The results obtained from this work suggest that short-term absenteeism among nurses is significantly correlated with quality of care in terms of patient satisfaction, and in a negative manner, in particular in relation to MI and RS. These exploratory results involve a large sample from 25 hospital facilities, both public and private, and differing in size, in several French regions.
These results are interesting for several reasons. Firstly they enable confirmation of the hypothesis of interdependence between dimensions of performance underpinning the WHO-PATH model, at least for the "patient centeredness" and "human resources" dimensions. Secondly, they consolidate and widen the scope of previously published work exploring relationships between quality of care provided and the satisfaction of professionals in the workplace. Finally, these results reemphasize the need to explore human resource indicators as explicative factors for satisfaction data, in the French context too.
Factors known to affect job satisfaction are burnout, stress, lack of autonomy, or poor cohesion in the team [20
] and these factors are linked to inadequate organisational and managerial support [23
]. This has been shown in studies on "magnet hospitals", where it is good to work and good to be cared for [15
]. Aiken showed a link between job satisfaction on the one hand and quality of care or patient safety on the other [24
], as did Clarke [25
]. The direct relationship between absenteeism and care quality was explored by Unruh [26
]: this author shows that absenteeism in conjunction with a heavy workload leads to a significant increase in incidents reported. Workload does not on its own affect patient safety, but appears liable to do so when in conjunction with staff burnout [27
]. Overworked nurses are more tired and find it harder to cope with pressures when there are extra efforts to be made. Another study was conducted by Aiken to determine the association between increased workload and care safety. Beyond a certain patient to nurse ratio, the increase of a single patient is associated with a 7% increase in the likelihood of dying within 30 days of admission [29
]. Several authors have shown the existence of a relationship between burnout among nurses and patient satisfaction [12
]. Thus the results of a survey among patients and nurses indicated that, in facilities that were described by professionals as having sufficient staff and in which relationships between doctors and nurses was good, patients were more inclined to report that they were satisfied with their care [13
]. The key role of nurses in patient satisfaction appears to be more relational than technical [12
]. Indeed, it is nurses who connect most to patients, because they take charge of day-to-day needs. They give physical care and emotional support to both patients and families.
The present study presents numerous limitations. The study protocol was not designed in order to verify the particular hypothesis explored. This possible relationship was studied because the literature contained work on the subject, and because a large number of the hospital facilities had exhaustive data available for the indicators analysed. The PATH Project didn't include a lot a adjustment variables concerning type of professionals, hospitals or patients, and adjustment has been only based on few variables. Results have to be interpreted cautiously.
Concerning satisfaction data, the mean response rate was about 40% as expected for a postal survey [9
]. However, the risk of a selection bias exists due to the relatively low number of responding patient per hospital.
Further to this, indicators' definitions for absenteeism are still the subject of debate. Short-term absenteeism as studied here takes account of WHO-PATH specifications, and concerns medically motivated absences of 2-7 days. This measure of absenteeism is assumed to reflect the social atmosphere in the workplace and the implication of staff, while long-term absenteeism is more likely to be an indicator of health status and the effects of conditions in the workplace on health. It would nevertheless be interesting to take account of non-justified or motivated absences of one day, which could reflect poor atmosphere in the workplace. According to [31
], short-term absenteeism is an absence of less than three days. Certain publications distinguish between "approved and non-approved" absenteeism [32
]. Some calculate the days, others the hours of unplanned absence [33
]. A national working group coordinated by our team is to be set up in 2011 to define a consensus on data collection for absenteeism. A further limitation in the present study is that the data collected concerned average figures for absenteeism over one semester, and they are set against patient data from a study over 2 months, for reasons of feasibility and in compliance with WHO requirements. Future studies should integrate patient and staff data over the same period. Finally, the indicators of the WHO-PATH project ultimately aim to be used by all member countries: they therefore need, beyond their necessary validity, to be simple to handle.