The assumption behind the study was that the sources of stress would be different between qualified and N-R C nursing staff, and that this would lead to N-R C staff being more prone to burnout than their formally better qualified colleagues. In both the univariate and multivariate analyses we found few and generally small differences between the two groups with regard to both sources of stress and burnout.
This was confirmed with respect to two of the stress-source variables, where the differences were univariately significant at the .05 level: N-R C staff experienced their social relations at work to be inferior (cliques, conflict, quarrelling, lack of agreement) compared the qualified staff. The qualified staff, on the other hand, complained more about high work demands (in terms of having to concentrate, task difficulty, availability of solutions etc).
The burnout subscales did not discriminate between the qualified and N-R C staff. This means that the hypotheses that guided the study (different sources of stress and different levels of burnout) only to a very limited extent (two out of 22 sources of stress, no differences in burnout) were confirmed. Also Jenkins' and Elliot's [31
] study of stressors and burnout among qualified and N-R C staff in acute mental health services showed that the two groups did not differ much in terms of burnout, but that their sources of stress were to some extent different (adequate staffing and dealing with threatening and difficult patients).
Compared to the MBI norms[66
], the percentages of qualified and N-R C staff that exceeded the normative values were on Emotional Exhaustion (High ≥ 27, Average 17 - 26 and Low ≤ 16) were 8.9 % vs. 12.5 %), on Depersonalisation (≥ 13, 7-12 and ≤ 6, resp.) .8 % vs. 9.7 %, and on Personal Accomplishment (≥ 39, ≥ 32-38 and ≤ 31 resp.) 25.4 % vs. 20.8 %. Thus, in terms of extreme scores, the differences between the two groups were greatest on Depersonalization. High burnout scores did not in general characterise the two groups. Personal Accomplishment was an area where a substantial proportion of both groups had high scores and here high score signifies feelings of achievement. A positive self-image may counteract burnout [74
] and feelings of Personal Accomplishment may help to explain the rather low average value on emotional exhaustion and depersonalisation.
On most of the stress source measuring variables the absolute scores were low which implied that neither groups were particularly dissatisfied or strained. There were some exceptions, most strikingly on the variables Work Load (too much work, many different tasks, lack of time, too long hours), Work Demands (having to concentrate, lots of things to recall, task difficulty, accessible solutions), Influence and co-determination (influence on organisational changes, organisation of the work, economy, strategies, allocation of resource), Acting Possibilities (be able to change bad working conditions, receive help and support from the management and from colleagues), Work Related Health (need to relax, concerns, problems due to incompetence, motivation etc.) and Lack of Resources (Inadequate staffing, physical environment, training opportunities). On all these variables both groups had high ratings. This suggests a common experience among both the qualified and the N-R C groups of staff - in spite of low average burnout and high personal accomplishment scores - of having limited influence on a work situation characterised by clinical pressure and insufficient resources. Similar results are reported in other studies of mainly nurses [22
The choice of classification tree as the multivariate statistics made identification of subgroups possible. The analyses identified three groups with (to some extent) extreme negative ratings:
(i) a numerically small group (a) (N = 18) with high scores on the MBI depersonalization variable. N-R C staff was strongly overrepresented. When the high-Depersonalisation group was compared to a random similar-sized group of staff with the lowest ratings on this variable, univariate analyses showed that high depersonalization-scorers were younger, more of them lacked professional qualifications of any sort and the majority came from one of the centres. One centre had 60 % of their participators in the high depersonalisation group. Depersonalization is the interpersonal aspect of the burn-out syndrome [77
] and refers to uncaring and impersonal attitudes towards care or service recipients. As mentioned, we also found that a higher proportion (9.7 %) of N-R C staff exceeded the normative values for high scores of the MBI Depersonalization-norms, but in terms of persons the number is only 7, which highlights the centre-specific character of the problem. Depersonalisation usually occurs among overburdened personnel who receive little positive feedback/rewards [50
]. A study by Hare and Pratt found high levels of emotional exhaustion and depersonalisation among nursing assistants [78
]. According to Kanste et al [79
] leadership that encourages the employees to look beyond their own self-interest for the good of the group (so-called transformational leadership) seem to protect from depersonalization.
