Lessons learned from measuring safety culture: An Australian case study

Abstract

Background

adverse events in maternity care are relatively common but often avoidable. International patient safety strategies advocate measuring safety culture as a strategy to improve patient safety. Evidence suggests it is necessary to fully understand the safety culture of an organisation to make improvements to patient safety.

Aim

this paper reports a case study examining the safety culture in one maternity service in Australia and considers the benefits of using surveys and interviews to understand safety culture as an approach to identify possible strategies to improve patient safety in this setting.

Setting

the study took place in one maternity service in two public hospitals in NSW, Australia. Concurrently, both hospitals were undergoing an organisational restructure which was part of a major health reform agenda. The priorities of the reform included improving the quality of care and patient safety; and, creating a more efficient health system by reducing administration inefficiencies and duplication.

Design

a descriptive case study using three approaches:

• Safety Attitudes Questionnaire and Safety Climate Scale surveys administered to maternity health professionals (59/210, 28% response rate) measured six safety culture domains: Safety climate, Teamwork climate, Job satisfaction, Perceptions of management, Stress recognition and Working conditions.

• Semi-structured interviews (15) with key maternity, clinical governance and policy stakeholders augmented the survey data and explored the complex issues associated with safety culture.

• A policy audit and chronological mapping of the key policies influencing safety culture identified through the surveys and interviews within the maternity service

Findings

the safety culture was identified to warrant improvement across all six safety culture domains. There was reduced infrastructure and capacity to support incident management activities required to improve safety, which was influenced by instability from the organisational restructure. There was a perceived lack of leadership at all levels to drive safety and quality and improving the safety culture was neither a key priority nor was it valued by the organisation.

Conclusion

the safety culture was complex as was undertaking this study. We were unable to achieve a desired 60% response rate highlighting the limitations of using safety culture surveys in isolation as a strategy to improve safety culture. Qualitative interviews provided greater insight into the factors influencing the safety culture. The findings of this study provide evidence of the benefits of including qualitative methods with quantitative surveys when examining safety culture. Undertaking research in this way requires local engagement, commitment and capacity from the study site. The absence of these factors is likely to limit the practicality of this approach in the clinical setting.

Significance

the use of safety culture surveys as the only method of assessing safety culture is of limited value in identifying strategies to potentially improve the safety culture.

Introduction

Adverse events in maternity care are relatively common but often avoidable. National and international patient safety strategies and policy identify the importance of developing positive safety cultures to improve patient safety (Australian Council for Safety and Quality, 2002, NPSA, 2004, NSW Health, 2004, Barraclough and Birch, 2006). Evidence suggests it is necessary to understand the safety culture of an organisation to make improvements to patient safety.

Improving safety culture in the health-care setting is a key strategy being implemented in a number of countries to improve patient safety in health care (Kohn et al., 2001, NPSA, 2004, NSW Health, 2004, Pronovost et al., 2005) and reviewing the safety culture is recommended as a patient safety strategy (NPSA, 2004, Kirk, 2005, Hindle et al., 2006). There is only limited evidence that the measurement of safety culture at the ward or clinical unit level has resulted in improvements in safety culture. Whilst the results of these few international studies are promising (Pronovost et al., 2005), it is unknown if this process is either generalisable or offers a practical method in the clinical maternity setting.

There are many variations in the interpretations and definitions of safety culture. Some define safety culture as a sub-facet of organisational culture that affects the attitudes and behaviours of members with regard to the health and safety performance of an organisation (Cooper, 2000, Kirk et al., 2007). Safety culture is also defined as ‘a product of individual and group values, attitudes, perceptions, competencies and patterns of behaviour that determine the commitment to, and the style and proficiency of an organisation’s health and safety management’ (Sexton et al., 2006a). The culture of an organisation, or the attitudes and beliefs of those health professionals within it, will influence its ability to respond to adverse events.

A common interpretation of culture is ‘the way things are done around here’ (Pronovost and Sexton, 2005, p. 231). Positive safety cultures in health care were identified to include strong leadership to drive the safety culture and a strong management commitment where safety is made the key priority for the organisation (Perry, 2002, NPSA, 2004, Hindle et al., 2006). Leadership and management commitment in this context was considered to be important as their actions and attitudes are thought to influence the perceptions, attitudes and behaviours of staff in the organisation towards safety culture (Flin, 2007). Organisations with positive safety cultures have: staff who are constantly aware that things can go wrong; have an acknowledgement at all levels of the organisation that mistakes occur; and a strong organisational commitment and ability to learn and take action to prevent their reoccurrence (NPSA, 2004).

