Nine per cent of nurses across Europe report intent to leave their profession, with burnout among the associated personal and professional factors
- Correspondence to
: Dr Michael Simon
Faculty of Health Sciences, University of Southampton, Highfield Campus Southampton, Southampton SO17 1BJ, UK;
Implications for practice and research
The association between leaving intentions and burnout gives policymakers and nurse managers the opportunity to pay attention to burnout and its precursors as potential drivers of professional leaving.
The study has the same limitations of many cross-sectional studies by focusing on work life and omitting contextual factors such as economic and family background.
Future studies should consider actual leaving rather than intentions, adopt more developed models of organisational leaving1 and prospective designs.
Nursing shortages are a recurrent problem in healthcare systems worldwide. Two basic mechanisms are considered to maintain and develop a sufficient nursing workforce: first, increasing the inflow of personnel into the active nursing workforce (eg, through training or immigration) and second, decreasing the outflow of nurses from the active workforce. A representative study of nurses in the USA found a relatively small percentage (3.5%) of nurses changing their career.2 The focus of this study was identifying the association of nurse and work environment characteristics and leaving intentions in comparison of the participating countries.
Ajzen's3 theory of planned behaviour is often referred to as a theoretical framework to explain the association of intentions and behaviour (actual leaving). Although some research supports the association between the intention and the behaviour of leaving the profession, there is a recognisable gap between intention and behaviour and the association of both is even less straightforward.4 There is an immense amount of primarily cross-sectional research in relation to organisational and professional leaving intentions in nursing. However, research relating specifically to intentions of leaving the profession is far less common with few studies examining actual leaving.
This cross-sectional survey of 23 159 nurses in 10 European countries employed the practice environment scale (PES-NWI) to assess the work environment and the Maslach Burnout Inventory to measure nursing burnout levels. All other factors were measured by single survey items (age, gender, education, working hours per week, perceived safety, perceived quality, nurse-to-patient ratio) or from other sources (hospital size). A multilevel regression model with leaving intentions due to job dissatisfaction as outcome and all other variables included as independent variables were conducted.
Between 5% and 17% of nurses in the 10 participating countries intended leaving due to dissatisfaction. When all countries were pooled, nurses with burnout were twice as likely (OR 2.0; 95% CI 1.9 to 2.1) to intend to leave nursing than those without burnout. This result held for the analysis by country (OR range 1.6–2.9). Nurse–physician relationships (OR 0.9; 95% CI 0.8 to 0.9), leadership (OR 0.8; 95% CI 0.7 to 0.9), participation in hospital affairs (OR 0.7; 95% CI 0.6 to 0.8), age (OR 1.1; 95% CI 1.1 to 1.2), gender (OR 0.7; 95% CI 0.6 to 0.8) and working hours per week (OR 0.8; 95% CI 0.7 to 0.9) were also associated with leaving intentions in the pooled analysis for all countries. Leadership (OR rang 0.6 –0.7), participation in hospital affairs (0.6 –0.7) and burnout (see above) were associated factors in at least 5 of the 10 countries.
The strength of this study is in its large, multinational sample using established instruments for the measurement of core constructs, such as work environment or burnout. From a theoretical viewpoint the model of leaving the profession is implicit and solely focused on the work environment without taking important context factors such as economic climate and family background into account. Although the binary single item, leaving intentions due to job dissatisfaction, has face validity, it is generally considered less robust from a measurement viewpoint. The multilevel model seems appropriate; however, the decision of including country and unit level as random effects and omitting the hospital level is not explained. In addition, indicators of the extent of clustering in the data on the various levels (eg, intraclass correlation) as well as the overall model fit were not presented.
The proportions of potential professional leavers (5–17%) are consistent with previous studies even though the operationalisation as ‘leaving due to dissatisfaction’ represents a subgroup of the overall group of professional leavers usually measured. The study also confirms associations (eg, with burnout) reported previously.5 Different levels of nurses intending to leave and differences in identified associated factors suggest cultural and health system differences between countries as important factors. However, external yet integral factors to the working life like the economic and family background, and the inherent limitations of a binary single item question in a multilanguage context, might have also contributed to this picture.