(ii) Qualified staff was overrepresented in a numerically large group that was low on Depersonalization and high on Work Demands. The Work Demand variable measures concentration, keeping many things in mind, task difficulty and having to deal with situations where solutions may be difficult to find). Nurses perceive stressors differently according to their grade [80
], and the obvious explanation for the high Work Demand score is that qualified staff more often have managerial positions and have more responsibility for clinical decisions. Comparing random samples of this group with colleagues who scored low on depersonalisation and on Work Demands, univariate analyses showed that the low depersonalisation/high Work Demand group had more often been recipients of violence at work and had higher scores on Emotional Exhaustion. Other studies have found correlations between high work load and Emotional Exhaustion [31
], but the Work Demand variable in the present study measures other characteristics of work than work load. Our Work Load variable showed no significant differences between the two groups.
(iii). A second small group (N = 12) also with a majority of N-R C staff had high scores on Professional Self-doubt, but were low on Work Demands and on Depersonalization. Professional Self-doubt is a MHPSS subscale that measures feelings of being inadequately skilled, uncertainty about own capabilities, doubting the efficacy of therapeutic endeavours and fear of making mistakes in the treatment. As with the high depersonalization group, extreme Professional Self-doubt was context related: 41% of this group came from one centre.
The traditional rigid workforce model is non-functional in complex interactional systems, not the least in community-based services. Thornley [29
] found that N-R Cs through experiential learning acquired substantial clinical competencies that were useful in a variety of different roles and she argues for a comprehensive re-evaluation of the competencies of non-registered caregivers. N-R C's position within the health system is said to be both central and marginal: the services depend upon them at the same time as N-R Cs are often considered to be substitutions for more valued staff. The uncertainty often surrounding the N-R C's position may reduce the interest in important professional issues such as training, career, regulation and inclusion of these persons into the health services. This may result in a situation where non-registered personnel are less appreciated because they are insufficiently trained, and - as a result - are insufficiently trained because they are not well valued. Whereas qualified personnel are responsible for the duties within their profession, non-registered personnel perform only functions that are delegated to them, and they work under the supervision and guidance of fully qualified personnel [82
]. This can result in more work-related strain, more burnout problems and less stable services.
Our results showed only small differences in burnout and sources of stress between N-R Cs and nurses. The main implication of the study is that no general measures addressed towards the N-R Cs with regard to stress and burnouts seem necessary. This has consequences for recruiting and retaining this staff group. However, the study also showed that centre-specific problems may cause more problems among N-R Cs than among the qualified staff. When N-R Cs from one of the centres reported high score on Professional self-doubt, this gave limited support to one of the hypothesis behind the study, namely that lacking formal qualifications in a highly specialised field might be experienced as stressful, and - when relevant - requires to be dealt with.
Both groups reported a limited influence on a work situation characterised by clinical pressure and insufficient resources. This means that alleviating measures should reach both groups. As we have argued above, the mental health services will continue to depend on less formally qualified staff and including these in the work force will require the development of special career-pathways, an example being the UK initiatives (NIMHE, 2005 #210}. A study by Caudhill and Patrick [83
] found that nursing assistants who were planning to leave their present jobs were younger, had been in their positions shorter, were paid less, ranked their nursing skills lower and were better educated than those who planned to stay. At least some of these factors can be met by more systematic career-planning. Better trained and educated N-R Cs can also to some extent alleviate the significantly higher Work load reported by qualified nurses. Wages, fringe benefits, job security, and alternative choices of employment are important determinants of job tenure that should be addressed, in addition to training and organizational culture [84
Limitations of Study
It was not possible to ensure that the participating countries were randomly drawn from all EU countries, nor that therefore their services were in any sense necessarily representative of the EU as a whole.
Thus it is difficult to generalise as to what degree any of the centres were representative of typical working conditions operative in their own countries on a wider basis.
Whilst the MBI already was translated reliably into all the languages pertinent to this study, this was not the case for the ODQ, PWSQ or the MHPSS. While it was anticipated that back translation would occur during the course of the project, the complexity and cost of the task could not be born by the project's funding. However, all of the key researchers were English speakers as a second or third language and in one site, Bodo (Norway), the researcher was a native English speaker. Consideration must also be given to the sample sizes achieved for each site and team type.