Patient safety culture is influenced by a number of factors collectively described as safety culture dimensions or domains (NPSA, 2004, Sexton et al., 2004, Kirk, 2005, Flin et al., 2006, Sexton et al., 2006a, Singla et al., 2006, Kirk et al., 2007). Safety culture domains are broadly reported to include organisational, work environment, team and staff factors (Vincent et al., 2000, Hindle et al., 2006). There is no agreed classification or definition describing these patient safety culture domains in the literature (Singla et al., 2006). A number of reviews of existing patient safety culture surveys conducted by various authors has resulted in views about the fundamental domains influencing patient safety culture (Flin, 2006; Singla et al., 2006; Colla, 2005; Fleming, 2008).

Dimensions of safety culture includes: management and supervision, safety systems, risk, work pressure, competence, procedures and rules, teamwork, communication, organisational learning, feedback and communication, beliefs about the cause of errors and adverse events, job satisfaction and overall perception of safety (Singla et al., 2006). Other work highlights the importance of leadership; safety systems and risk perception; job demands; organisational reporting; teamwork, communication and feedback, physical resources and safety attitudes in understanding safety culture (Colla et al., 2005; Flin et al., 2006, Singla et al., 2006; Fleming and Wentzell, 2008). Sexton and colleagues postulate that there are six specific patient safety culture related domains in the health-care setting: (1) Safety climate, (2) Teamwork climate, (3) Job satisfaction, (4) Perceptions of management, (5) stress recognition and (6) Working conditions (Sexton et al., 2004). The factors included in each of the six safety culture domains are summarised below (Table 1).

Safety culture surveys examining the various dimensions are being advocated as one way to examine health services (NPSA, 2004, Kirk, 2005, Hindle et al., 2006). These surveys quantitatively measure various domains safety culture by identifying strengths or weaknesses perceived by respondents (Flin, 2007). Safety culture surveys however, only provide a snapshot of an organisation’s safety culture sometimes also described in the literature as climate (Pronovost and Sexton, 2005, Sexton et al., 2006b). As such, they are limited and often only provide superficial understanding about the safety culture (Kirk et al., 2007). In addition, qualitative research should be undertaken to examine the human factor components of cultures (Braithwaite et al., 2005, Flin, 2007, Kirk et al., 2007). Qualitative methods can explore the safety culture and to identify specific interventions (Flin et al., 2006, Perneger, 2006). There are limited studies which describe safety cultures in clinical settings using both safety culture surveys and qualitative methods such as interviews.

There is little known about the safety culture in Australian maternity services. In an effort to understand whether measuring and examining the safety culture using both surveys and interviews was a practical method which could help improve patient safety in the clinical maternity setting, we undertook a case study to explore the safety culture in one maternity service in Australia.

We originally designed the study to measure at two time points and examine the safety culture to inform the development and implementation of improvement strategies and then remeasure the culture for changes post intervention. However, early challenges in gaining ongoing local stakeholder engagement and limited capacity to support the study combined with and low response rates to surveys indicated that this plan was premature. The study was then revised to measure and examine the safety culture and identify the barriers and challenges to improving the safety culture in this setting. This paper therefore reports the study as a case study and considers the processes designed to understand safety culture as an approach to identify possible strategies to improve patient safety in this setting.

The study took place in one maternity service located in two public hospitals in NSW, Australia. Concurrently during the study, a health reform agenda in NSW, Planning Better Health (NSW Health, 2004) was being rolled out. The priorities of the reform agenda included, improving the quality of care and patient safety; and, creating a more efficient health system by reducing administration inefficiencies and duplication. The reform agenda included a number of policies such as, The Patient Safety and Clinical Quality Program providing a framework for the systematic identification, reporting and management of events and risks; and, the organisational restructure of all NSW Health Services. The restructure resulted in changes across the whole of NSW, where the existing 17 Area Health Services (AHS) were restructured into eight new AHS (NSW Health, 2004).

The restructure resulted in two AHS amalgamating. This required significant reorganisation of clinical services into area clinical streams and smaller divisions within new governance structures. At the study site, this resulted in one maternity service amalgamating with a second in a facility located eight kilometres away. The maternity service operated as one maternity service across two sites. The restructure resulted in changes to, and the displacement of, key local stakeholders shortly after the study commenced. The changes to leadership at the study site meant a period of instability for the existing quality and safety infrastructure of the service and the staff.