Degree Name

Doctor of Philosophy


Faculty of Health & Behavioural Science


Background: It is increasingly apparent that the existing health professional workforce is insufficient to meet the growing demands of health care. Within Australia and globally, an increasing demand for quality health care workers will impact substantially upon the future of the industry (International Council of Nurses (ICN) 2010). Driven by this trend, some healthcare stakeholders are giving consideration to the magnet hospital structure as a framework for the development of effective professional practice environments. The identified features of a magnet hospital include a decentralised administration, participation in decision making, supportive managers and autonomy and career development opportunities (McClure Poulin, Sovie & Wandelt 1983). There has been extensive research evidence illustrating the success of magnet hospitals in retaining nursing staff, which has been found to contribute to positive patient outcomes (Aiken, Smith & Lake 1994; Aiken, Sloane & Lake 1997; Aiken et al 2001; Upenieks 2003). However, the transferability of the concept to Australia has yet to be comprehensively examined.

Aim: The aims of this research were to: (1) adapt a tool for measuring magnet features that relates to the Australian context; (2) test the reliability and validity of this adapted tool; and (3) use the tool to measure magnet features and investigate their relationship to measures of job satisfaction and staff intention to leave, among a sample of nurses in Australian health facilities.

Research Design: The research included qualitative and quantitative research approaches and mixed methods approach in the adaption and testing of an Australian tool for measuring magnet features. Three inter connected studies were undertaken to address the research aims and questions. The first study involved focus groups with registered nurses who reviewed an established North American tool to assess its applicability for use in the Australian context. The second study was a pilot survey that tested the Australian tool ‘The Nursing Work Index–Revised: Australian (NWI-R:A)’ establishing it as a credible measure of magnet features. The third study, a larger scale survey, examined the NWI-R:A as a descriptive measure of magnet features in four Australian facilities in conjunction with measures of staff job satisfaction and intention to leave for the purpose of exploring relationships between these variables.

Results: Overall the findings suggest that the NWI-R:A is a valid and reliable measure of magnet features with statistically acceptable internal consistency (Cronbach Alpha 0.76) for the tool. Study One established that the Australian version of the tool needed to address issues with the language, presentation and meaning to improve the transferability to the Australian context. In studies two and three the perceptions of the registered nurse sample (n=326) consistently showed that these cohorts viewed the quality of care, nursing management and leadership, and collegial relationships of their respective workplaces favourably, but perceived that nurse participation in decision making; staffing; and resources were lacking. Finally, Study Three results identified statistically significant relationships between registered nurses’ perceptions of magnetism, job satisfaction and their reported intention to leave.

Discussion: The establishment of a tool for the reliable measurement of magnetism in Australian facilities enables the magnet concept to be effectively transferred to the Australian healthcare environment. A more comprehensive understanding of the ways in which nursing staff perceive existing magnet features in their workplace potentially informs targeted development of nursing practice environments in Australian health facilities. If the magnet framework was to become central to the development of organisational structure and governance, healthcare facilities in Australia would be well positioned to improve nursing staff retention.



Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily represent the views of the University of Wollongong.