led managers to espouse more of the preferred strategies of patient-safety specialists.

The ‘engaged’ groups held significantly different preferences regarding four strategies. Significantly more patient-safety specialists supported the adoption of reviews and guidelines and the reporting of adverse incidents. This may in part be a consequence of their role and position in healthcare rather than their specialised knowledge. The workforce groups were more in favour of improvements to staffing, particularly increasing staff numbers; and acquiring and maintaining equipment and infrastructure. These findings highlight a gulf between safety specialists and workplace groups regarding patient-safety initiatives. There was a much greater emphasis on the need for increased resources, both human and physical, by the groups ‘at the coalface.’ The specialists may be restrained in making such suggestions by knowledge of the actual limitation of finances available to provide such resources. The two strategies preferred by specialists, viz regulatory strategies and incident reporting, may be perceived as less desirable by workplace groups, as their implementation may add to their workloads. Such patient-safety strategies, however desirable, come at a burden for workplace groups—for example, reporting an incident includes finding time and a computer. Data entry may be difficult if the staff member has low IT skills or finds the programme difficult to navigate. Given the emphasis on incident reporting in the patient-safety literature and in specific safety programmes,10 it is perhaps surprising that although specialists were more likely to recommend reporting than were other groups, they only ranked it 5.5 out of the nine strategies. The ‘engaged’ members of the three groups held similar attitudes on the importance of increased staff education and supervision, improved management and leadership, better communication and teamwork, more focus on patients' needs, and targeting specific safety problems. Thus, there were important areas of agreement between the groups.

Examination of the categories of patient-safety suggestions elicited via the content analysis shows that they include seven of the eight research validated strategies recommended by Wong and Beglaryan.2 Disclosure did not emerge as a category, although it was occasionally mentioned by participants in the context of providing patient-centred care. Our category of supply and maintenance of equipment and infrastructure was not on Wong and Beglaryan's list but was on Hindle et al's.3 The latter commented that the Institute of Medicine11 study also ‘made little reference to resource shortfalls, on the grounds that the challenge is to make better use of available material no matter how limited’ (p. 22). In terms of Jha et al's1 classification of safety strategies, those arising from the content analysis were mainly aimed at allievating structural causes of harm. Strategies concerned with process causes and outcomes of unsafe care such as misdiagnosis and inadequate follow-up were addressed to some extent through mechanisms such as incident reporting and regulations such as guidelines and audits. Additionally, some of the suggestions classified in the content analysis as 'targetting specific issues' would be assigned to these two categories proposed by Jha et al. Examples were: set up a falls programme, double-checking all medications and improving hand hygiene. To some extent, respondents' reduced emphasis on strategies for dealing with process and outcome problems might be attributed to the effects of the considerable resources allocations and efforts that have been made to address these problems in developed countries.

A limitation of the research was differences in the content and context of the surveys undertaken by the participants. All completed a questionnaire that assessed their workplace conditions before giving their suggestions for improving patient safety. Although the items in both questionnaires were developed to assess a workplace's safety by comparing answers with what is known about conditions that affect such safety, none of the questions explicitly refer to improving patient safety or strategies for doing so. However, the content of items may have influenced ideas that came to mind when paticipants gave their suggestions. This does not seem to have occurred. The workplace groups' questionnaire had two items about error reporting and one about guidelines while the specialists' questionnaire had none, but significantly more specialists advocated improving incident reporting and use of guidelines. There were more items on consumer focus, continuing education and teamwork in the specialists' questionnaire, but all survey groups made similar proportions of suggestions related to these topics. Unlike the workforce groups, the specialists were not in the workplace at the time of their completion of the questionnaires.


Our examination of the attitudes of healthcare workforce staff, managers and patient-safety specialists revealed levels of disagreement about the patient-safety improvement agenda. Those at the clinical coalface, and to a lesser extent those responsible for managing it, were less likely to express views about improving patient safety than were specialists. When they made suggestions, staff and managers were much more likely to advocate acquiring more staff, better conditions for staff, and improved equipment and infrastructure. In reality, lack of resources is likely to limit the degree to which specialist-initiated patient-safety programmes can achieve these goals. On the other hand, patient-safety specialists favour some strategies which may involve staff commitment of time and effort. The lesser engagement among workplace groups, particularly staff, suggests lack of knowledge of safety issues. However, the similar emphasis given by ‘engaged’ members of all healthcare groups to education, communication, patient-focused care, and improved management and leadership suggests that common ground can be achieved in the implementation of safety-improvement programmes. Attitudinal differences between groups responsible for patient safety in the health system seem to be about categorising the agenda and prioritising some issues over others, rather than the agenda itself.


We thank participants for their contributions to the surveys, D Debono for assistance with inter-rater reliability testing, and two referees for their useful comments on an earlier draft of the paper.



  • Funding JB receives funding for research into patient safety from the National Health and Medical Research Council. This research is supported in part by National Health and Medical Research Council Programme Grant 568612.

  • Competing interests None.

  • Ethics approval Ethics approval was provided by the South Australian Department of Health's Human Research Ethics Committee.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.